FR2839894A1 - Integrated radiotherapy equipment for obtaining instant diagnostic images, comprises five sources of photon beams on rotating frames and six sources of photon beams on fixed porticos - Google Patents

Abstract

The equipment has five or other odd number of photon beams sources mounted on rotating frames guided by 4-D radioscanography and used for multi-beam orthovoltage isotropic cyclotherapy. These beams are used in combination with six sources of photon beams (8,9,10,11,12,13) mounted on fixed frames (15,16;17,18;19,20) which may however be rocked laterally.

Description

<Desc / Clms Page number 1>

Methods, image-guided Cyclotherapy apparatus and mode for obtaining instant diagnostic CT images for online planning and dosimetry
TECHNICAL FIELD The present invention relates not only to a plurality of volumetric computed tomography apparatus and to continuous guided radiotherapy by instantaneous isotropic multislice radioscanography and radioscanometry, whose rotational gantry includes a number of odds of multiple sources associated with a radiotherapy system. control imaging of the dosimetric characteristics of the processing energy; but also methods of implementing the acquisition of the anatomical data and the contours of the target volume, with a view to a ballistic optimization of the multiple simultaneous external beam delivery of the therapeutic irradiation dose as well as of repeated verification. by 3-D dynamic scanning methods, for the purpose of highly accurate conformation radiotherapy, clinical simulation and behavioral modeling.

2. State of the art Radiotherapy is, from the point of view of the only technical concepts, in constant evolution was only because of the necessary adaptations related to the homogenization of the regulations and the development of the techniques of irradiation. But evaluations of the actual effectiveness of radiation and the development of long-term radiovigilance do not always go along with the first finding, if it is not rather obvious that the marginal costs of this quest for perfection. The therapeutic arsenal of cancers uses, in about 60% of the cases of which approximately 2/3 are currently cured, the radiation treatment. This nonetheless constitutes a step reached for a while now. It is nevertheless possible to go beyond this therapeutic ratio, notably by improving the ballistic aim and the difference between tumor dose and dose to the surrounding tissues not involved in the tumor process. It is the concept of Conformational Radiotherapy that has given birth to this new hope.

The objective of the present invention is precisely to leave on the basis of the whole

<Desc / Clms Page number 2>

 new concept of CT guidance of the Bilinac Imatron device of the patent filed a year ago under No. 01/01133 to further reduce, with even more precise radiation treatment techniques, the range of dose fluctuation in specific regions of the body around a given target lesion, given that high-dose X-rays have the potential to reduce mortality and morbidity in a large number of patients; this, provided that the delivery of the rays is as accurate as possible. The source of the radiation may be in the form of the external beams of the ionizing particles or better, from the point of view of the radiation energy deposit, the radioactive sources internal to the patient. Only external beams, usually produced by machines acting as particle accelerators, telechiotherapy or formerly orthovoltage X-rays, have an interest in the description of the clinical simulation and the physiological behavioral modeling of the treatment.

External beam radiotherapy (x) is performed with several types of ionizing radiation, of which approximately 80% of patients are photon-treated. Nevertheless, the tumor residues of this photon therapy are generally treated with electron overprinting. There are also facilities for neutron therapy, proton therapy, and even heavy ions, which, as particle accelerators, have been used around the world, and their latest, very expensive sophistications are further accentuating the concentration along with the shortage of logistics. sanitary facilities to cope with cancer. Providing, in the face of the burning questions of reducing health costs, to this state of relative deficiency is another of the objectives of the development of a tool goes everywhere and a reasonable price of acquisition and maintenance. We also know that the major obstacles to the widespread diffusion of radiation therapy are, besides the cost price of the acts, the operation of the facilities and their dependencies, a perfect technique and a precise dosimetry.

Moreover, the role of the prediction of the best treatment conditions is in today's art of computer-assisted planning of radiation treatments. This process of treatment preparation consists first of all in characterizing the patient's individual anatomy (this is more often done using not only a separate computed tomography (CT) examination, but also determining its shape, intensity, that the positioning of the sources of radiation with references to data from different sites and devices, and then calculating after all on another

<Desc / Clms Page number 3>

apparatus the distribution of the dose of radiation absorbed by the patient). To calculate the dose in the patient's body, the most commonly used methods are based on measurements of the dose made in a water box. Heterogeneities, such as bones and airways, are treated roughly or are simply ignored. There is therefore no complete symmetry between the simulated situation and the reality of what is simulated, nor between the representation that one strives to make and the real. We can not predict accurately and accurately the result, neither in time nor in space, of what we have implemented!
2. 1. Concept of image-guided therapy The concept of image-guided therapy of art includes a set of tasks that are performed, with regard to voluntary or involuntary movements of the objective, both pre- and intra- -operative in order to achieve a true physiological behavioral modeling. These tasks include acquiring images as well as their variability over time, description, treatment planning, and navigation into a prospective PACS network. The concept of guided image therapy, on the other hand, requires CT or NMR tomodensitometric sections to be obtained, usually in forced apnea.

Situation that we know is not physiological! On the basis of 3D images, obtained preoperatively, these digitized images delivered from the scanning devices to specialized consoles are then used in a later stage to guide the intervention in the different operating rooms or in the radiotherapy rooms. It is obvious that a successful daily application of this technology requires a specialized infrastructure that shares many components with a departmental archiving system associated with a communication system (PACS). Under these conditions, the planning of the therapy requires a description and an analysis, which are performed on different consoles working in a network, according to the specific requirements of a given case. The performance in the operating room or radiotherapy of image-guided therapy is performed with navigation instruments that harmonize the patient's therapeutic plan in a controlled environment. Multiple image overlay and interaction with specialized consoles are integrated into the clinical routine. Among the new innovations, considered the most edifying and the most important, there is the real-time fusion of ultrasonographic (US) images with 3D models of the patient's geometry as it is performed with the

<Desc / Clms Page number 4>

 CT sections, NMR images, or the fusion of the two. US images generally serve to refresh the database of stored images and require to be useful, according to John W. Haller et al. (2001), an instantaneous multimodality fusion.

2. 1.1. - Image-Guided Networking Systems The modern Radiology Departments have CT and NMR scanners. CT and fast gradient multi-slice CT imaging devices are used to obtain images during pre-operative examinations.

All these data are transmitted via DICOM interfaces (digital imaging and communications in medicine) to PACS archives and then in the local area network (LAN) for image-guided therapy. The connection of imaging devices and PACS interfaces is, in a practical perspective, the most important aspect of image-guided therapy, especially for surgery. Almost any imaging device, when properly connected and interfaced, can provide the data needed to perform an image-guided therapy, once an appropriate intervention protocol and calibration procedures have been established. This is why Radiology technicians must be, according to John W. Haller (2001), informed of the strict requirements of the protocol of intervention intended to reduce the interference of human error, or even that joint of the machine or machines.

Alternatively, for direct measurements of the therapeutic irradiation dose, Monte-Carlo transport is, according to Chandler et al., Currently considered to be the most accurate method of the art for determining dose distributions in heterogeneous media. In this Monte Carlo transport method, a computer is used separately to simulate the passage of particles through an object of interest (William P. Chandler et al .: PCTIUS 97/03328). It is the effects of radiation on the body which, as a dose of radiation, are thus quantified.

The dose of absorbed radiation is therefore defined as the ratio of energy deposited per unit mass of tissues. Since tumors and sensitive structures are often located in close proximity, the computational accuracy of dose distributions is therefore eminently important, as the results depend on them. The goal of radiation therapy is therefore precisely to deliver a lethal dose to the tumor, while maintaining an acceptable level, if not much lower, in the surrounding sensitive structures. If this goal is easily achieved with interstitial or intracavitary brachytherapy,

<Desc / Clms Page number 5>

 is far from it with the external radiation beams, where the philosophy of weakening the depth of the dose abuts on a certain number of the radiobiological limits, to allow by the fractionation the occurrence of the tissue repair whose kinetics is higher for tissues. The object of the present invention is therefore to achieve irradiation conditions close to or better than interstitial or intracavitary brachytherapy.

2. 1.2. The advent of radiotherapy imaging in radiotherapy Radiation therapy has long proved its effectiveness in destroying certain tumor cells and currently includes in its arsenal, during evaluation or development, intensity modulation and tomotherapy. .

The other side of the coin is that it also works to the detriment of healthy neighboring tissues, so side effects are sometimes so severe that they stop treatment. A pitfall that can now be in the art bypassed by modulable intensity radiation therapy to which we will return in more detail.

 Knowing that in Dynamic Radiation Therapy the distribution of the dose delivered to the patient depends not only on the characteristics of the external bundles used but also on the shape and composition of the body in the vicinity of the treated area. It is only since the appearance of the scanners that one is able to know precisely the form, the position and the density of the heterogeneities around the target. We must certainly take advantage of these new tools; but it is also necessary, as we will realize below, that the conditions of acquisition and exploitation of the CT images respect specific rules of perfect symmetry between the object and its representation. It is also known that the methods for determining the dose are still based on geometric calculations where the depth of the points of interest is expressed in "water-equivalent thickness" obtained from the geometric distances corrected by the electronic density.

These calculations require, as of today, the preliminary (static) contouring of the external surface of the body and the determination of the main heterogeneities in the field. On more recent systems of the art, it is the matrix of voxels constituted from the images obtained in adjacent sections which is, as such, exploited at a later date. Other forms of imaging, magnetic resonance imaging (nuclear) in particular, are only complementary to improve the identification of target structures and organs at risk, but they do not provide information as relevant for the calculation. doses than those

<Desc / Clms Page number 6>

 obtained even by deferred scanning. Indeed, coupled by the techniques of registration and fusion with CT images, the heterogeneities are of considerable interest, but still remain situations judged at a distance from each other by not integrating into the concept of conformation the dynamic notion. fundamental evolution over time of all these clinical situations.

2. 2. - The technical concept of Conformal Radiotherapy Dynamic conformational radiotherapy requires otherwise a perfect equipment and methodology allowing, with a view to optimizing the ballistics and the therapeutic strategy, precise materialization by an infinitely adjustable millimetric ballistic study. in situ of the target rather than that traditionally superficial of the prior art which proceeds, to achieve the so-called crossfire technique, by entrance gates of the beams of a single beam oriented differently in time. By deviating from the rigid and inflexible ballistics of current art, we have introduced the simultaneity of the several synchronous beams of therapeutic irradiation with the so-called Bilinac Imatron device of the patent n 01/01133. This is why, having also placed radiological semiotics as an infrastructure of the radiological sciences, we do not see a good way of slipping towards the virtual whole of ancient art, ignoring the real object as well as the necessary symmetry of the object representation transformation. This question is at the center of the confrontation of the precise object, the lesion to be destroyed, with its representation, the target volume, geometric place and place where the formalism of art is elaborated at the same time as a theoretical articulation is constructed. to the significant practice studied; this essential question whose nonresolution greatly reduces the scope and the scientificity of the so-called virtual simulation, as it is practiced today; because, there is a prerequisite: that of having to recognize one day a certain vulnerability, namely that we are, with all this arsenal of Radiotherapy, aiming at a constantly shifting target and that we must give ourselves accordingly. means, not to suppress what is physiological, movement, but to quantify it exactly at any time to act accordingly with more precision. This new definition introduces a dynamic character: a transformation, an evolution in space and time. It's physiological behavioral modeling that he

<Desc / Clms Page number 7>

 is in these words. Now, in the form of a multitude of algorithms, only the self-censored mathematical formalism of current art subsists on frozen impressions of a given moment. In what follows, we will indicate, without the counterpart of the precise objectivation of the target, some moments of the simulation operation of the ancient art that led him or did not lead him to the place we have just to sketch, namely the aim without the precise materialization of the targeted target whose most immediate risk is the sighting bias, to which we will return in more detail. The simulation must nevertheless seek, as in the example of the modulable intensity irradiation, IMRT or RIM, described below, a representation of an object, the target volume (in the sense of the Stoics, a sign which establishes a relation between two terms, one of which evokes the other in a singular way, provided that these terms are concomitant, so it is not at all arbitrary that we find this symmetry and this precision), in order to optimize the ballistic aim and to update, as we will also see, the possible presence of sighting biases (see figure la).

2. 2.1. - Modulated intensity beam techniques (IMRT or RIM) A computer divides, in all these techniques, the area to be irradiated in fields, themselves subdivided into segments, and simulates for each field and each segment, all the possible irradiations. A tailored irradiation plan can thus be established by increasing the intensity of the radiation passing through the tumor and decreasing that of radiation passing into healthy tissue. This modulable intensity radiation therapy (or RIM) would thus deliver exactly the indicated dose where it is needed and wanted.

While modulated intensity beams now offer many new possibilities, in modern radiotherapy, to optimize treatment results.

A modulated intensity beam has a prescribed intensity variation (actually a dose variation) over the cross section of the radiation field. Dynamic radiation therapy is a way of achieving such variations in dose distribution within a field. One or more parameters of the dose-delivery system are varying during dose administration. Examples of such parameters are the collimator (blade) position, the blocking position, or the position of a brush beam within a field. The time that the field or the beam is in a certain configuration is prescribed by a weighting factor. The advantages of a modulated intensity beam, and thus of a dynamic beam delivery, do not

<Desc / Clms Page number 8>

 of course not only apply to photon beams, but also (and even especially) to electron beam, proton beam, and heavy ion beam applications (E. Pedroni et al, 1995, Th. Habarer et al, 1993). They make it possible to adapt for the first time in a precise and dynamic way to the three dimensions, without however making the part of the voluntary and involuntary mobility of the patient; to protect healthy tissue, the irradiation field shaped by low-melting point alloy blocks or multi-blade collimator conforms to anatomical structures and tumor size.

2. 2.2. - Dosimetry of conformation and verification of the absorbed dose
It is to be expected that by using a dynamic treatment technique, and especially with proton beams, the shape of the total distribution of the dose delivered in a single field may be rather complex. A detailed knowledge of the dose distribution is therefore essential to harmonize many of these fields with the distribution of the desired total dose. The purpose of such a ghost verification is threefold. First, it can be used to control whether the shape and magnitude of a complex dose distribution in a homogeneous phantom agrees with the outcome of the treatment planning. For various reasons, the actual dose distribution may differ from that calculated. Second, the goal may then be to use this system to understand the effects that may be responsible for such deviations, so much so that the accuracy of the dose calculations can be improved. Thirdly, phantom dose verification can be used as a means of implicitly checking the operation of the beam delivery system, if the performed dose has actually been administered as prescribed.

Since standard dosimetry systems are not very suitable for dynamic beam delivery systems. Dosimetric measurements are usually made by means of an ionization chamber placed at various positions in an aqueous phantom. However, since dose delivery varies over time, a measure of the three-dimensional dose distribution is very time-consuming, since for each measuring point the entire beam delivery sequence must be repeated. . Another disadvantage of such methods is that an error in beam delivery, if it caused only a dose error at a distance from the measurement point, is not detected. The arguments above show that a simultaneous measure is, in many respects, desirable.

A quantitative impression can of course be obtained using a film (W.L.

<Desc / Clms Page number 9>

McLaughlin et al, 1976; S. M. Vatnitsky et al, 1997; I. Daftari et al, 1999). However, other approaches in favor of uni-, bi- or three-dimensional dosimetric measurements, such as a linear array of 1-D or 2-D diodes or ion chambers (HH Liu et al, 1997; CP Karger et al, 1999) stacked, multi-anode ionization chambers (C. Brusasco et al, 1997), and 3-D MRI gel dosimetry (Oison et al, 1990, MJ Maryansky et al. , 1993) are currently being investigated as well.

The instrument discussed by SN Bonn et al. (1998) is for example a modern equivalent to 2-D film dosimetry. It consists of a scintillator or screen
Gd202S: Fluorescent Tb, mounted perpendicular to the beam, on the beam exit side. Light from the screen is detected by a low dark CCD camera with long integration time capabilities. The light output observed at one location on the screen is proportional to the dose in the phantom of a stack of equivalent water material of adjustable thickness. This system allows an integrated measure of the time of the dose distribution simultaneously in both directions. SN Boon et al (1998) demonstrated that this system is for accurate relative dosimetry in a static beam delivery system. Furthermore, dynamic systems, such as the screen + CCD system tested in particular in Switzerland, in real conditions, as an additional dosimetry device, independent of the scanning mode of the beam. The Paul Scherrer Institute (PSI) uses a spot-scanning technique with 200 MeV proton beams at the stand. The CCD-scintillator system is mounted on the patient's table (E. Pedroni et al, 1995). Component details and system characteristics were reported in 1998 by SN Boon et al. (1998). In the configuration discussed here, a pixel in the image (which consists of 4 CCD encapsulated pixels) corresponds to less than expressed differently at approximately 1.2 mm2 on the screen.

Examples of inhomogeneous dose-to-measure distributions have been discussed with applications of the system, including inhomogeneities in the dose distribution. The sufficient combination of sensitivity and spatial resolution is needed to detect small distortions in a dose distribution. The verification of the calculated dose distributions is as follows: The results of the measurements of the lateral shape of the brush beam at the different arrangements of the scanning magnet demonstrate the applicability of the CCD-screen system as an exact tool to curb ionic optical aberrations in the

<Desc / Clms Page number 10>

 beam delivery system. A shape of the brush beam, exactly known for each disposition of the scanning magnet, can be incorporated into a treatment planning program, but not during a treatment, to increase the accuracy of the dose calculations. The scintillator-CCD system finally proved to be a useful tool in this team for dosimetric verification of proton scanning beams. Measurements with this system have thus been found to be complementary to calibration measurements with a reference ionization chamber, which requires, however, a homogeneous dose distribution at the point of measurement.

2.2.2.1. - MLC vs. Conventional Alloy Blocks Since February 1995, the dosimetric and clinical implications of the Clinac 2100 C / D Linear Accelerator multi-leaf collimator used to replace conventional low-melting alloy blocks have been studied in Turin. The festooning effects of the isodoses and the effective punctures produced in an irregular field by the multi-leaf collimator were analyzed in a precise way. Radiation leakage through the tungsten plates was measured and compared to the values of the low melting alloy blocks.

Different dosimetric steps were followed to obtain the units / dose ratio of the monitor. Because of the irregular fields protected by the MLC, the unique measurements also showed several physical and dosimetric disadvantages in relation to the effective upper penumbra than with the conventional protection when the angle between the edge of the field and the normal at the direction of the path of the blade was 45. In clinical practice, MLC can be widely used for conformal radiotherapy of pelvic and thoracic tumors. While low-melting blocks and alloys can be replaced by the Multi-Blade Collimator (MLC) for radiotherapy of selected cancers of the brain, head and neck.

It has been established that frequent use of a multi-leaf collimator improves the accuracy and efficiency of radiation therapy and reduces the time for each treatment dose, potentially increasing the number of patients treated each day. The multi-blade collimator is now considered an important technical tool either to replace conventional protection for static conformal radiotherapy or to deliver a planned dynamic 3-D radiotherapy. The target localization conformation as well as the planning dosimetry of the treatment of Conformational Radiotherapy were initiated in

<Desc / Clms Page number 11>

 using NMR imaging. A detailed methodology was developed to verify three-dimensional radiation dose (3-D) mapping of treatment planning for conformal radiotherapy with FeMRI dosimetry, ie a phantom, consisting of a human skull. filled with a 1-mM ferrous sulfate solution and 0.1-N sulfuric acid gelified with 7.5% (by weight) of gelatin was employed. With a spherical target volume in the phantom of the head, five non-coplanar conformational fields were designed through the use of a homemade 3-D treatment planning system. The phantom was irradiated with a 6 MV linear accelerator at a total dose of 25 Gy, delivered at the periphery of the target volume. The phantom and a series of calibration vials were simultaneously scanned by a 1.5T GE imager with six different multiscan sequences in inversion-recovery. Tl values were evaluated on a pixel-by-pixel basis through the use of special ad hoc software on a UNIX console and were converted into a dose using the calibration data. Comparisons of dose distributions between those measured by FeMRI and those calculated by treatment planning by these authors (M. F. Chan and K. M. Ayyangar, 1995) showed good agreement.

2.2.2.2. - Time tracking of tumor mobility In another study by Shirato et al. (2000), a new radiotherapy system and its treatment planning system have been developed and used by these authors in clinical practice, to perform precise three-dimensional (3-D) conformational radiotherapy of mobile tumors. A linear accelerator synchronized with a real-time tumor tracing fluoroscopic system, by which the 3-D coordinates of a 2.0 mm gold marker inside the tumor can be determined every 0, 03 second, has been developed. The 3-D ratios between marker and tumor at different respiratory phases are evaluated, using CT image at each breathing phase, whose optimal phase can be selected to synchronize with irradiation (this is what we call 4-D treatment planning). The linear accelerator is triggered to irradiate the tumor only when the marker is located within the region of the planned coordinates relative to the isocenter.

The coordinates of the marker were detected, during radiotherapy, with an accuracy of 1 mm in the phantom devoid of experimental movement.

The time between recognition of the marker position and starting or stopping the megavoltage X-ray irradiation was 0.03 seconds. 14

<Desc / Clms Page number 12>

 Patients with various tumors were treated with conformational radiotherapy with a projected volume margin (PTV), well adjusted. They were surviving without relapse or other complications with a median follow-up of 6 months. Real-time tracer fluoroscopy radiotherapy following 4-D treatment planning was developed and found to be feasible to improve the accuracy of radiotherapy of mobile tumors. It is the temporal dimension of the radiotherapeutic practice which appears here in broad daylight and which will henceforth have to be taken into account.

2. 2.3. - Modulation of the beam by the multi-blade configuration (MLC)
In order to optimize, moreover, conformal radiotherapy by external beam at the level of the thorax. It should be possible, it is recommended in the art, to minimize the physiological and non-physiological movements of the patient during the treatment. For treatment on the upper torso, the target organs are known to move substantially due to the breathing of the patient or even transmitted cardio-vascular beats. The technical aspects of synchronization of the beam of Radiotherapy with the breathing, namely: the optimal system to control the breathing, the measurements of displacement of the organ and the synchronism (gating) of the linear accelerator was studied by HD Kubo and BC

Hill (1996). Several respiratory sensors including a thermistor, thermocouple, strain gauge (strain) and a pneumograph were examined to find the optimal sensor. The size of the breast, the chest wall, and the lungs were determined using the X-ray radiofluoroscopic images played back on a VCR during routine Radiotherapy simulation. Total dose, beam symmetry, and beam uniformity were examined to determine, by synchronism or gating, the effects on the Varian 2100C linear accelerator.

The dosimetric characteristics of a multi-leaf MLC collimator on this 6 and 18 MV dual energy accelerator, installed in the Department of Radiotherapy of Mauriziano Umberto I Hospital in Turin, starting its use in clinical practice. The measures included in particular the transmission through and between the slides as well as under the junction of the closed slides, the percentage of depth dose on the central axis, the flow factors and the effective penumbra (M. Stasi et al, 1999). The MLC installed on the dual energy accelerator, Varian Clinac 2100 C / D (6 and 8 MV) used is an added component placed under the standard jaws; it consists of 40 opposing pairs of tungsten blades of

<Desc / Clms Page number 13>

 5 cm thick, controlled by computer, each projecting a width of 1 cm to the isocenter, and provides a maximum treatment field of 40 x 40 cm2 to 100 cm of SAD. Values for transmission, darkness and festooning were measured with the standard radiographic film used routinely. A laser scanning photodensitometer (WP102, Wellhofer) with a 450 micron spot was used to obtain the optical density and the profile of the relative gamma dose to determine the dose. X-ray films had been calibrated with an ionization chamber by irradiating samples; this calibration was used to correct the scanner readings of the dose film. The percentages of the depth doses were also measured in an automatic aqueous phantom (WP600, Wellhofer) for irregular fields defined either by the MLC collimator or by alloy blocks, in order to test the differences in the acrylic region. Measured and calculated flow factors were compared for some irregular fields defined by the MLC collimator. This comparison tested the accuracy of the CadPlan 3.1.1 Treatment Planning System 3-D algorithm. Varian Dosetek.

For both energies, approximately 2% of incident radiation on the MLC is transmitted and an additional leakage of 0.5% occurs between the adjacent blades. The leak under the junction of closed blades is considerable: about 25-33%. The relative relative dose curves are similar for the two fields shaped either by the MLC or by the conventional jaws. The dose to the skin with the MLC shaped field is less (3.5%) than that of the cerrobend shaped fields. The treatment planning calculation procedure can be applied to the MLC (the difference is less than 1%). The effective penumbra in MLC shaped irregular fields is 11mm on average, which is slightly wider (2-3mm) than the dim light of conventional cerrobend blocks. It is established that the effective penumbra increases with the depth, the width of the field and the positioning of the blades.

The hope aroused in the current state of the art by the MLC, which, when used properly (the rotation of the collimator, the position of the jaws and blades, the high number of fields), can be applied with good results to conformational radiotherapy. MLC is considered better than conventional cerrobend blocks in improving reproducibility and accuracy of dosimetric treatment as well as dose and skin dose. The use of the MLC to modulate the beam fluence (IMRT) makes it possible to modify the intensity of the beam, to improve

<Desc / Clms Page number 14>

 volume treated and surpass, modeling the dosimetric limits of static therapy.

The experimental demonstration of the delivery of a 3-D distribution of the dose of Conformal Radiotherapy was made by T. Bortfeld et al.

(1994), using the nine-field full-field modulation profiles of a fixed stand. Two-dimensional, field-to-field, full field, two-dimensional modulation profiles were made, using the prototype of a multi-blade collimator, by commercially available quasi-dynamic multi-blade collimation installed on a linear accelerator medical. The profiles were calculated to deliver an optimal dose distribution in a patient with prostate carcinoma. The surface of the target volume was invaginated and bifurcated. The calculated dose distribution was delivered to a homogeneous polystyrene phantom, consisting of 1 cm thick slices that were cut to equalize the patient's outer contour. In this study, seven therapy verification films were placed between the slices of the phantom.

 Analysis of these films revealed a high degree of conformation, which would not be possible with the conformational therapy of the unmodulated dose region to target form. However, the small spatial displacements of the observed dose distribution confirm the need for very accurate placement. It is therefore . It is possible clinically to deliver relevant three-dimensional dose distributions, which conform to invaginated and bifurcated target volumes, using fields modulated by multi-leaf collimators. And in the case of a Bilinac-Imatron device, a secondary diffusion mirror modulation of the particles of the two simultaneous beams emitted from the two irradiation heads can be envisaged. synchronization, with an even better result as to the simultaneous double modulation of the two therapeutic irradiation heads.

2. 2.4. - Dosimetric monitoring by dedicated electronic systems
Knowing that the escalation of the dose requires in Conformational Radiotherapy an exact placement of the fields, gantry electronic systems, Electronic. gantry imaging devices or EPIDs, have been developed and are used to verify the location of the fields but remain diagnostically limited by the low contrast of the subject's bone anatomy, the large imaging dose and the small size of irradiation fields at megavoltage (MV) energies. Adaptation of a linear accelerator to obtain a radiographic and tomographic location

<Desc / Clms Page number 15>

 bone and soft tissue targets in a reference structure of the accelerator. This system clearly separates the verification of the delivery of the beam, that of the machine installations, that of shaping the field (machine settings, field shaping), that of the location of the patient and that of the target.

 At this time, there are tentative attempts at real-time dosimetry on dynamic images of dosimetric control by gantry imaging of radiotherapy, as reported by D. P. Huyskens (Belgium) (2000). In this last article, the physical problem of the relationship between the dose at the exit of the patient and the dose at the gantry detector is presented in three examples of dosimetric control by gantry imaging: the in vivo dosimetry in the axis of the beam, the two-dimensional dosimetry profiles and finally gantry dosimetry with intensity modulation. It turns out that this type of dosimetric control by gantry imaging of a radiotherapy is in principle interesting, because the dosimetry of the gantry image contains more information on the quality of the treatment than the conventional methods of dosimetry in vivo, ( either by semiconductor detector or by thermoluminescent detector) which are always punctual and give a one-point estimate of the absorbed dose, whether it is the dose at the entrance or the dose at the exit or the intracavity dose. (R. Boellaard et al., 1997, Bogaerts et al., 2000a). It is even possible to combine in this way the dose at the entrance and the dose at the exit, to estimate thanks to more or less sophisticated algorithms the dose at the height of the center of the volume occupied by the patient, or at the level of the target (R. Bogaerts et al, 2000b, Essers et al, 1995a). This two-dimensional detector offers, by definition, the possibility for gantry imaging to study and. control dose distributions, which could be measured either in relative mode or still in absolute dose, that is to say in Gy.If we know the physical relationship between the dose at the exit of the patient and the dose at the level of the film or the electronic imager, and if one controls the calibration of the detector, determined by gantry imaging, the dose in the axis could replace the measurements of dose with the. exit. For the treatment with static beams, gantry imaging can in fact give additional information that can not be determined by the standard in vivo method on treatments concerning the orientation of the corner filter, knowing that the measurements on the axis are not very informative, whether at the entry or exit of the beam, on the orientation of the filter. It is better not to take into account the filter, from a dosimetric point of view, than to use one, whose

<Desc / Clms Page number 16>

 tation is false. Secondly, gantry imaging can also - apart from its use for visual control - serve to optimize treatment. In this case, let's leave aside the process of quality control stricto sensu, to value the dosimetric information of the detector which is integrated in the process of treatment. The well-known example is the use in the breast treatment of the control cliché information for the development of individual compensators (M.

 Essers et al., 1995b). It would also be possible to optimize, according to Essers et al.

 (1999) the choice of a wedge filter, using the dosimetric information of gantry imaging.

 The interest of dosimetric control by imaging will, however, be reinforced in the future by the use of the method of the Modulation Intensity in Radiotherapy (IMRT or RIM). The beams used in IMRT, be it the step-and-shoot sequences of about ten segments or those of the sliding-window type, are so complex that the in vivo point measurements only inform the dosimetry quality of the treatment. . Gated dosimetric control is currently considered the only one capable of verifying whether the dose distribution varies, both in time and in space, in accordance with the planned plan. However, the number of degrees of freedom increases, as in IMRT, with the number of segments, one is therefore necessarily faced with an increased need for dosimetric control of these sophisticated beams. It is obvious that, from an ergonomic point of view as well as from the point of view of the visual control of the treatment, it finally makes more sense to use the same detector that does not accumulate measurement errors. Practical difficulties in the use of dosimetric control of gantry imaging, however, are evident. This is, on the one hand, more specifically the physical problem of the relationship between the dose at the exit and the dose at the level of the imager and, on the other hand, the calibration of the detector. The film dosimetry shows, moreover, a good correlation between the ionometric measurements, thanks to an ionization chamber and densitometric measurements, converted into relative doses, obtained from photographic films. In these specific cases, the dosimetry information obtained by the detector can, according to R. Boellaard et al. (1997) to be geometrically retroprojected at the patient's exit, in order to estimate the distribution of said dose at the exit, which theoretically lies within the patient, that is to say, according to the definition, at the distance dmax of the output.

Subsequent studies, on the other hand, have proved that this simple relation or rather this identification, in the case where the inhomogeneities are important by

<Desc / Clms Page number 17>

 relative to the thickness of the phantom or where the thickness of the irradiated phase would vary substantially, would no longer be valid. In the irradiation of an orthorhombic phantom containing large aluminum particles and air cavities, a variation exceeding 5% is observed, when one compares the dosimetric profile obtained at a short distance from the profile obtained at a great distance. Moreover, for inhomogeneities at low density, ie with air cavity, the densitometric profile overestimates the dose at a large distance, whereas for high density inhomogeneities, such as aluminum, the subemptimal dosimetric profile the dose relative to the dosimetric profile at the exit of the ghost. These results were corroborated by the Amsterdam group (NKI) by measurements on patients in a publication on the use of the imager, as a quality control tool and the dose measured at great distance with the detector (C. Fiorino et al, 1996), which were in the order of 8% on patients treated for lung cancer.

In the case where the thicknesses of the treated region vary significantly, as in the treatment of breast cancer or head and neck, the physical phenomenon is accentuated. To solve this problem of dose variation, depending on the distance, there are only two solutions: either the dosimetric profiles are measured with the detector at a short distance, or an algorithm is developed that takes into account the differential behavior of the detector. radiation scattered behind the patient. The first option is not always easy to achieve. For example, in the lateral bundles of the head and neck treatments, the shoulder constitutes a geometrical obstacle. The last option is more reckless, because most algorithms and software offer to treat the dose distribution inside the patient and not behind him.

 One could also introduce into a software a large air cavity behind the patient. Current software is not in this kind of manipulation fully satisfactory.

Several research groups have, over the past decade, studied this problem and have begun to look for formalism that may, in the form of algorithms, control the phenomenon in a quantitative way or bypassing it empirically. It is interesting to note that for this purpose two approaches are possible. Starting from the distribution of the dose obtained by the detector at a great distance, one can develop a method to recalculate the dose distribution at the exit. One can also use a method estimating, from the distribution of the dose at the exit, the distribution of the dose at a distance. These are where

<Desc / Clms Page number 18>

 two different schools. In a quality assurance program, the choice is eminently philosophical. The three methods currently practiced in Europe are as follows: In the approach to the center of Rotterdam (Daniel den Hoed), according to A. Rizzoti et al. (1985) information from the dose planning system to predict at the imager the dose distribution. More specifically, the patient's CT sections are first converted to water equivalents (in the presence of inhomogeneity, such as the lung, the patient is virtually compressed). This compression or dilation makes it possible to define, for each point at the exit of the patient, a transmission, separated into primary transmission and broadcast transmission. The change in transmission scattered at a certain distance behind the patient is then estimated by an empirical kernel, determined using different ghost measurements where three parameters are considered: the thickness of the phantom, the extent of the field and the space between the ghost and the detector. By recombining the long-distance transmission of the scattered radiation, an estimate is obtained at the detector of the dose distribution.

While the method of J. van Dam and G. Marinello (1994) of Amsterdam (NKI) starts from the dosimetric information contained in the imaging, to go back to an estimate of the dose distribution at the exit of the patient. even in a plane inside the patient. The method assumes that the image is taken more than 50 cm from the patient's exit. Under these conditions, it can be shown that the scattered radiation that contributes to the image is relatively small (less than 5%) and is distributed more or less homogeneously. In this case, the scattered radiation can be subtracted from the image in order to obtain the primary dose.

 Then, the primary is corrected according to the inverse of the square of distances. The convolution is done with a diffuse kernel, determined experimentally, to obtain the dose at the exit. This does not require, according to these authors, CT data.

In the Leuven approach (J. van Dam et al, 1992), finally, the information starts from the dose planning system to predict the dose distribution at the gantry detector. The dose at the exit is separated into primary dose and diffused dose. The primary dose is modulated, following the inverse of the square of the distance, and the diffused dose is modulated by a new analytic function that has been experimentally controlled. This function admits that the number of particles emitted by a set of secondary sources, located inside and at the exit

<Desc / Clms Page number 19>

 of the patient cross the surface per unit of time, remains, by analogy with the Gaussian theorem in electromagnetism, the same across a larger area corresponding to the greater distance. In other words, the flow of secondary particles through these surfaces is constant. This constancy of the flux implies that the ratio of the surfaces to shoot, as a function of the distance to the output, the necessary information on the decay of the scattered radiation. From the ionometric data, in the axis of the radiation behind different ghosts, an analytic function is established. This function has only one parameter, determined experimentally, according to the energy, and makes it possible to estimate the contribution of the scattered radiation at any distance.

The detector is calibrated here with photographic films for dosimetric use. The photographic detection, which is very accurate for the determination of the relative dose, is less suitable for the measurement of the absolute dose in Gy. It is essential to establish beforehand a quality control program, not only of the developer, the fixer and densitometer, but also that of the software. It must then be ensured that the films used in the dosimetry study belong to the same batch, that is to say from the same production line having the same badge. With regard to electronic imagers, depending on the type of imager, different image correction factors must be applied to obtain a relative or absolute dose distribution. An example of the use of gantry imaging to replace in vivo output measurements has recently been published by J. van Dam et al. (1992). This last study demonstrated the possibility of using the optical density of the control plate, to estimate from a dosimetric point of view the quality of the treatment of the breast.

Gantry imaging can be used to replace so-called in vivo output measurements. However, the accuracy of these measurements is lower because we must add errors introduced by the algorithms to recalculate the dose at the gantry detector from the dose to the output or vice versa. Also, this kind of measurement is not, on a larger scale, possible, provided that the process is well or very well automated, both in terms of detector calibrations, and that of the distance of the detector. detector relative to the patient, and that of the software, etc. The simple two-dimensional portal dosimetry would effectively identify in the treatment, as has been shown above in the treatment of breast cancer, errors that do not appear in conventional in vivo dosimetry (for example, a poor orientation of the corner filter). This kind of mistake can very well be

<Desc / Clms Page number 20>

eliminated, by applying to the system additional filters of record and verify (Varis) without having to treat the patient, at the first session, according to an inadequate procedure. Portal dosimetry of this type involves a statistical approach, where the standards of good treatment are defined. For the development of individual compensators, for example, the question is whether the use of portal dosimetry is the optimal way to develop a compensator or simply to control a compensator well?
Some schools do not advocate that the development of compensators or the choice of a wedge filter is the responsibility of the system of planning doses or modules associated with the system of planning doses. Ideally, physiological imaging of diagnostic quality should be used to visualize.

2. 3. - The advent of physiological imaging
One can rightly argue that functional imaging has been around for a long time, namely: in nuclear medicine, where gamma cameras offering pulmonary perfusion ventilation sections; in cardiac ultrasonography, showing the movement of the walls and valves of the heart; or in standard biplane fluoroscopy revealing joint kinematics or gastrointestinal motility.

 So why this renewed interest just now for CT imaging. physiological? This is because CT CT has evolved to a point where with a high resolution CT scanner we can begin to structure, revealing for example regional lung pathology with sub-millimeter resolution. Conventional multilayer helical CT already offers the ability to scan a significant axial volume of the chest or abdomen in a single apnea.

. Magnetic resonance imaging (MRI) has now reached imaging speeds where cutting acquisition times as small as 67 msec. This, together with the emergence of electron beam CT (EBCT) and polyconic multi-layer CT 4-D CT scans of less than 50 msec, allows the capture of beating heart motion, diaphragm displacement , the . passage of blood through the tissues or perfusion, and the distribution of a marked insufflation with a gas radiodense. The effects of the air-water interfaces on the NMR signal are being used to estimate the dimensions of the pulmonary alveoli while it seems to be at the # a small signal of nothing at all from the pulmonary field in an image. NMR of the thorax, and the use of hyperpolarized helium or xenon gas is permitting via NMR imaging of the structure

<Desc / Clms Page number 21>

 pulmonary. The experimental resolution of NMR and CT CT was constructed in such a way that it allows micron resolution, providing visualization of the vessels at the true interface of oxygen delivered to tissues. Positron emission tomography (PET), coupled with the anatomical detail derived from X-ray CT and MRI, is producing images of the structure / function relationships involved. The practice of Radiology has always been highly dependent on the collaboration between physicists and electrical engineers to achieve an integrated physiology. In the following descriptions, we review current imaging and image analysis approaches that allow the physician and scientist to begin to quantify and visualize, in a network or on a single device, the structure at the same time. that function.

2.3.1. Physiological Imaging in a Computer Network It is possible to have a system of networked devices and a centralized or non-centralized PACS architecture including in the first alternative a storage unit with a short delay or else a system of data storage. decentralized archives with multiple archiving. The successful implementation of network-based image-guided therapy on a large scale requires specialized infrastructure, especially a high-performance network, in addition to mapping the PACS of a hospital complex or even several hospitals. In a search environment, this network must have specialized storage areas across multiple sites that provide sharp and immediate access necessary for keyframes. To meet any new emerging requirements, such as when new systems and accessories are integrated into the existing infrastructure, the flexibility of the network and its components is also required. It is therefore not possible to deliver images to a complex and intensive therapeutic environment without the collaboration and involvement of a Department of Radiology. There is therefore an inevitable increase in the workload for the staff and therefore the cost of the acts. The goal, however, is to create, describe, manipulate, and apply, at minimal additional cost, images to a therapeutic environment with high availability, reliability, and accuracy. It is and is the key to success - a multidisciplinary team that includes surgeons, radiologists, radiation oncologists, physicists, and appropriate image-providing engineers along with a knowledgeable technical support staff, which is constantly mobilized around this system.

It is also known that a central architecture of such a nature is vulnerable and

<Desc / Clms Page number 22>

whereas, on the other hand, a decentralized structure is redundant and, depending on the source, debits data at very different speeds. And, who else especially that the
Radio-oncologist also needs a technical file of irradiation of a patient? Such an organization would therefore be oversized for routine use in radiation oncology. In addition, the patient is unique and can not be, if not successively in time, spread over several machines at the same time. The time adding to the time and representing an hourly labor cost and a longer care, when the patient must successively pass from one to the other with multiplication of gestures, measures and their conditions . execution; multiplication of the number of interventions and stakeholders; finally multiplication of machine-operator interfaces (number of convolution necessary increases) and opportunities for error. The network does not synchronize the set but superimposes the various gestures and adds their errors and inaccuracies. All this is, with regard to Radiotherapy, harmful and. very complicated: Finding a way to simplify the overall handling and make it available to the greatest number is the primary objective of this new invention.

 Nevertheless, where a computer network dedicated to image-guided surgery connects, as at the University of Iowa (John W. Haller et al., 2001), all imaging devices ( CT and NMR) to treatment planning devices, radiosurgery consoles, and surgical consoles, can of course provide a quick and efficient transfer of images from an imaging device to the surgical navigation console. operation. The network server also provides images to the PCs on the control panel. where pre-surgical planning can be done. This is not without problems in radiotherapeutic practice.

Until now, it was not possible to avoid irradiating otherwise healthy tissue during radiotherapy of cancer, because there were only a limited number of potential angles of fire, the irradiation field being defined in a fixed manner. With evidence accumulating in the literature that Radiation Therapy may be associated with a range of late sequelae, some of which may occur as late as 20 years later or more after initial therapy, it may seem interesting that the cancer patient with long survival is not, as SM Bentzen and S. Dische (2001) believe, in the position of the King
Dionysus 1st envious of Damocles rejoicing in greatness and joy more

<Desc / Clms Page number 23>

 brief of a luxurious banquet but sitting at the same time under a sword hanging by a single horse hair over his head.

2. 3.2. - Imaging Physiology Integrated to Radiotherapy
On the other hand the control of more sophisticated treatments, such as intensity modulation, it is obvious that the only way to control the dosimetric accuracy is to measure on a phantom with a precise detector the intensity modulation. the dose that would correspond to the desired distribution. If such measures are part of the criteria of acceptance of a treatment and that one will have quickly to pass, from a daily point of view, to the employment in practice of the. measured dose distribution, during treatment, by gantry imaging. There is no doubt that the permanent collection of data from the acquisition system, as proposed here by the electron beam scanner (Imatron's C-
150 electron beam CT scanner. Electron beam CT scanning.) Associated with X-ray cameras measuring, in a set called Bi-linac, the dosimetric characteristics. beams; those on the inhomogeneities of the volume crossed by the radiations, readable on the scanographic sections which would be, in the matter, a must. And this is exactly what is constantly realized, as we will see in the description below, the device Bilinac-Imatron coupled to a server type console dosimetry online and real time patent n 01/01133. As it is possible. In conventional external beam radiotherapy, even in the most widely used cross-fire technique of the present state of the art, the effects of particle braking in the intratissular medium are used.

 In fact, in one of the embodiments of the Bilinac-Imatron device, such as the SLC collider, it is conceivable that for irradiation with electron beams. that the collision occurs between the electron beam on the one hand and the positron beam on the other and that for the photon irradiation the collision occurs between the two beams of the same energy or different energies . It is the SLAC (Stanford Linear Accelerator Center) transformed into a medical linear collider or better the J.L.C. Japanese Linear Collider: JLC-1, KEK Report 92-16, 1992.). The dosimetry is here focused on quantifying, in the target volume, the electromagnetic interactions of the two beams of simultaneous irradiation. That is why we propose to use in the new invention software that have already proven themselves on all the dosimetry and virtual simulation consoles available in the art and have been on the market for at least 20 years. years.

<Desc / Clms Page number 24>

Increasingly quality assurance procedures are being developed for modulated-intensity X-ray beam (IMRT) conformal radiotherapy procedures using dynamic multi-leaf collimators (MLCs). The procedure verifies an X-ray modulated beam type prescribed with the beam eye's view (BEV) of the beam, before the treatment procedure is applied to a patient. It verifies that a) the computer files sequencing the slides are correctly transferred to the Linear Accelerator Control Computer; b) the processing can be performed correctly without any machine error. The Beam Imaging System (BIS) is a fast beam imaging system consisting of a Gd202S scintillation screen, a charge-coupled device (CCD) camera, and a personal laptop (portable PC,
Wellhofer Dosimetry, Schwarzenbruck, Germany) was proposed for this purpose. Reference images are derived from the slide sequencing files that are used to drive a dynamic MLC system, such as Varian Oncology.
Systems (Palo Alto, CA). A correlation method has been developed by Ma et al. to compare the BIS measurements with the calculated reference images. A correlation coefficient is calculated, using 26 correct intensity-modulated fields, which has been shown to be a reliable discriminator for identifying inaccurate treatment delivery files. This study by L. Ma et al. (1997) demonstrated the feasibility of using the BIS and the correlation method to perform BEV quality assurance tasks online for RIM or IMRT treatment fields.

 However, the desire to improve the local tumor control and healing of many more cancer patients, coupled with advances in computer technology and linear accelerator models, has stimulated the development of radiotherapy techniques. three-dimensional conformational. Optimized treatment plans, aimed at delivering a high dose to the target, while minimizing the dose to the surrounding tissues, can be delivered by multiple fields, each with modulated beam intensity (IMRT), space point of view, or with multiple slice treatments or tomotherapy. To deliver optimized treatment plans, the latest method is Intensity Modulated Arc Therapy (IMAT) to further improve the therapeutic ratio. It uses a continuous movement of the stand, such as in conventional arc therapy. Unlike conventional arc therapy, the shape of the field, which is shaped by the colli-

<Desc / Clms Page number 25>

 Multi-blade controller, changes during rotation of the stand. Arbitrary intensity distributions 2 D of the beam, at different angles of the beam, are thus delivered with multiple arcs of the beam superimposed. This is why a system capable of delivering IMAT has been implemented. An example illustrating the feasibility of this new tomotherapy technique is given and discussed in relation to other delivery schemes, based on the slice (C.X. Yu, 1995).

Except for tomotherapy, the Bi-linac concept is also capable of performing both IMRT and IMAT, or even a combination of both, with considerable time savings and no increase for workload staff. .

Modulated intensity radiation therapy (IMRT) is a new form of conformational processing, made possible by advances in multi-blade collimator (MLC) technology, brush-beam modeling and inverse optimization algorithms. With IMRT, the benefits derived by the patient are the improved backup of the dose to normal tissues and the possibility of escalation of the dose. With the IMRT, the risk to the patient would come, it is estimated, from the error on the dose. To do this, several potential sources of dose error were, on the one hand, enumerated: (a) inaccuracies in the controls of an accelerator and / or MLC; (b) those on the patient's facility and / or organ movements; and (c) the inaccuracy of a dose calculation algorithm. And it is, on the other hand, the human error never evaluated in daily practice.

There are, however, only three levels of verification of the accuracy of a dose calculation by an IMRT treatment planning system, namely: (a) at the level of the manufacturer of a commercial planning system treatment ; (b) in the clinic, during a proposal phase of an IMRT system; and (c) in studies of the specific phantom of the patient. For all this to be useful, real-time implementation must be achieved. This is precisely what is proposed in a specific study of the patient's phantom, the above description where an IMRT plan is first generated with a single CT scan of the patient, containing the target volume. The optimized fluences of the patient are then applied to a CT scan of a water equivalent phantom. The essential step to recalculate the doses involves using, with respect to the geometry of the phantom, the IMRT system. Dosimetric verification is for the phantom between measured and calculated doses. A specific study of the patient's ghost is intense work. For the issuance of IMRT by a sliding window, a

<Desc / Clms Page number 26>

 difference in the installation of the monitor unit between the phantom planes and the patient can complicate the validation of the study of the ghost. A bonus feature of a ghost study is that the study checks whether the controls of an accelerator and / or MLC are behaving, at least for the day the study was conducted, adequately.

2. 3.3. - Method and targeted irradiation strategy by modulated intensity The actual therapeutic irradiation is always preceded by a careful preparation. It is necessary to locate first, with precision using CT and MRI, the tumor. The doctor can then define the volume of healthy tissue to preserve. He also defines for each volume the dose of rays to be delivered. Then he simulates, using the planning computer-RIM, the planned irradiation. The computer divides the area to be irradiated into different fields, which can themselves be subdivided up to 44,000 segments. The computer then proposes for each field a large choice of different irradiations, taking into account the effects on the neighboring fields as well as on the different segments, and thus making it possible to develop a tailored irradiation plan. When the radiation passes through the tumor, the irradiation intensity of this segment is increased. If, on the other hand, the irradiation passes into healthy tissues, the intensity is reduced. Thus, a large number of small fields of isodirectional irradiation, each having a different intensity, are created.

 When the computer has finally calculated the best irradiation strategy, the actual irradiation by means of the linear electron accelerator can then begin. The beam of rays produced by this linear accelerator is modeled as soon as it leaves the apparatus in three dimensions by a lamellar collimator, formed of 120 lamellae of lead or tungsten, each of which can be actuated individually, and which turn when emission in the irradiation fields. It is thus possible to save certain tissue areas with great precision and to vary the radiation intensity in the irradiation field - in the three dimensions of space: length, width and depth. It is to the precision of the irradiation and especially to the increase of this precision at the same time as to a greater differentiation between the tumor dose and the dose to the healthy tissues, which would open by this way the way to a better knowledge of the tolerance of healthy tissues, which is still working as we have already said this new invention. To achieve this accuracy, certain elements are essential: accessories probably adapted to a good immobilization of the patient; exact three-dimensional acquisition

<Desc / Clms Page number 27>

 not only external contours, tumoral volume and surrounding healthy tissues, but also their variations over time; a computer system and software to reconstruct three-dimensional and time-based anatomical data, to optimize the irradiation technique and to cal- lot the 3-D distribution of doses. The RIM certainly allows to obtain the same cancer cure rate with the usual radiotherapy, but with much fewer side effects, since the healthy tissue seems to be better preserved. This method increases the dose of radiation and gives, in some types of cancer, a better cure rate. RIM is particularly useful for tumors located near healthy organs or having a three-dimensional branched structure. It is possible, inasmuch as this new technique is currently used especially for cancers of the ENT sphere, to extend within the scope of the present invention its field of application to the treatment of prostate cancers, gynecological tumors and cancers. of the child, where it is particularly important to preserve healthy tissue. But, we are told, the lack of qualified personnel and available devices may limit its use. At best one to two people a month could benefit in the current state of the art of the RIM. This limits access drastically. Hence the necessity in the context of the present invention of constantly materializing the target itself, and not at a given moment, as in the ancient art, its drop shadow, and determining its isocenter. This is what we mean, as opposed to the so-called virtual simulation done in the absence of the patient, by clinical simulation (C. Mwanza-Chabunda, PhD Thesis, ULP, 2000), where there is a symmetry of transformation, in the extent to which symmetry today means a similarity of the opposite parts, the exact reproduction on the left of what is on the right (Violet-le-Duc) 2. 3.4. - Simulation with physiological imaging In this area also the aspiration to do well is not rewarded, for lack of a suitable technology for planning Radiotherapy. The distinctive theme of this new invention will therefore be the notion of the permanence of the radiological "demonstration" in situ of a target with its different dynamic surface references, where the ancient art has always resorted to the fixed anthropomorphic ghost and to historical representations. The conceptual family, to which the word "demonstrate" refers, comprehensively encompasses not only the hard variants, whose fields of application are objective; but also the concepts with variable contours, among others, but statistically well defined that, under the technique of

<Desc / Clms Page number 28>

 virtual simulation, ultimately introduce the terms of forecast target volume, PTV, macroscopic target volume, GTV, and anatomo-clinical target volume, CTV, etc. All these terminologies, however, aim to establish the properties of the object of the simulation, the target conformational volume.

One quickly realizes that the clinical act can not always be objective, in all these last cases where the inanimate phantom is objectified, and the precision of the scanned image well read can not be all the time of rigor; the scientific excellence of the method can not find its account in this repository with variable geometry. Three-dimensional treatment planning and CT simulation are nowadays widely used for the treatment of a variety of cancers. As for example at Stanford University Medical Center, where a CT treatment planning review is obtained prior to therapeutic breast irradiation to optimize the distribution of the dose to the treated breast and to limit the irradiation of the opposite breast, c #ur, and lungs. In an article whose results below, Mehta and Goffinet (2001) review the fortuitous discoveries they made, under posterior reading conditions, on a careful review of slices used for planning purposes from a single reading of a single CT scan.

Between 1997 and 1999, 153 patients were referred for radiotherapy of the breast or chest wall and therefore underwent a CT scan. The planning cuts were immediately expanded to include not only the breast, but also the neck, thorax, and liver. A resident and a Radiation Oncologist had carefully reviewed each section before approving the treatment plan. No abnormal findings were made then, in the Department of Radiology, by a caregiver or additional diagnostic imaging or another assessment either, when it was necessary, was requested. It was these 153 sequential sections that were reviewed, and 17 non-suspected abnormalities were observed (11%). The abnormalities included the lung (n = 4), the liver (n = 3), the gallbladder (n = 4), the esophagus (n = 2), the lymph nodes (n = 3), and the breast. All of these abnormalities that stigmatize human error were evaluated with additional imaging studies and / or appropriate consultations. Four of these abnormalities represented additional cancer foci (3%) and modified the treatment plan accordingly.

The results immediately argued for the fact that three-dimensional CT scan of breast cancer

<Desc / Clms Page number 29>

 should be carefully reviewed and if necessary redone. If in this institution, 11% of these planning studies contained abnormalities, and 3% demonstrated unanticipated sites of breast cancer involvement, it is in repeated extemporaneous CT scanning at each fraction of treatment that resides at this time. the true nerve of evidence of bias in relation to the readability of the target lesion and its spontaneous evolution over time or under the influence of treatment. The lesion should not be, unpredictably, a variable at random coordinates, such that the more this variable is dispersed the greater the variance will be. And it is this idea of the permanent proof and the topographic variability of the lesion that we will have to clarify and concretize with this new invention and establish a symmetry between the subject and the object aimed at, while stripping from the same blow of any captious artifice connotation the assertion of virtual simulation.

2. 4. - 3D treatment planning and monitoring
2. 4.1. Delayed CT Scanner and Extemporaneous Electronic Control Imaging Radiotherapy planning (RTP) has historically become with the advent in the recent years of Conformational Radiotherapy an intensive image discipline and one of the first areas in which to From an imagery the 3-D information is, even though remarkably dependent on the exact 3-D definition of the target and non-target structures, more and more clinically exploited. In addition to the interactive display of the wire frame or shaped models of the anatomical surface of objects, radiation beams, beam modifying devices and calculated dose distributions have been proposed, the significant use of the relevant direct visualization of anatomy made recently from image data, to geometrically optimize beam placement, dedicated systems, implementing the functionality of the standards of radiotherapy simulators in virtual reality are commercially available. CT simulation systems that rely heavily on 3-D visualization and projection of image data to ensure that simultaneous x-ray comparisons, performed with diagnostic-quality X-ray images on a simulator, are performed with morphometric images of megavoltage, using the therapeutic bundles themselves. Although the calculation and analysis of dose distributions is an important component of

<Desc / Clms Page number 30>

 In the design of radiotherapy, geometric targeting with optimization based on anatomical 3-D information is frequently performed, according to Pelizzari and Chen (2000), as a completely separate step independent of dose calculations.

On the other hand, the combination of CT for computed tomography (CT) and treatment simulator, such as in Sim-CT, Elektra Oncology Systems, has been evaluated in a clinical situation in Belgium. Image quality, hounsfield units (HU) and linearity were evaluated for both implantations and treatment planning. The additional dose to the patient was also brought to light. The image data is acquired using a row of solid X-ray detectors attached to the outside to intensify the image of the simulator. Three different fields of view (FOV: 25.0 cm, 35.0 cm and 50.0 cm) with 0.2 cm; 0.5 cm and 1.0 cm of cutting thickness can be selected and the system has a standard opening diameter of 92.0 cm at standard isocentric height. The performance of the CT was characterized by several criteria: spatial resolution, contrast sensitivity, geometric accuracy, reliability of hounsfield units and the level of radiation production. The NAQP (Nuclear Associates Quality Phatom) quality phantom spatial resolution gauge as well as the Modulation Transfer Functions (MTFs) were applied to evaluate the spatial resolution. The contrast sensitivity and UH measurements were performed using NAQP and a conversion phantom that allows insertions with different electron densities. CT-option computed tomography (CTDI) dose index was monitored with a pencil-shaped ionization chamber. Treatment planning and dose calculations for heterogeneity correction based on Sim-CT images, generated from an anthropomorphic phantom and from patients, were compared to similar treatment plans, based on contrast identical images, diagnostic CT (DCT).

The last row of holes, which are distinguished by the NAQP spatial resolution gauge, is either 0.150 cm wide or 0.175 cm wide, depending on the FOV and applied reconstruction filter. The latter consist of FTM curves showing cutoff frequencies ranging from 5.3 Ip / cm to 7.1 Ip / cm.

Analysis of the linear regression of UH versus electron densities revealed a correlation coefficient of 0.99. Contrast, pixel size and geometric accuracy are within the specifications. The values of the index of

<Desc / Clms Page number 31>

 Computer-assisted computed tomography of 0.204 Gy / As and 0.069 Gy / As were observed with dose measurements at the center of a 16 cm and 32 cm diameter phantom of the large FOV. The CTDI values of the small FOV fields of 0.925 Gy / As and 0.358 Gy / As are, under similar conditions of acquisition, a factor ten times higher than the results obtained from a DCT. These ghost studies showed excellent agreement between the distributions generated with the Sim-CT and the UH of the DCT. The deviations between the calculated monitor unit settings and the maximum dose in all three dimensions were for treatment plans based on these two HUs of less than 1%, whether for pelvic or thoracic simulations. The patient studies subsequently confirmed these results.

If the CT-option can be considered on the one hand as an added value to the simulation process and the images acquired on the Sim-CT system considered adequate for calculating with correction of the heterogeneities of the tissues of the dose, the good quality of image is compromised, however, by the high values of the dose to the patient. On the other hand, the considerable load of the conventional X-ray tube frequently limits the Sim-CT to only seven image acquisitions per patient, and the system is therefore limited in its ability to perform the entire three-dimensional reconstruction as a whole. This is not the case of the 4-D radioscanners involved in the present invention.

Other teams have used gantry CT, in which the radiation treatment bundles contain valuable information for verifying the patient's fit. These bundles can be used for gantry CT reconstruction. However, the direct use for reconstruction of the beam data can produce inadequate CT images, simply because these beams only cover a part of the patient's body thus necessitating the use of artifacts for the synthesis of missing data. This is why H. Guan et al. used, to reconstruct a locally enhanced gantry CT image, said processing beams in addition to a series of regular CT projection beams.

 They named this approach by adaptive gantry CT reconstruction. The reconstruction of the image was thus performed by the technique of a schema at multiple levels of algebraic reconstruction. The results indicated, according to H.

Guan et al. (2000) that the quality of adaptive portal image reconstruction may be valid not only for verification of three-dimensional radiotherapy, but also applicable to diagnostic CT imaging.

<Desc / Clms Page number 32>

[I is a fact that when Radiation Therapy (RT) is used to treat cancer patients, its therapeutic index can be significantly increased if the volume of the dose is tailored to match the target volume while sparing normal tissue. adjacent. This is why Moderate Conformal Radiotherapy seeks to achieve this goal by using nodulated intensity X-ray beams to deliver high doses of radiation to the target while protecting the normal outcome. However, this method uses rigorous specifications, including stability with respect to treatment beams, of the target volume.

Controlling a conformational treatment with a combination of computer assisted megavoltage tomography and gantry imaging has certainly increased in frequency (S. Webb, 1997) but not necessarily in efficiency. In recent years, indeed, many of the feasibility studies for using CT reconstruction for megavoltage have been reported (A. Brahme et al, 1987, DG Lewis et al, 1992, H. Guan and Y. Zhu, 1998, Mosleh -Shirazi et al, 1998, Hesse et al, 1998, S. Midgley et al, 1998, KJ Kuchala et al, 1999). These studies have, for example, used conical beams (MA Mosleh-Shirazi et al., 1998, S. Midgley et al., 1998) or simply fan-shaped beams (A. Brahme et al., 1987, DG Lewis et al., 1992; Guan and Y. Zhu 1998, BM Hesse et al., 1998, S. Midgley et al., [998) and the continuous data acquisition (KJ Kuchala et al., 1999) or a sequential approach, step-and -shoot (A. Brahme et al, 1987, DG Lewis et al, 1992, H. Guan and Y. Zhu, 1998, MA Mosleh-Shirazi et al, 1998;

 1: esse et al, 1998; S. Midgley et al., 1998). They also used detectors comprising a GO scintillator (A. Brahme et al., 1987, DG Lewis et al., 1992) a CsI crystal (TI) array, a xenon chamber (KJ Kuchala et al, 1999), and portal images (H. Guan and Y. Zhu 1998, BM Hesse et al 1998, S. vlidgley et al 1998). The dosage involved in these studies, however, restricted their extension to clinical application. The direct use of radiation treatment beams in CT portal reconstruction would be desirable, since these beams can not give an additional dose to the patient in practice. This new approach, however, is not without problem, since the treatment beams only cover a portion of the patient's transverse section.

Not only the truncated data can not give an exact map read attenuation coefficient, but still the reconstructed local region does not have such surrounding surrounding essential structures. Such a reconstruction may have limited use in verifying the patient's facility and its images are far

<Desc / Clms Page number 33>

 to have a diagnostic quality. Even if this projection is not truncated, the number of beams may indeed be too small (usually less than 10) to iteratively provide CT images of useful quality, unless the arc or cyclotherapy is used. To tackle this problem, Huaiqun Guan et al. propose to acquire, before radiotherapy, a small number of the entire projections (covering for example the cutaneous contour of the patient) with a low dose of radiation to serve for the synthesis of the missing image data. The acquired projections, from the treatment beams, could then be added to the full projections to locally and adaptively enhance the quality of the image around and within the tumor. This adaptive gantry CT technique is an image composition that can also apparently be applied to the diagnostic CT (local tomography or tomography of the region of interest), with a view to reducing the dose. However, this study focuses on the verification of radiotherapy from this composition image. Previous studies of local tomography (Ogawa et al., 1984, PV Sankar et al, 1982, RM Lewitt, 1979, JC Gore and S. Leeman, 1980, AI Katsevitch, 1966, WJT Spyra et al, 1982) indicated that truncated projection data distorted the quantitative accuracy of image reconstruction (JC Gore and S. Leeman, 1980). A general approach to improve the accuracy of the computer-generated image was to extrapolate the truncated projections using either smooth continuity of the data or some prior information related to the shape of the object (Ogawa, K. et al. 1984, RM Lewitt, 1979). However, the discontinuities of the projections derivatives can still produce characteristic artifacts. Therefore, these authors consider that they could not apply this approach to clinical situations, but that they could take, however, in addition to truncated screenings a limited number of complete projections to improve the reconstruction of the patient's anatomy. . It is this reconstruction technique they have called the adaptive portal CT technique for radiotherapy.

If not in CT, the technique of algebraic reconstruction (ART = Algebraic technical reconstruction) seems to provide today, according to the literature (R. Gordon, 1974, GT Herman, 1980) a better reconstruction either from very few projections or to from the incomplete projections that the conventional convolution method of retroprojections (CBP, Convolution backprojec-tion). The disadvantage of ART is its low computation speed (which also affects the accuracy of convergence), caused by the strong correlation between consecutive projections

<Desc / Clms Page number 34>

 A correlation can be minimized when two orthogonal projections are adjusted consecutively. To this end, a new multi-level algebraic reconstruction (MLS-ART) scheme was developed to substantially improve the computational efficiency as well as the accuracy of the reconstruction of the ART algorithm (H. Guan and R. Gordon, 1994). When VILS is used, the maximum orthogonality of projections may be maintained during ART during ART. The layout of the MLS, although it works best when the number is a power of two (while the MLS is the exact order for the fast one-way Fourier transform), applies to any number of projections . Recently, the MLS-ART algorithm could be done in an equivalent way for adaptive gantry CT with a limited number of complete projections plus truncated projections, since it has for example an additional advantage, flexibility. A simple update of these pixels throughout the ray paths, which are available from truncated projections and which leave out other image pixels. This is not a trivial task for the conventional CBP technique. In this preliminary study reported here, the idea I adaptive CT reconstruction was tested using computer simulation in two-dimensional cases. As Conformational Radiotherapy uses, as a component of the integral or optimal feedback of the linear accelerator, the commercially available Multi-Blade Collimator (MLC), which allows automatic protection against irregular fields by movement. computer assisted multiple tungsten slides (R. Corvo et al.).

2.4.1.1. Delayed use of CT-scanner data in Radiotherapy
In this area, the development and performance of a cone beam computed tomography (CTCT) imager (CBCT), using a flat panel with indirect detection (FPI = flat-panel imager), was recently presented by DA Jaffray and JH Siewerdsen (DA Jaffray and JH

Siewerdsen: Cone-beam computed tomography with a flat-panel imager: Initial performance characterization. American Association of Physicists in Median 2000; S0094-2405 (00) 01306-7. In Bibliographic Review of /. Radiol 81; 2000: 1774.).

The Multi-detector row Computed Tomography (MDCT) Multi-Detector Scanner with 3-dimensional volume (3D) rendering offers a unique perspective on anatomy and pathology. The volume rendering now allows the real time, the interactive modification of the relative attenuation of the pixel, in an infinite number of

<Desc / Clms Page number 35>

 plans and projections. Developed still in the prototype state for a preliminary investigation of the conical projection system, CBCT, which we know moreover that this precise type of images is, because of the collimation of the beam, multiparametric, and in view of the localization , based on the FPI of bones and soft tissues, in Radiotherapy. It is from the projections acquired, during a single rotation, that the system provides three-dimensional volumetric image data. This system employs a 512 x 512 active matrix of thin a-Si: H transistors films and photodiodes, in combination with luminescent phosphor.

 The performance of a tomographic imaging is quantified, in comparison with each one with the performance measured on a conventional CT scanner (which is in itself today a bad reference), in terms of uniformity of response, linearity response, voxel noise, noise-energy spectrum (NPS), and modulation transfer function (MTF). The answer is judged, for the geometry used and the objects considered, uniform to 2% and linear to within 1%. The noise of the voxel is, at the level of 20 HU, comparable to the conventional CT scanner. NPS and MTF results from Jaffray et al. highlight the characteristics dependent on the transfer frequency, thus confirming that the CBCT system can provide, under these conditions, a high spatial resolution and can not suffer greatly from the levels of additional noise. For larger objects and / or low exposures, additional noise levels must be reduced to maintain high performance. Imaging studies of a low-contrast phantom and a small animal (a rat euthanasia) qualitatively demonstrate excellent soft tissue visibility and high spatial resolution (this latter is a function, as we know, of sampling of the object at the time of data acquisition and the discretization of said object). The quality of the image seems to be, for these authors, comparable or superior to that of a conventional scanner. These quantitative and qualitative results clearly demonstrate the potential of CBCT scanning-beam systems, flat panel imagers. Advances in FPI technology (enhanced X-ray converters and enhanced electronics, for example) are anticipated to enable high performance in CBCT for FPI-based medical imaging.

However, to meet the more specific needs of Radiotherapy, the search for an optimal irradiation technique is not an idea

<Desc / Clms Page number 36>

 new and since the beginning of this discipline has been the primary objective of the radiotherapist, the physicist and the technician in charge of developing the treatment plan. The current performance of imaging and computer systems associated with the development of specific software, simply allow to design and approach in a different and especially fast way the act of simulation. Doing better, cheaper and faster, this is where the epistemological problem lies, concerning among other things the determination of the target volumes (GTV, CTV, PTV) addressed to the resolution of which is attached by the present invention; clinical or virtual simulations, 2-D or 3-D; the accuracy of the ballistic aim especially; and finally the distribution of the dose, its heterogeneities as well as its dose gradient between healthy tissue and tumor, in view of the uncontrolled heterogeneities, as well as the mathematical tool that has to spend a lot of time calculating everything, or even to integrate in this calculation all the uncertainties and various parameters, accumulating inexorably in considerable sums of calculations already very tedious with their respective additional errors on all the respective valuations deferred and executed separately on different tools with different instruments.

In the prior art, the system is made in such a way as to define the target in a cloud of large and imprecise density density that is able to contain the lesion, like the Heisenberg uncertainty principle, in quantum mechanics. relative to the observation time and the exact position at a given moment of the object.

Only a control of scientific excellence, like that proposed by this new invention, can allow an objective verification of the aim, an approach outside of which there is no science to speak of. This disturbing element breaks the symmetry of transformation allowing to seek a balance of another nature. This to the extent that more uncertainty is equal to more unnecessary irradiation to normal tissues, and especially more tumor tissue missed by imperfect targeting. Concession made to the philosophical foundation see through which presides over the destinies of Radiodiagnosis, whose refusal to undergo these uncertainties has as a corollary the decrease of the level of useless irradiation of the healthy tissues. It is therefore necessary at all times to be able, in the radiological sciences, to see through and control exactly scientific excellence and the quality of what we do.

Current art is in this a fiction where the coincidence of the elements of the target volume having existed or existing is, insofar as things evolve constantly, quite fortuitous and escapes any control or even any precise prediction. In

<Desc / Clms Page number 37>

 Heisenberg's uncertainty, indeed, the probability of finding the electron where the density density cloud is certainly stronger but not necessarily accurate. However, more extensive will be this more precise volume cloud will be the equation used. In terms of target volume, less thick will be this more conformational cloud will be the irradiation strategy, but the probability of missing the lesion or a large part of this lesion increases all the more paradoxically the ballistic accuracy move a little further. Knowing that this type of uncertainty applied to the Target Volume, because of the poor targeting, and to the heterogeneity factor of the dose as well as the geometrical inaccuracy in the repeated positioning of the patient can vary, during external radiotherapy, in a proportion today, as long as there will not be a reference invariant to the said non-quantifiable transformation.

2.4.1.2. Fusion of CT-NMR and DDR data for delayed radio therapy
The use of MRI information in the process planning process is still expensive and time consuming. Electro-optical devices or post-processing software have been specially designed by W. Wagner et al. (2000) against the risk of distortion of the image. Thus, between June 1998 and June 1999, these authors would have carried out CT and MRI treatment planning with a patient positioning considered identical in 48 patients with brain tumors and in 11 patients with prostate carcinoma. using the same devices as during the simulation. The organ transposition and the tumor volumes between the MRI and the simulation snapshot were set up, superimposing them together on the brightness of the screen by means of a grid.

 MRI and simulation films were produced, using the same magnification factor. For the nine patients with brain tumors and two patients with prostate carcinoma in the Wagner study, the transposition of MRI information on the simulation images showed the need for a change in protection or treatment entrance doors.

The simple method of direct transposition of MRI structures into the simulation snaps is, according to these authors, the result of the process itself of treatment planning. An individualized protection can also be directly realized.

To demonstrate the clinical utility of digitally reconstructed radiographs (DRRs) based on magnetic resonance imaging (NMR) in the

<Desc / Clms Page number 38>

 Establishment and verification of patients with intracranial lesions, Ramsey et al. (1999) studied, for treatment planning and (in view) virtual simulation, the NMR images of 16 patients with various intracranial lesions were obtained. Adjacent contiguous cross-sections of the Tl-weighted 5 mm thickness were obtained, using a standard head antenna in a 1.5T General Electric Signa NMR scanner. The NMR-DRRs were generated on a console. Existing treatment planning computer using, without any special modification, the pseudo-density technique. Verification snapshots in anterior and lateral views were taken by Ramsey et al. (2000) for each patient, for visual comparison with MRI - based DRRs. They made sure, thanks to bone markers including orbits, frontal sinuses, sphenoidal sinuses, auditory meatus, nasal bone, vomer bone, mastoid processes, and skull, the visual alignment used by doctors, physicists, and therapists, to check the patient's positioning. They also visually identified and corrected the misalignment from 3 to 10 mm, using, during the simulation, the clinically developed and clinically developed MRI-based DRR technique and method for use visually. The quality of the NMR-DRRs generated using this technique is such that physicians, physicists, and therapists can easily and routinely compare them side by side with the NMR-DRRs in delayed simulation. But it remains to be verified in the course of treatment, without guarantee of scientific excellence, the accuracy of all these forecasts and the precision of their clinical application.

2. 4.2. Verification in progress of forecast accuracy
It is based on electronic portal imagery (EPID). The U. study

Mock et al. (1999), for example, estimated the accuracy of setting up, for the purpose of determining the planning target volume (PTV), during pelvic irradiation for gynecological malignancies. 25 patients undergoing pelvic irradiation by 4 fields for gynecological malignancies were analyzed, considering the accuracy of installation throughout the spreading of the treatment. Regularly made EPID images have been used to systematically evaluate the systematic and random component of the implementation. The anatomical harmonization of the verification images with those of the simulation was followed by the measurement of the corresponding distances between the central axis and the anatomical characteristics. analysis

<Desc / Clms Page number 39>

<Desc / Clms Page number 40>

 circular cross-section beam is measured with respect to a steel sphere attached to the patient's positioning table and which must coincide exactly with the isocenter. Since measurements can be made at all angles of the stand (if available) as well as at all possible orientations of the patient's table, this system is considered particularly suitable for quick and accurate measurements of the isocentricity of the stand. and / or patient support. As it is measured directly at the beam offset positioner, the system provides an implicit verification of the total processing system. This system was first developed for use with proton beams, but it can also be used equivalently for X-ray beam alignment checks of a linear accelerator or other source. The measuring instrument consists of a scintillating screen viewed by a CCD camera, mounted on the stand downstream of the sphere. The steel sphere is not big enough to stop the protons of all energies of interest; however, it still modifies the energy and direction of the protons that interest it, creating a region of lower intensity (a shadow) in the bright spot created by the beam of protons striking the screen. The position is a measure in relation to the luminous point of the shadow of the system alignment. An image analysis algorithm has been developed for automatic determination of the position, with respect to the luminous point, of the shadow. The specifications and the theoretical analysis of the system were derived from the Monte Carlo simulations, which are validated by the measurements. Barkhof et al. (1999) demonstrated that the device detects misalignment of the beam with an accuracy (1 s d), in agreement with the expected performance of 0.05 mm. This accuracy is considered more than sufficient to detect the maximum misalignment of 0.5 mm allowed.

 2. 4.3. - Verification of the alignment of a therapeutic irradiation beam.

Random errors in placement can lead to a patient's erroneous prediction of the calculated dose distribution, using a computer-assisted static (CT) model; this insofar as things are not frozen in time and space. Multiple recalculations of the dose distribution covering a range of expected patient positions provide a way of estimating a course of treatment. However, because of the statistical nature of the implementation uncertainties, many treatment evolutions must be simulated to calculate a distribution of the average values of the dose delivered to a patient. Thus, direct simulation methods can be consumed

<Desc / Clms Page number 41>

 time matrices and may be impractical for clinical routine applications. Methods have been proposed to efficiently calculate via convolution with a probability distribution function (usually Gaussian), the distribution of the mean dose values that describes the nature of the uncertainty of dose distribution (calculated on a CT model). static). In an article, J. Barkhof et al. extend the convolution-based computation to compute the standard deviation sigma D (x, y, z) of the potential results for the distribution of the mean values of the dose, and they characterize the statistical significance of this quantity using the limit theorem Central.

Inverse three-dimensional planning of modulated beam processing was also applied to the dosimetric optimization of a complex (concave-convex), predicted target volume (PTV), fashioned in the upper cervical and mediastinal regions by Esik et al. (1997). The planning of the 9 coplanar beams regularly spaced by 40 intervals was carried out. The properties of the 15 MV photons of a linear accelerator were simulated.

The optimization of the fluence modulation profiles was for each beam based on a definition of the desired / allowed levels of relative dose in the forecast PTV volume as well as in the organs at risk, as well as a definition of the forces of constraint to achieve these objectives. . Adequate dose delivery to PTV and protection of the spinal cord have been considered completely feasible in view of the risk of radiation-induced pneumonia. For physical reasons, no further reduction in PTV coverage was possible. This is why the deposit of radiant energy on certain critical parts of normal tissues has effectively been reduced by application of a dummy organ at risk. Inverse planning of conformational radiation therapy for large tumors is, of course, an equally effective method. However, the power of the technique is insufficient when the tolerance dose of the surrounding normal tissue is too low and its volume effect is high. Although requiring more interaction with the operator, the introduction of dummy organs at risk may be, for Esik et al. (1997), an aid in reducing radiation energy deposition on normal tissues. The dose at one point of the tissues depends on the one hand on the irradiated volume, which intervenes by its shape, its thickness and its composition, and on the other hand on the radiant energy of the irradiation beam.

These two factors can not be considered independently because the beam is modified by the presence of the medium and the dose at one point depends on the quantity

<Desc / Clms Page number 42>

 as well as the quality of the incident radiation. Just as the physical characteristics of the medium can, however, in the absence of any irradiation, be defined just as the characteristics of the incident beam can be defined in the absence of the irradiated volume. The beam leaving the generator is then considered and has not passed through any material medium (neglecting absorption and diffusion by air). The dose distribution in the irradiated medium is dependent on the radiation energy and geometric conditions, i.e. the shape and dimensions of the beam and the medium. The system consists, for two-dimensional dosimetry, of a scintillating (fluorescent) screen mounted on the output side of the. beam of a ghost and it is observed by a camera-coupled coupled device (CCD). The light distribution observed on the screen is equivalent to the two-dimensional (2-D) dose distribution at the screen position. However, and without questioning the benefits of all these artifices exposed above, it is established that they do not allow the eradication of the bias and do not control the physiological mobility. or not of the body.

2. 5. Virtual simulation and predictive dosimetry
2. 5.1. Simulation of the trans-action of radiant energy
It is clear that the required simulation accuracy for a linear accelerator. the actual art depends on the required accuracy of the quantities of points to be explored at the end. The same is not the case for multi-interaction radiotherapy for simultaneous beams converging on a single target that operates rather by energy diffusion kernels. For the planning of the Monte Carlo processing, the accuracy of the electronic data of the phase space is usually ensured. so that the calculated Monte Carlo dose distributions can be consistent with the measured data in the range of 2% of the maximum dose (A. Kapur et al 1998, CM Ma et al 1999). These criteria appear to be practical and as appropriate as the uncertainty in the 1% distributions of the clinically measured dose is, according to the recommended dosimetry procedures (IAEA 1987, AAPM 1983, 1991), usually around 2%. .

 The uncertainty of the dose in a routine clinical measurement point in a phantom is more likely to be about 3% (Khan 1994). Experience has shown that the agreement between the Monte Carlo simulations and the fine-tuning measurements of the parameters used in the accelerator simulations could be

<Desc / Clms Page number 43>

better than 1% (A. Kapur et al 1998, CM Ma et al 1999). High precision measurements could be predicted for many authors (CM Ma et al
1993b, CM. Ma and Seuntjens 1997), within 0.3% by simulations of
Monte Carlo. Such agreement seems to be excessive for clinical simulations of an accelerator, if dosimetric uncertainty remains at 2-3% in the dose data used for machine control. Faddegon et al (1998) estimated the various electron emission delivery uncertainties and concluded that an overall dose accuracy of 5% or 5 mm in the location of the isodose lines or 5% / 5 mm in short, could be feasible if '. relative doses were calculated at 3% / 3 mm (point-to-point variation of the dose within a given dose distribution) and the dose in water compared to 2% / 2 mm (Chang-Ming Ma and Steve B. Jiang, 1999). This without taking into account possible bias with respect to the lesion not permanently visualized, with reference only to an image delayed in time with respect to the moment of simula- tion. and therefore necessarily inappropriate for the situation just under way.

2. 5.1. Simulation time for energy deposition by trans-action
The major disadvantage of the Monte Carlo method is the computation time required to obtain an acceptable statistical uncertainty in the simulated quantities. CM.

Ma et al. (1997a) reported that to achieve statistical uncertainty in the 1% dose distribution, around the 1 x 104 electron phase space, in an aqueous consistency phantom, 1 cm3 cubic voxels were required for each surface in the 1 cm2 field. The number of particles per unit area should increase for smaller voxels, to achieve the same statistical uncertainty. Rogers et al. (1995a) have also shown that the. ratio of the number of electrons per cm 2 of the labeled phase space on the labeling plane for a field of 10 × 10 cm 2 with respect to the number of electrons incident at the vacuum exit window and the incident electron energy. For an accelerator of the Varian Clinac 2100C type, this ratio was for a 6 MeV electron beam of about 0.0002 and 0.0008 for a 20 MeV electron beam. This means that simulations of 5 x 107 in simulation of an accelerator must be simulated to achieve the 1% statistical uncertainty specified in the dose calculation. More than 108 electron histories will be needed for a 25 x 25 cm2 field of 6 MeV electrons. This may require 300 hours of time
CPU on a SGI R4400 200 MHz console (Rogers et al 1995a) or a few days of CPU time on a 200 MHz Pentium Pro PC (Kapur et al 1998). The

<Desc / Clms Page number 44>

 Total required CPU time to boost five nominal energies of electrons and five applicators would be close to a month. Many iterative process of fine tuning of the incident energy of the electrons and other parameters of the accelerator, the overall CPU time for the control of all the beams can properly speaking be of several months. It is clear that most cancer centers do not have such computational resources or expertise needed for the Monte Carlo process (Chang-Ming Ma and Steve B. Jiang, 1999).

Over the last two decades, significant efforts have been made in Monte Carlo modeling of clinical electron beams from medical accelerators. Many of the major results of this procedure have been addressed through detailed Monte Carlo simulations. Such efforts have been facilitated by the development of the BEAM code. Electron beams of various clinical accelerators have been for different purposes more systematically studied by different research groups. To improve the development of accelerators, accelerator manufacturers have studied beam characteristics, although they have not published any detailed reports in the literature. Accelerator users have also researched beam phase space information to derive dosimetric data needed, to make dosimetric measurement more accurate, and to perform accurate radiation treatment schedule planning calculations. There is, however, still a variety of practical problems associated with the accuracy of the simulation using accelerator Monte-Carlo techniques (Chang-Ming Ma and Steve B. Jiang, 1999).

2. 5.2. Accuracy in a trans-action situation of the radiation treatment simulation
The most basic information required by a Monte Carlo simulation for a treatment head is the specifications of the parts of the accelerator, such as their locations, dimensions and materials. Without knowing precisely the specifications of the components, the results of the simulation can be of limited use. So it seems by far difficult to obtain specific information. from manufacturers on an accelerator (Udale-Smith 1992,
DWO Rogers et al 1995a). Due to the commercial value of the detailed specifications of parts of the accelerator, manufacturers are usually reluctant to provide the information with the details necessary for Monte-Carlo modeling.

 In addition, newly acquired accelerators are often adjusted on site for individual users, for example by selecting scattering sheets or

<Desc / Clms Page number 45>

 adjustment of the waveguide, to harmonize the characteristics of the beam of the machine to be replaced, or existing machines. The situation can be further complicated by refurbishing, upgrading or updating the accelerator. Accelerators of the same model could not understand exactly the same components. For example, the same type of accelerator may have different diffusion sheets, flattening filters, chambers or monitor applicators (Chang-Ming Ma and Steve B. Jiang, 1999).

It is also difficult for users to know in practice precise information, such as the size and location of the light point (impact), the distribution. electron energy and the angular distribution, from the phase space to the exit window of the electron vacuum. In executing the Monte-Carlo simulations of the electron beam, it was convenient to assume ordinarily that at the vacuum window of the accelerator there was a beam of mono- energetic electrons. The initial energy of the brush beam is determined in harmony. the calculated and measured values of certain dosimetric quantities, say,
R50 or Rp (Udale, 1988, Udale-Smith, 1990, 1992, Keall and Hoban, 1994;
DWO Rogers et al, 1995a; A. Kapur et al, 1998; CM. Ma et al, 1999). It was found that the final curve of incident energy of the dose, unless the spectrum is very broad (DWO Rogers et al, 1995a). This approach has proved practical and. adequate for dose calculations in radiotherapy. The introduction of these empirical procedures, however, calls into question the accuracy of the simulation results, such as the actual data characteristics of the beam phase space, even when the final dose distribution agrees with the measured data. . For example, the calculated energy spectrum is not necessarily one. exact representation of the real spectrum. In addition, as noted by DWO Rogers et al. (1995a), there may be other more subtle effects, for which this approach can be defeated. There is also poverty, according to Deasy et al. (1996) experimental data on the fluence of clinical electron beams. Thus, there is a reason to establish yet another benchmark of the simulation of the. treatment head based on fluence and dose measurements, using a search accelerator with a source and well-known details of the treatment head (Faddegon et al, 1998, quoted by Chang-Ming Ma and Steve B. Jiang,
1999).

<Desc / Clms Page number 46>

2. 5.3. - Characterization of the beam and modeling of the single source of radiant energy
The important reason for simulating a treatment head is to obtain the desired electron phase phase space information for Monte Carlo processing planning. The phase space data generated with the code
BEAM contains almost all the information of the desired phase space (Rogers et al 1995a). However, there are some problems associated with the direct application of phase space data in Monte Carlo dose calculations (Ma
Rogers, 1995b, c; Ma et al, 1997a; Ma, 1998; Jiang et al, 1999). Since the beam characteristics are usually different, even for the same type of accelerator, it is necessary to simulate each individual accelerator, in order to plan the processing, to obtain the data from the Monte Carlo phase space.
Carlo. The production of phase space information, simulating the processing head and its quality assurance is not an easy task, even with the co-, of BEAM. It requires both the experience of Monte Carlo simulation and much more time than is required to provide a conventional electronic processing planning system. This places, according to Chang-Ming Ma and Steve B. Jiang (1999), a practical limit in an ordinary cancer clinic on the clinical implementation of Monte Carlo treatment planning.

 Storing simulated data from the Monte Carlo phase space is another problem. It has been discovered that the beam characteristics of the different electron applicators can because of the variation of the collimated electrons differ significantly by the movable and diffused jaws from the applicator attachments (Zhang et al, 1998; Kapur et al. , 1998). This means that, the phase space files must be obtained for each electronic energy by combination of applicators. In order to achieve the 1-2% statistical uncertainty in a phantom consisting of 0.1-lcm3 voxels, 106-107 phase space particles are usually required, in a Monte Carlo simulation, to the size of the electron beam field (Rogers et al, 1995a, Zhanget al, 1998, Kapur et al, 1998, Ma et al, 1999). For a clinical linear accelerator with five nominal energies and five applicator sizes, this means more than 108 phase space particles or a few gigabytes of computer disk space. More disk space is required when the photon phase space data is also stored. This represents a significant burden for the limited computer resources in most clinical centers.

<Desc / Clms Page number 47>

 Obviously a more concise description, with sufficient accuracy of the clinical electron beam is needed to routinely perform a Carlo Watch treatment planning (Chang-Ming Ma and Steve B. Jiang, 1999).

This is why a methodology has been proposed to replace the phase space file with a multiple source model with parameters derived from simulated phase space data (Ma et al 1993a, 1994, 1997a, Ma and Rogers 1995b,
Ma 1998, Jiang et al 1999). Based on the observations that particles from different components of an accelerator have energy, different angular and spatial distributions as the particles from). of the same component have very similar characteristics, it was considered that the particles of different parts of an accelerator come from different sub-sources. Each sub-source represented a critical component of the treatment head and its geometric dimensions were determined by the component in question. Each sub-source had its own spectral and planar fluence distributions derived from simulated phase space data. By sampling the position of the particles on the sub-sources and on the surface of the phantom, the correlation between the position of the particles and the incident angle was naturally retained. A special program, the BEAMDP, was written to help derive the source model parameters from the phase space data, using BEAM (CM Ma and DWO Rogers 1995a, c). This multiple source model was evaluated against the data of the whole phase space for the different accelerators and the set up of the beam arrangements (CM Ma et al, 1997a, C.-M. Ma, 1998, cited by Chang-Ming Ma and Steve B. Jiang, 1999).

 The same reasoning can be applied to several simultaneous sources of therapeutic irradiation of the present invention.

A convenient ordering procedure of Monte-Carlo treatment planning
Carlo of the electron beam was proposed by CM. Ma et al. (1997c). The idea was to derive the parameters of the basic source model of the simulated phase space data and then modify the model parameters, based on the measured beam source data, for clinical implementation. . CM. Ma (1998) demonstrated the feasibility of this approach, using a previously developed multiple source model (CM Ma et al, 1997a). By varying the energy spectrum edges of the bin for each sub-source, the depth-calculated dose distributions using the reconstructed phase space from the multiple source model for nominal beam energy harmonized with

<Desc / Clms Page number 48>

in-depth dose distributions for another nominal energy. By varying the edges of each planar fluence distribution bin for each sub-source, the calculated dose side profiles using the reconstructed phase space from the multiple source models for a harmonized size of the fields with the dose profiles for each another size of field (CM Ma 1998, quoted by Chang-Ming
Ma and Steve B. Jiang, 1999).

 Jiang et al. (1999) further improved the methodology for modeling and proposing the electron beam (C-Ma Ma et al 1997a, c, C-Ma Ma 1998). They improved the representation of the source model as well as the recognition algorithms. struction and explored the minimum number of sub-sources of dosimetric importance to obtain the required beam proposal accuracy. This method was applied to the electron beams from the Varian Clinac 2100C machine with a Type III applicator. A four-source model was found to be adequate for all energy / applicator combinations. The source model. consisting of a point source of electrons, a point source of photons and two extended sources of electrons. The marking of the plane directly above the upper surface of the last scraper was not manipulated by the source model but simulated separately in an additional BEAM simulation or in the calculation of the patient dose. The model was spotted in. comparing calculated dose distributions with model and whole phase space data (Chang-Ming Ma and Steve B. Jiang, 1999). But for the target it is necessary to acquire anatomical data.

 2. 6. - Acquisition of anatomical data. The precise delineation of the tumor volume requires a three-dimensional acquisition of the different volumes of interest, which is obtained thanks to the imaging systems. Computed tomography remains the most appropriate diagnostic test, not only in terms of contrast, to the required density and spatial resolution criteria, but also in terms of mobility in space and time.

. However, nuclear magnetic resonance imaging has hitherto allowed a better appreciation of the tumor volume and can be supplemented by a computerized image fusion in order to simultaneously exploit the data acquired by computed tomography (CT) and computed tomography. magnetic resonance imaging including the excellent detail definition of TDM 4-D. S. Treves et al. (1998) described

<Desc / Clms Page number 49>

 a technique developed in their laboratory allowing the incorporation of all available information, such as SPECT and PET, into the imagery data, etc. ; that V. Mongioj et al. (1998) evaluated the accuracy of CT, MRI and SPECT image fusion using three different commercial software packages.

 These techniques have a real potential for improving the specificity of tumor tissues; but remains of the order of the only 4-D scanner, such as that embedded in the present invention.

As far as immobilization is concerned, the first step is the making of a thermoformed mask, on which external marks are placed, indicating. in particular the median line and the lateral projection of the center of the tumor volume and then the imaging is carried out under these conditions. It is a question of the present invention of the means of making computer assisted preparation from the only TDM - 4D. However, some software packages are able to merge TDM
MRI, although both exams were performed in different positions. From. this is because the acquisition characteristics are the consequence of a compromise between the desired precision for the visualization of the tumor volumes and the need to include risk organs situated at a distance from the macroscopic tumor volume and the anatomo-clinical target volume. Support allowing irradiation techniques by non-coplanar beams, of any orientation of incidence.

2.6.1. - Determination of target volumes (GTV, CTV, PTV)
All these denominations are in fact related to the precision or more specifically, in the ancient art, to the approximation of the instrument (clinical examination, macroscopic examination, 2-D or 3-D imaging, X or NMR, etc.). ), which determines them. They have indeed been designed, according to the determination proposed in operation by any one. each and according to the accessibility of the local technical platform. Given the theoretical evolutions and the inadequacies that this necessary evolution has caused, it seemed to us that we need to update at a certain level, the methodology condition of all conceptual thinking that we have engaged in this invention of the duality operation. - object, the definition of the radiotherapy target. at the same time as the tool. Insofar as it seems to us to proceed from a certain illusion that does not diminish all the statistical guarantees that conformationally surrounds itself, this simulation procedure is not part of a truly revisional plan. Insofar as this logical determination of the operating systems is no longer referred to in present art as only a shadow or a promise of objects. The meta-level where philosophical knowledge unfolds

<Desc / Clms Page number 50>

 of the object is not that, as the very subject of the conformation suggests, but rather the form as a system of probability of existence of the organization of the real. It is therefore necessary to examine the nature of the consistency and coherence of the fabric of the technical concepts which it elaborates, and to show how the aspects of isolated facts discover, when we are careful, their fundamental determinations. From there, to logically chained conclusions, which account for the symmetrical internal connection between the different transformation phenomena, we must put the concept of the target lesion on foot, which is exactly what, by definition, the -. tructions 3-D CT scanner. This is because the topography of the lesion does not remain a random variable that varies unpredictably, and that renders meaningless the notion of projected target volume. Today, the climate of uncertainty created by this decline in clinical simulation with the help of difficult ballistic biases even with the sophisticated sophistication of simulation. virtual permanent and immutable to evaluate, in time and space.

 Unlike the so-called virtual simulation, the new technical concept of clinical simulation introduced by this new invention, which accounts for not only physiological variations but also voluntary movements of the patient on the table around a physical invariant, the target lesion,. as viewed at time t, by the CT. At this point in our analysis, it seems that we discern more precisely the contours of the technical difficulty.

 Because we now have the means to understand its meaning, since an evolutionary expansive process can not, as is the case with the target in radiotherapy, be a frozen being in time on images even repeated several times. to be carried out at a distance and thus censored at a more or less long interval, with respect to the time to, and presented differently in the form of DRRs and CT or non-CT gantry images, which for months will be for a long time still the permanent reference between the initial diagnostic examination, through 2-D and / or 3-D simulations, as well as contouring, until the last end session. ray treatment. Period during which the contours of the lesion and its environment oscillate around the values that only a good estimator can give exactly at the same time as they change as a function of time.

However, computer-based virtual simulation has been conceived as a three-dimensional simulation of an irradiation technique, for which on a representation
3-D of the patient is given a 3-D representation of the irradiation beams

<Desc / Clms Page number 51>

 therapeutic. The mathematical tools of the virtual simulation then give the possibility of optimizing the geometrical characteristics of the irradiation beams, since they make it possible for example to approach the optimal incidence, capable of excluding, for this incidence of a given beam , organs at risk and define the shape of the irradiation fields best suited to the volume to be irradiated.

But only to have the target clearly identified, and to do it right, the only way to do it is to work in real time under CT-guidance. It is around this technical concept that articulates, as in Patent No. 01/01133, the new invention of CT-guided therapeutic irradiation presented below.

2. 6.2. - Delimitation of the volumes of interest The external contours and the spinal cord can be automatically recognized by the delimitation program according to an isodensity line indicated by the operator. The macroscopic tumor volume must still be manually recorded. In general, macroscopic tumor volume includes contrast-capturing tissues and peritumoral edema for high-grade gliomas, for example. This evaluation is, in a postoperative situation, more delicate when the resection has been complete and there is no more contrast enhancement. In this case, the evaluation of the macroscopic tumor volume is based on preoperative diagnostic imaging data, despite the uncertainties due to the different conditions of acquisition of the images (patient position, orientation of the cutting planes). For low-grade gliomas that do not capture the contrast medium, MRI imaging, supplemented by a fusion of CT-MRI images, is found to be critical in improving the delineation accuracy of macroscopic tumor volume. However, something must be found in the viewfinder. Why radiate the margin as much as the tumor, when there are fewer cancer cells at this level? It's using a hammer to kill a fly in the periphery of the tumor. Advantage of the divergent effect of the collision of the particles + isotropic diffusion of radiant energy + effect of the penumbra. Knowing that diffusion essentially means the flux of secondary radiating particles.

* In practice the margins around the volume depend on the histological nature of the tumor, the type of spatial configuration (if this information is known) of the quality of the surgical excision and the proximity of a risk organ. The software must allow anisotropic expansions, that is to say different in the directions with invaginations and bifurcations, etc. the 3D morphology of the lesion, in order to take into account these different constraints or to exclude a

<Desc / Clms Page number 52>

organ whose probability of microscopic tumor invasion is nil.
Some publications emphasize the importance of variations in the evaluation of tumor volume between operators or when the same operator is asked to repeat, after a sufficient interval of time, a tumor volume. When nine different observers (neurosurgeons, neuroradiologists and radiotherapists) reported macroscopic tumor volume in five patients with inoperable brain tumor (C. Weltens et al, 1998). It is the part of subjectivity of the deal! This observation illustrates the difficulties related to the objective determination of tumor volumes as well as the effort that clinicians should make, based on histological, surgical and imaging data, to better standardize the anatomo-clinical limits of the target volume. In this respect, the concept of anatomo-clinical target volume is particularly fundamental in the particular field of brain tumors. The Target Volume 3-D goal must prevail at all. The predicted target volume (PTV) which is, in the current state of the art, inxdifferately 3-D or 2-D is defined by three-dimensional automatic expansion around the anatomo-clinical target volume. The value of the margin, which depends mainly on the immobilization system used, the means and the frequency of the positioning controls used, is generally 5 mm because it is automatically considered that the patients are properly immobilized by the thermoformed mask. and the physiological movements of tumor volume would be negligible (good profession of faith). All these data are then used in the procedures of
Monte Carlo for dosimetry purposes.

2. 6.3. Monte Carlo simulation
The Monte Carlo simulation of radiation transport commonly used is. considered as being, in the ancient art, one of the most exact methods of calculating the dose of radiotherapy. With the rapid development of computer technology, treatment planning, for radiation therapy, based on the Monte Carlo technique consists of detailed knowledge of radiation beams from medical accelerators or other sources different from. radiation. A practical approach to achieve this end is to perform the Monte Carlo simulation of the radiation transport of the medical accelerator or said source. In addition, the Monte-Carlo modeling of the head of the processing machine can for example also improve the understanding of the clinical characteristics of the beam, assist in the development of the accelerator, as well as improve the accuracy of the dosimetry clinic by providing more data

<Desc / Clms Page number 53>

realistic beam. The work on Monte Carlo simulation of clinical electron beams from the past two decades from medical accelerators is summarized in the article by Ma and Jiang (1999). The Monte-Carlo method
Carlo is essentially a statistical simulation method. With regard to the radiation transport problems, it simulates in a given direction of the irradiation beam the traces of the individual particles by sampling appropriate quantities from the probability distributions governing the individual physical processes, using the (pseudo) numbers. -) chances generated by the machine. Mean values of macroscopic quantities, such as fluence. Primary and secondary photons, the energy spectrum and the absorbed dose distribution can be calculated by simulating a large number of particle histories. The Monte Carlo method and its applications in physics of medical radiation have been discussed, especially in physics.
Radiotherapy (Radiophysics), in many publications (Burlin et al, 1973, Raeside, 1976, Nelson and Jenkins, 1980, and Bielajew 1984, 1990, Turner et al, 1985, 1985, et al, 1988, 1990, 1991; Andreo, 1991). This one needs to be supported by a rigorous methodology of ballistic aiming.

 2.6.3.1. - Ballistics of irradiation in therapeutic clinic). At the end of the survey of the different contours, the ballistics of irradiation is determined on the reconstruction of a virtual patient. However, we must find something in the viewfinder by a use of the patient itself which is only possible on a device able to do, during the therapeutic irradiation, real-time images.

 The value of virtual simulation software for customizing and optimizing the processing is not discussed, subject to a good methodology. As such, visualization beam's eye view (BEV) is an interactive tool, which without having the qualities of a diagnostic imaging is considered essential, since it allows the choice while excluding organs at risk of optimal impacts of the beams.

 But, we must deplore otherwise that the object is not in the viewfinder system irra-). therapeutic therapy of art! if not its shadow more or less reach. Similarly, the visualization room's view is an essential representation mode visualizing the envelope, on a three-dimensional representation of the patient, of all the beams. The divergence of the non-coplanar beams and the extent of overlap of the! fields on the cutaneous plane. These methods lead to the choice of the number of beams to be used, their optimal incidence

<Desc / Clms Page number 54>

compared to the outline of the contours of the complex fields. The operator must, in the latter case, compare the adjustment obtained by the use of a multi-blade collimator to that obtained by cerrobend custom caches. The presence in the immediate vicinity of the tumor volume of organs at risk frequently leads to use cerrobend caches, to avoid a stair effect due to the width of the blades (MLC). On the other hand, MLC multi-blade collimators make it possible to automate more and more complex irradiation techniques and can associate, in the same patient, coplanar and non-coplanar beams, beams of beams of different energies, corner filters with collimator rotations, etc. The proposed ballistics is validated by a careful study of the dose distribution in different planes supplemented by the construction of dose-volume histograms (HDV) to quantify the proportions of tumor volume and irradiated healthy tissue. The most commonly used method is Monte-Carlo
Carlo.

;. 2.6.3.2. - Modeling by the Monte-Carlo method
The Monte Carlo method can accurately model the physical processes involved in radiotherapy and is powerful in the processing of any complex geometry. It is widely accepted that Monte-Carlo simulation of radiation transport is one of the most accurate methods for predicting 1. radiotherapy distributions of absorbed dose. In particular, the simulation of
Monte-Carlo can handle backscatter from high-density materials such as bone, or more accurately than any other existing dose calculation model, the disturbance of airborne cavities (Rogers and Bielajew
1990, Nahum 1985, 1988, Mackie 1990, Rogers 1991, Andreo 1991, Bielajew 1994, Mohan 1997a). The major defect of the Monte Carlo method has become, in a nominal way of intensive computing, much less severe because of the rapid increase in the speed and the fall of the cost of the computers, and the use of the innovative techniques of variance reduction (Ma and Nahum 1993, Holmes et al 1993).

The Monte Carlo simulation is thus rapidly becoming practical). Routine clinical planning for radiation treatment planning systems The next generation of dose-calculating machines (Manfredotti et al, 1987, 1990, Al-Beteri and Raeside, 1992, et al, 1995, van der Zee, 1996). Kawrakow et al, 1996, 1997a, 1997b, Hartmann-Siantar et al, 1997;
Ayyangar and Jiang, 1998; De Marco et al, 1998; Solberg et al, 1998; Wallace and
Allen, 1998; Wang et al, 1998; Ma et al, 1999).

<Desc / Clms Page number 55>

A Monte-Carlo processing planning system for incident (s) beam (s) indeed requires detailed information on the patient. In order to initiate particle transport in the patient's CT model, accurate information on the patient's surface of the phase space pertaining to the particles is required. Direct measurement of this information is very difficult, if not impossible, for a clinical beam due to the very high radiation intensities produced in the clinical bundles (Deasy et al, 1996). The calculation of phase phase parameters of the beam using the analytical methods is not flexible and is usually used, for example by ignoring the approximations of the events of the broadcast. higher order Compton (Desorby and Boyer, 1994). Commonly, the most practical way to obtain detailed information on the incident beam is the Monte Carlo simulation of the treatment head (Petti et al, 1983a, b, et al, 1985, et al, 1987). Rogers et al, 1988, 1995a, Udate, 1988, Udate-Smith
1990, 1992; Chaney et al, 1994; Kassaee et al, 1994; Lovelock et al, 1994, 1995; Sixel and Faddegon, 1995; Ma et al, 1997a; Lee, 1997; Hartmann-Siantar et al,
1997 ; Liu et al, 1997; De Marco et al, 1998; Jiang and Ayyangar, 1998; Balog et al, 1999). In addition, the detailed information on radiotherapy beams has a wide variety of applications in clinical physics. The simulation of Monte-
Carlo of medical accelerators can increase the understanding of the characteristics. beams, help in the development of an accelerator and improve the accuracy of clinical dosimetry by providing more realistic beam data (Ma and Jiang 1999) (beams), but it can not nothing on the other hand for the precision of the ballistic aiming.

2.6.3.2.1. Monte-Carlo simulations for photons 5. But many investigations have, during the past 20 years, been carried out on the Monte-Carlo simulation of photon beams from medical accelerators or 60Co teletherapy units. McCall et al. (1978), for their part, explored the effects of the various targets on the average energy of the flattening filters of the photon beams, using, according to Patau et al. (1978) the). EGS3 code (Ford and Nelson, 1978). They did a pioneering work on the Monte Carlo simulation of an entire photon accelerator. They first simulated the production of photons on a W-Cu target, then the transport through a flattening filter of said photons between the collimators, as well as the attenuation of these photons in the plates of the various materials. Nilsson and
Brahme (1981) explored contaminating photons, scattered from

<Desc / Clms Page number 56>

 flatteners and collimators. Other studies have been conducted. Thus to investigate the electron contamination of photon beams, Petti et al. (1983a, b) have for example simulated in great detail a head of a processing machine, using a conditioning of cylindrical geometry to approach the various components of the linear accelerator. While for calculating photon spectra and fluence distributions from multiple accelerators, Mohan et al. (1985) performed similar detailed simulations based on the EGS3 system. A special conditioning geometry has been developed to model the exact shape of the flattening filter. Han et al. (1987). also used the EGS3 code to simulate in detail the processing head of a 60Co unit. On this occasion, they approached the complex geometry of a 60Co unit, the Theratron-780 source sheath as the source capsule and the entire collimator. Similar studies of clinical photon beams have been performed, using the EGS4 code system (Nelson et al 1985). Rogers and. al. (1988) instead explored sources of electronic contamination in a 60Co beam. Chaney et al. (1994) simulated the beams of a 6 MV photon accelerator to study the origins of the diffused head.

While Lovelock et al. (1994) simulated photon beams from a Scanditronix MM50 machine to obtain beam characteristics. For the treatment planning, an EGS4 user code, McRad, was the generic photon Monte Carlo model of the linear accelerator, and was also developed by Lovelock et al. (1995). Sixel and Faddegon (1995) simulated a Therac-6 treatment head, in radiosurgery mode, using the symmetric EGS4 user code, FLURZ, with the PRESTA algorithm (Bielajew and Rogers, 1987). They calculated the 6 VM MV radiosurgical X-ray spectra with and without a flattening filter and compared the results with the analytical spectra of the Schiff thin target and the flattened Monte Carlo spectrum calculated by
Mohan et al (1985). To study in the meantime the effect of differential hardening of the beam by the flattening filter, Lee (1997) simulated the beam of 6 MV from. a Varian Clinac 2100C Accelerator using the EGS4 code. Liu et al.

(1995) used Combinatorial Geometry conditioning to characterize the components of the processing head, which as modeled 3-D objects were combined using Boolean algebra. Simulations were performed with the ACCEPT code of the Monte Carlo ITS 3.0 System (Halbeib and Mehlhorn,
1984). The MNCP mode of Monte-Carlo (Hendricks and Briesmeister

<Desc / Clms Page number 57>

 1992; Briesmeister, 1993) was also used to model clinical photon beams. To determine the parameters used for the dose calculation in their PEREGRINE photon source model, HartmannSiantar et al. (1997) simulated linear accelerators (linacs), using MCNP and EGS4 / BEAM code (Rogers et al 1995a). De Marco et al. (1998) simulated the photon beams from the Philips SL-15/25 linear accelerators for phase-space information to calculate the dose in a CT-phantom of the patient. Lewis et al. (1999) also simulated a Philips SL 75/5 linear accelerator, using the MCNP code.

Another Monte-Carlo system, PENELOPE (Salvat et al, 1996), has also been used to simulate photon beams from a Saturn accelerator 43 (Mazurier et al, 1999). The EGS4 / BEAM code (Rogers et al., 1995a) has been used extensively to simulate the photon beams of the various types of linear accelerators Liu et al. (1997) simulated photon beams from a Varian Clinac 2100C machine and developed an extrafocal source model, analyzing the results of the simulation, for the calculation of the dose using a convolution / superposition method (Mackie et al. al., 1985). Jiang and Ayyangar (1998) simulated a Varian Clinac 1800 accelerator and studied the effect of disturbance of the compensator on the characteristics of the photon beam. Their simulation results were also used to explore the feasibility and the need to develop a Monte-Carlo radiological surgical treatment system (Ayyangar and Jiang, 1998) and to study induced dose disruption in small radio beams. -surgical, by the inhomogeneities of high density (Rustig et al, 1998).

A more detailed report on the simulation of clinical photon beams using the EGS4 / BEAM code was given by Sheikh-Bagheri (1998). Balog et al. (1999) have more recently investigated inter-leaflet multi-blade collimator (MLC) transmission by simulating the NOMOS MIMIC MLC attached to a GEORION 4 MV linear accelerator. Their results have been incorporated into the design of a MLC used in a prototype tomotherapy machine (Mackie et al, 1993, 1995). By simulating the processing head of a Siemens MXE accelerator, Faddegon et al.

(1999) developed a new flattening filter for the 6 MV photon beam of this machine. Verhaegen et al. (1999) applied the EGS4 / BEAM code to the simulation of X-ray kilovoltage (kV) radiotherapy units. Ebert et al. (1996) also reviewed some recent work on modeling both

<Desc / Clms Page number 58>

 photonic and electronic clinical bundles together, using both analytical and Monte Carlo methods (Ma and Jiang, 1999). It is therefore a statistical modeling and a mode of operation that adapts to the different conditions of therapeutic irradiation, particularly to electro-therapy, and which can be applied to the simultaneous multi-beam orthovoltage X-ray of the present invention.

2.6.3.2.2. Monte Carlo simulations for electron beams
The electron-beam Monte Carlo dose calculation was considered to require more detailed information of the beam phase space at the surface. of the patient as the photon beams (Mackie 1990). The difficulty was considered, compared to the simulations of the photon beams, as that of the simulation of the beams of the accelerator model (Rogers 1991). The work on Monte-Carlo simulation of electron beams can be rigorously grouped into two categories. The simulations include in the first category 5. were mainly executed, there are about ten years, with the limited computing power. A medical accelerator needed to be simplified, in order to perform in a reasonable time frame and acceptable accuracy, the simulation of
Carlo, one or two components of more dosimetric significance, (Berger and
Seltzer, 1978, Borell-Carbonell et al., 1980, Rogers and Bielajew, 1986, Manfr-). dotti et al., 1987, Andreo et al., 1989). Some studies have been carried out in this way, or even more recently, for investigator the influence of a component of the accelerator on the characteristics of the electron beam, at a relatively low calculation cost (Keall and Hoban, 1994; and Hoban,
1995b). The second category includes the Monte Carlo simulation of an integral accelerator 5. (Udale, 1988, Udale-Smith, 1990, 1992, Kassau et al, 1994). This category has been, since the OMEGA BEAM system was developed (Rogers et al, 1990,
1995a, b; Mackie et al, 1990), greatly enhanced. Since then, the modeling of
Monte Carlo medical accelerators has become easier and more systematic. Because of its significance and the large number of applications, the). OMEGA BEAM system is a separate system just as the modeling of
Monte Carlo simulated by electron beams using different source models, which is an approach to form the link between the Monte Carlo simulation
Carlo of electron beams and Monte Carlo processing planning (Ma and Rogers 1995a, b, c, Ma et al 1997a, Faddegon et al 1998, Ma 1998, Jiang et al 1999). Ma and Jiang (1999) finally discuss some existing problems as well as

<Desc / Clms Page number 59>

 as possible future directions for the Monte Carlo simulation of electron beams generated by medical linear accelerators.

Modulated electron radiation therapy (MERT) has been proposed as a means of delivering the conformational dose for irradiating shallow tumors while sparing distal structures and surrounding tissues. Conventional electron beam collimation systems are a laborious and intense moment in their construction and so are, according to Mr. C. Lee, S. B. Jiang and C-M. Ma (Monte Carlo and Experimental Investigations of Multileaf Collimated Electron Beams for Modulated Electron Radiation Therapy, Med Phys 2000 Dec; 2708-18.) Unsuitable for use in the sequential delivery of multiple complex fields required by the MERT. In addition, high energy electron beams in the 150-250 MeV range are being studied to evaluate the feasibility of radiotherapy. The Monte Carlo simulation results from the PENELOPE code are presented and used by DesRosiers et al. (2000) to determine the lateral scatter and the penetration of these beams. Thus it has been shown that penumbra is comparable to photon beams at depths of less than 10 cm and the practical range (Rp, practical Range) of these beams is higher than 40 cm. The depth distribution of the electron beam dose compares favorably with the photon beams (DesRosiers et al, 2000).

 If the virtual simulation still opens, in the choice and the blind optimization of the therapeutic irradiation technique, new perspectives that say then of the statistical method of simulation and clinical estimation? In addition to important technological means, this approach, which is called virtual simulation, which is known to be inaccurate, nevertheless requires a suitable quality assurance program and a multidisciplinary team. On the ground the reality is moreover much more complex than it seems. The very first steps outlined above allowed to deepen the initial interrogations and to temper the temptations of immediate answers, with regard to the quality of the modeled objects.

But these were also, as we will see below, confirmed and they finally allowed to organize in a coherent speech, thanks to a certain number of assumptions that were injected as such in the proposals for daily work. of
Radiotherapist. But, we know all too well, we can not with impunity combine so many cumulative errors of measures and estimates, whose sanction is nothing but increasing vagueness, and still aspire to aspire to a certain science of science. quality, of which there is no tool other than subjective to control and

<Desc / Clms Page number 60>

 confirm the self-satisfaction of each other with regard to their own process.

 In every age, scientific thought evolves within the limitations imposed by conceptual tools and available methods of observation.

 This is to say what a society gives itself as real and as possible. Otherwise you have to do better, faster and cheaper! Here the virtual simulation, made possible by the computerization rampant and the statistical method of the clinical simulation which is lacking appropriate tool more and more distant from the daily practice of Radiotherapy. The lack of precision of the analytical apparatus used in current art necessarily leads to gaps and most importantly. bias that can not be solved by statistical physics alone. There is a demand at this time unfulfilled of the use of the physically measurable reality (the clinical reality) where there would be a total absence of variability in the coordinates and outlines of the target or target lesion, as well as the ballistics of the rays that 'reach like an object frozen in time and space.

. In these conditions, the variances around the lesion are always zero and the topography of the lesion is not only falsely predictable but never instantly updated and therefore effectively updated. Current approximations of the art around the target lesion in its environment and the different variations that accompany them, including contours and structural density,. no longer allow to observe this clinical situation in a simple way.

 The current approach is such that if the facts are authentic, they are taken at random from among a population of ballistic phenomena whose deviations are large and unquantified in relation to the object aimed at. While accurate estimates with near-zero variances are, as we will demonstrate in the invention that. we describe below, today possible. It is not the virtual simulation that should be performed but rather a pre-therapeutic physiological behavioral modeling by the instantaneous statistical simulation and estimation method of the present invention.

 The partial control of all these analyzes by a single individual is of course not possible. It is only thanks to the numerous loans made here and there to the large number of recent publications of the literature which have had the avowed objective, from a 3-D perspective, of perfecting the treatment planning and thus making the actual development of the phase of treatment, in its different stages. So restoring on their philosophical and scientific background, the simulation and planning theories of the organization and preparation of the treatment itself.

<Desc / Clms Page number 61>

 rays could be considered, with the new invention, as major points of rupture of radiotherapeutic practice. In this new framework, the pioneering research and statistical hypotheses recounted in the abundant literature briefly cited above are certainly better received. They are so much more statistically significant that they do not propose a radical upheaval, but rather a relevant reorganization and a rigorous combination of diverse and diversified operations that do not accumulate their methodological errors. But we are not able to systematically study such fine distinctions. Only this way of doing things by breaking with the old art makes it possible, by spreading relationships, to study during treatment execution, between target and non-target structures, knowing that the central notion here is conformational radiotherapy, in which no aiming bias can no longer be tolerated.

 The well-localized proton dose deposition involves, in clinical applications. the need for good geometric accuracy of proton beams.

In most proton therapy installations the proton beam is delivered with a fixed line, usually horizontal, of the incident beam and multiple fields applied by rotations and / or only translations of the patient's positioning system. With the introduction for proton beams @. (Slater et al., 1992, et al., 1995, Flanz et al, 1995), sometimes in combination with a patient positioning table having more than three degrees of freedom, the irradiations from many of the directions different, including non-coplanar directions, were greatly facilitated. In addition to this, these stands are substantially larger and / or heavier than the electron stands of medical linear accelerators. These developments have made quality control of the geometric accuracy of the integrated system indispensable.
The accuracy of the beam incidence relative to the target volume with the patient is clinically important. Putting aside the difficult problem of knowing exactly where the target volume is relative to the positioner part (eg, the top of the patient support) on which the patient is installed, this means that one wishes to perform a good beam-to-positioner accuracy. It is common to control separately the movements of the stand and those of the patient support.

 However, in such an approach, measurement errors may accumulate and it is preferable according to Barkhof et al. (1999) to directly measure the accuracy of the beam-at-positioner.

<Desc / Clms Page number 62>

 This is the case of the Northeast Proton Therapy Center (NPTC) patient positioning system (PPS) with three translational and three rotational movements (Flanz et al, 1999), which provide six degrees of freedom, but do not offer the possibility of native rotation around a specific point in space. Movements rather composed of several commands, under computer control, maneuvers are required to perform a rotation around a given point in space. The isocenter of the PPS can be located anywhere inside the manipulator's workspace; but it is, of course, chosen to coincide with the isocenter of the stand. However, the accuracy with which this occurs depends on the mechanics and the software working properly. To ensure correct alignment, it may be necessary to perform daily isocentricity and system alignment checks, including the stand as well as the PPS. A tool has been developed, with respect to a reference point, for checking the alignment of the beam which is attached to the top of the patient support of the PPS, so that it can coincide with the isocenter of the stand. The tool allows, with respect to this reference point, rapid beam alignment measurements at all stile angles and at all possible PPS orientations. This approach has evolved on the linear accelerator to a system used in stereotactic radiosurgery (Lutz et al, 1988), where a snapshot is used to determine the shadow of a steel ball. This prototype has been evaluated for use with photon beams, but could be used equivalently for X-ray beam alignment checks of a linear accelerator or other source (Barkhof et al, 1999).

 2. 6.4. - Reality of the delivery of the dose The permanent organization of the ballistic-target confrontation is in the image of this double movement: increase the total dose to the target volume and reduce the dose to the surrounding structures. The tumor volume to target volume ratio on the one hand and the target volume ratio on surrounding healthy tissues materialized by a series of 3-D cuts and reconstructions, related to the target, visualizing in practice the different incidences of the beam with their characteristics. is, in line with the patented embodiment of No. 01/01133, the main objective of this new invention. Several trials have also confirmed the value of evaluating radiotherapy errors.

<Desc / Clms Page number 63>

 2.6.4.1. Margins of security and uncertainties of the location of the target But, uncertainties in the repositioning of the patient and the movement of the organs are certainly taken into account, defining in the prior art a target volume of planning (PTV), without any means however, to evaluate them exactly in particular with respect to the variability, in space and time, of their relative topography. Recommendations are found, according to Craig et al. (2001), in existing protocols not explicitly discussed here for the development of PTV. A quantity called coverage has also been defined by these authors to quantify how much a PTV actually surrounds the clinical target volume and how it is applied, in order to examine the impact of several factors. Stochastic simulation is used to determine the coverage required for a desired balance between the probability of tumor control (TCP) and the irradiated volume. They evaluate using a sample of the anatomy, the importance of the method used to then add the uncertainties, the form of distribution of uncertainties, the effect of systematic uncertainties, and the use of non-uniform margins. Craig et al. (2001) also examined the benefit of patient immobilization techniques. This example indicates that a goal of 95% coverage is, for these authors, reasonable (but difficult to verify during treatment) for treatment planning. Using this example as a comparison value to indicate a quadrature addition of uncertainties, they predict smaller margins (7 mm) than by linear addition (11 mm). If the Gaussian distribution of uncertainties (7 mm) requires the same margin as a uniform distribution (7 mm), the systematic uncertainties have, according to these same authors, a small effect below a threshold value (4 mm) on the TCP and non-uniform margins allow, according to Craig et al. (2001), a slight reduction in irradiated volume. They recommend that uncertainties can generally be added to squared, that the exact form of the uncertainty distribution is not critical, that systematic uncertainties can be kept below a threshold value, and that non-uniform margins can be when the uncertainties are anisotropic and unquantifiable with respect to a target that is never yet, to be effective in space and in time, as we imagine it, frozen. So there is already at this level a reasoning bias to which will be added many more measurement.

On the other hand, the need to oxygenate said safety margins around the target volume makes the solid and necrotic tumors of the brain respond by

<Desc / Clms Page number 64>

 for example, although necrotic lesions often contain a significant proportion of hypoxic cells that can not be re-oxygenated during the short time of single dose application of overall treatment more easily to radiosurgery. But, the peripheral vascularization provides enough. In addition to the direct cytotoxic action, these hypoxic cells would retard, according to Kocher et al., Vascular occlusion followed by ischemic cell death of the tumor and they could contribute to the effect of radiosurgery. Our belief is that too much sprawl giving way to a cellular repair that is too effective and very harmful for the peripheral neovascularization of the surrounding structures does not go in the direction of more tumor control. Therefore, to determine the impact on the tumor response of the two possible effects, a 3-dimensional computer simulation was developed and performed by Craig et al. (2001) to respond to data obtained from 90 patients treated with a linear accelerator by radiosurgery for 1 - 3 brain metastases with average marginal doses of 20 Gy. Complete response rates were, for small solid lesions (diameter 0.4 - 1 cm), 52% (12/23); for large solid lesions (1.1 - 5.2 cm) of 28% (17/60); and, for large necrotic lesions, 12% (6/50). The 3-dimensional model of the computer simulated the size of small solid tumors and large tumors, solid and necrotic, located in a vascularized stroma. The supply of oxygen, the cell division of the tumor by irradiation of a single beam at a single dose (linear / quadratic model, alpha / beta being equal to 10 Gy, the ratio of oxygen uptake being equal to 3.0) and time-dependent vascular occlusion (alpha / beta = 3 Gy) were modeled by Monte Carlo simulation techniques, the results of which are shown below.

 2.6.4.2. The Oxygen Effect on the Efficacy of Radiotherapy In the presence of neovascularization, solid tumors developed a hypoxic fraction of 1-2%. Without neo-angiogenesis, central necrosis occurred, and the tumors had a hypoxic fraction of 20-25%. By simulating the pure cytotoxic effect of radiosurgery, neither the dose-response relationship of solid lesions of different sizes nor that of large lesions having a solid, necrotic appearance could be reproduced at any given level of radiosensitivity. This occurred only by introducing a vascular effect, which led to the occlusion at the periphery of the target volume of more than 99% of these vessels, within a year after irradiation. In the presence of vascular effect, the apparent radiosensitivity of the tumor cells was increased by 50-100%. Calculations of the equivalent dose

<Desc / Clms Page number 65>

 The vascular effect shows that it contributes 19-33% to the overall effect of single-dose radiosurgery. We can not therefore bet that the reduction in the number of fractions will benefit, as in Radiosurgery, from the oxygen effect. This simulation study suggests that the therapeutic effects of single radiosurgery in malignant brain tumors can not be understood without taking into account vascular effects. The computer model could serve as a basis for exploring new treatment modalities that modify the two cytotoxic and vascular effects of radiosurgery (Kocher et al, 2000), particularly as we have already proposed by the Bi-linac Imatron device. adopting resolves. negative safety margins or the HBO technique.

 In another publication (Y. Kinoshita et al 2000), hyperbaric oxygen (HBO) has been proposed to reduce hypoxia by increasing dissolved molecular oxygen in tissues. Using a magnetic resonance imaging (MRI) technique, Y. Kinoshita et al. showed that the changes after. HBO exposure, MRI signal intensity is explained to the extent that paramagnetic molecular oxygen dissolves and shortens the T1 relaxation time. To do this SCCVII cells transplanted into mice were used. NMR enhanced molecular oxygen images were acquired using a 4.7 Tesla NMR system with an inversion-preparation sequence. fast, slow, and firing angle recovery (IR-FLASH) sensitizing the paramagnetic effects of molecular oxygen. The NMR signal of the muscles decreased rapidly and returned within 40 minutes after decompression at the control level, whereas that of the tumors decreased gradually and remained 60 minutes after HBO exposure at a high level. In contrast, in the normo oxygen group. bare, the signal from the tumors showed no significant changes.

 These data suggested that changes in the NMR signal of tumors and muscles represent an extra-vascular oxygenation alternative. The preservation of the tumor concentration of oxygen, after exposure to HBO, may be in view of the adjuvant therapy of important cancer patients. That is why . Since this is very important and is important in the likelihood of tumor control or complication, we have always advocated taking negative safety margins so that, in the absence of HBO, a peripheral zone of neovascularization, which is reached only indirectly by the peak of diffusion of the electromagnetic interaction in a tumor medium.

<Desc / Clms Page number 66>

2.6.4.3. Probability of tumor control and probability of complication Radiation treatment is based on a more or less physiological statistical method, a statistical variable with two modalities: tumor tissue and non-tumoral tissue. We know that the probability distribution of a dichotomous variable is entirely characterized by the probabilities PI on the one hand, for the proportion of sick cells killed, and P2 on the other hand, for the proportion of healthy cells killed or complications. It is here that the interval of fluctuation of the dose takes all sense at the same time as the importance of the instantaneous or scanoscopic visualization of the lesion with respect to its environment becomes clear. In the case of concave-shaped PTVs including, for example, the prostate (P) and seminal vesicles (VS), modulated intensity radiotherapy (IMRT) is expected to improve the therapeutic treatment ratio of cancer. of the prostate. With this in mind, Fiorino et al. (2000), to compare IMRT by simple 1-D modulations and conventional 3-D conformational therapy (ie non-IMRT) in the treatment of concave-shaped PTVs including P + VS, previously treated with five patients with PTV concave form (P + VS) fully treated at their Institute with conformational radiotherapy, conformation plans of the 3 and 4 fields were compared, in terms of biological indices, with the IMRT plans. The IMRT plans were generated using five equidistant beams with partial rectal protection available through the single-absorber or single-absorber modulation technique (Fiorino et al,
1995). The modulation was uni-dimensional and the shape of the bundles was at least unique in correspondence with the rectal 'core'; in this minimum, the intensity of the beam was 20 or 40% of the intensity of the aperture beam.

All the plans were simulated, using a brush-based algorithm (with 18 MV radii) on the CADPLAN TPS. The approach is such that if the tumor control probability events (TCP = Tumor Control
Probability) and the probabilities of complication of normal tissue (NTCPs = Normal
Tissue Complication Probabilities) are authentic, they are taken at random from a population of accounting phenomena in the variances from which intervene, as we will see later, several biases difficult to quantify.

 However, because of the growing power of computing, these unbiased estimates are now statistically possible. Another major point to discuss is the therapeutic escape. Treatments as powerful as the dose escalation have indeed their flaw on which the ancient art does not have the means

<Desc / Clms Page number 67>

 to put one's finger on it, without a prior extemporaneous visualization of the objective aimed at firing the therapeutic irradiation fraction proposed by the present invention.

In order to evaluate, moreover, the use of inhomogeneous distributions of the iose-target, obtained with conformational 3-D radiotherapy, with or without beam intensity modulation, WR de Gersem et al. (1999) planned, in order to offer the possibility of not increasing the normal tissue toxicity indices and / or increasing the tumor control indices of stage III non-small cell lung cancer (NSCLC = non-small cell lung cancer), the treatment of 10 patients with stage III NSCLC cancer, using a 3-
Conventional D and a technique involving non-coplanar beam intensity modulation (noncoplanar beam intensity modulation = BIM). To do this, two target planning volumes (PTVs) were defined: PTV1 included macroscopic tumor volume and PTV2 included both macroscopic and microscopic tumor volume. The virtual simulation defined beam shapes and impacts for both the wedge (3-D) and the segment contour (BIM) orientations. The weights of the filtered beams, the unfiltered bundles, and the segments were determined by optimization using an objective function with a biological component and a physical component
The biological component included the probability of tumor control (TCP) for PTV1 (TCP1) and for PTV2 (TCP2), as well as the probability of complication of normal tissue (NTCP) of the lung, spinal cord, and c # ur.
The physical component included both the maximum and minimum doses as well as the standard deviation of the PTV1 dose. The most inhomogeneous distributions of the target dose were obtained, using only the biological component of the objective function (biological optimization). By allowing in addition to the biological component the physical component, the inhomogeneity of PTV1 appeared to be reduced (biophysical optimization). Like normal tissue toxicity indices, NTCP values as well as maximum doses or dose levels to the relevant functions of organ volume were used.

When the optimization, in relation to the objective biological function and the biophysical function, was carried out the inhomogeneity of the PTV1 decreased for the 3D- from 13 (8-23)% to 4 (2-9)% (p = 0.00009) and for BIM plans of 44 (35-56)% to 20 (9-34)% (p <0.00001). The minimum doses of PTV1 (expressed as the lowest dose-voxel) were similar for both objective functions. Target doses

<Desc / Clms Page number 68>

mean and maximum with biological optimization were significantly higher for 3-D as well as for BIM (p <0.001, for all values).
The probability of tumor control (estimated by TCP1 x TCP2) was 4.7% (3-D) and 6.2% (BIM) higher for biological optimization (p = 0.01 and p =
0.00002 respectively). The NTCP (lung) as well as the percentage of lung volume exceeding 20 Gy were, according to WR de Gersem et al. (1999), higher with the use of objective biophysical function. NTCP (heart) was also higher with the use of the objective biophysical function. The percentage of heart volume exceeding 40 Gy tended to be higher, but the. difference was not significant. For the spinal cord, the maximum dose as well as the NTCP (cord) were, for 3-D (D (max): p = 0.04, NTCP: p = 0.2) similar, but when the function objective biophysics was used, they were significantly lower for BIM (D (max): p = 0.002; NTCP: p =
0.008). Using conventional 3-D techniques, it seems that the distribu-. Inhibitory doses offer potential to further increase the probability of uncomplicated local control. Using techniques that, such as the
BIM would lead to a high escalation of maximum target doses, they appear to be indicated to limit the inhomogeneity of the target dose in order to avoid dose levels that are so high that safety becomes problematic (WR de Gersem et al. 1999).

These authors considered that optimization improves the homogeneity of 3-D planes, the
BIM of the dose at the target and the minimum dose at PTV2 was especially diminished by optimization in 3-D planes. After optimization, the dose-volume histograms (DVHs) of the lung and heart were shifted (changed) to the doses. lower for 80-90% of the volume of the organ. Since the lung is the dose limiting organ in stage III NSCLC, a minimal dose to PTV1 increased at the same time as a decreased dose to the main lung volume suggested an improved therapeutic ratio. Optimization would allow 10% dose escalation for 3-D plans and 20% for BIM plans at levels. isotoxicity of the lung and spinal cord. With respect to dose escalation, the # esophagus could become a dose-limiting structure when PTV1 extends close to the dose. Optimization using an objective biophysical function allowed, according to WR de Gersem et al. (2000) an increase for the
NSCLC Stage III Therapeutic Ratio of Radiotherapy Planning. It emerges from this study that taking into account the physical facetors

<Desc / Clms Page number 69>

 that anatomical has the virtue of refining even more the delivery of the dose to the target volume.

 2.6.4.4. - Delayed simulation on dedicated radioscanners The integration of a CT scanner and a treatment simulator, Sim-CT, Elektra Oncology Systems, Crawley (UK), for tomodensitometry was studied in a clinical situation by D. Verellen et al. (1999). Image quality, hounsfield units (HU) and linearity were evaluated for both implantations and treatment planning. The additional dose to the patient was also brought to light by these authors (D. Verellen et al, 1999). The prior art CT simulation system offers the possibility of an integration of CT-based planning target volumes, which makes it possible to work without conventional simulation. Schiebe et al. report the quantitative evaluation of the alignment accuracy of the field and the position of the isocenter, using the new virtual simulation system, AdvantageSim 4.1 (GE), for the different treatment regions. The image data is, in the study by D. Verellen et al. (1999), acquired using a row of X-ray solid detectors attached to the outside to intensify the image of the simulator. Three different fields of view (FOV: 25.0 cm, 35.0 cm and 50.0 cm) with 0.2 cm; 0.5 cm and 1.0 cm of cutting thickness can be selected and the system has an opening diameter of 92.0 cm, at standard isocentric height. The performance of the CT scanner was characterized by several criteria: spatial resolution, contrast sensitivity, geometric accuracy, hounsfield unit reliability, and radiation production level. The spatial resolution gauge of the Nuclear Associates Quality Phatom (NAQP) as well as the Modulation Transfer Functions (MTFs) were applied to evaluate the spatial resolution. The contrast sensitivity and UH measurements were performed using NAQP and a conversion phantom that allow insertions of the different electron densities. CT-option computed tomography dose index (CTDI) was monitored with a pencil-shaped ionization chamber. Treatment planning and calculations, for dose heterogeneity correction, based on Sim-CT images, generated from an anthropomorphic phantom as well as from the 10 patients, were compared to the similar treatment, based on identical images, but delayed from a diagnostic CT (DCT).

<Desc / Clms Page number 70>

 While to study the variation of computed tomography CT (UH) CT scans and the effect of this variation on dose calculations, J.C. Chu et al. (2000) proceeded as follows: CT images of a cylindrical phantom with multiple insertions were obtained, using a commercially available CT simulator (Ximatron: Varian, Palo Alto, CA).

Linear and Chi-square correlation methods of the coefficient were used in a phantom to determine the effective X-ray energy. The CT figures in Hounsfield units (HU) were a function of the size of the phantom, orientation, the field of view (FOV), the distance from the center, and the measured time for the various insertions. The modification of the dose calculations due to the variation of CT scan numbers was then determined, using equivalent path length (EPL = equivalent pathlength) and compressed conical convolution methods. A significant beam-hardening effect was observed for the CT simulator. The CT figures of the Sim-CT were therefore more sensitive to ghost size than those of a conventional CT. The Sim-CT figure is not sensitive to internal phantom locations and is stable over a 6 week period. It is important to use the appropriate field of view (FOV) for Sim-CT studies; by scanning a small polystyrene phantom a large field of view may result in an increase of 120 HU in the CT number at the center of the field. However, the variations in the dose calculation, due to the uncertainty of the CT number, do not exceed 2-3% for photon beams of 6-18 MV. CT images of the simulator have been acquired with patients in the treatment position, and these CT numbers are, for these authors, useful for dose calculations based on CT CT (Chu et al, 2000).

The last row of holes, which are resolved, in the spatial resolution gauge of the NAQP, according to the field of view and the reconstruction filter applied, either at 0.150 cm or at 0.175 cm. Analysis of linear regression of UH versus electron density revealed a correlation coefficient of 0.99. Contrast, pixel size and geometric accuracy are part of the specifications. The values of the computed tomography index of 0.204 Gy / As and 0.069 Gy / As were observed at the center of a phantom with dose measurements of 16 cm and 32 cm respectively. diameter for the large field of view. The CTDI values of small field of view fields of 0.925 Gy / As and 0.358 Gy / As which are, under similar acquisition conditions, a factor ten times higher than the results obtained from a field of view. diagnostic radioscanner (DCT).

<Desc / Clms Page number 71>

 Ghost studies have shown an excellent agreement between the distributions generated by the Sim-CT and the UH of the diagnostic radioscanner. The deviations between the calculated monitor unit settings as well as the maximum dose in the three dimensions were lower, based on either pelvic or thoracic simulations, for the 1% treatment plans on these two HUs together. Patient studies seem to have confirmed these results. The CT-option can be considered as an added value to the simulation process and the images acquired on the Sim-CT system, which are suitable for calculating the dose with correction of tissue heterogeneities. The good image quality is, however, compromised by the values of the high patient dose. The considerable load of the conventional X-ray tube, according to D. Verellen et al. (1999), frequently the Sim-CT with seven image acquisitions per patient and the system is therefore limited in its ability to perform full three-dimensional reconstruction.

, Another dedicated scanner: the ACQSIM Marconi has a surgical arch on the side of the scanner associated with a stereotactic arm, for the interventional act, with a platform Pinpoint. This is a 4th generation radioscanner dedicated to the
Radiotherapy with all bundle building software and integrated dose calculations. It can only consider images from Marconi scanners because of the patient reference problems that occur on the scanner. Radioscanners dedicated or not, responding to these applications are available at most manufacturers. It is necessary, at this point in the description, to take stock of the information provided by the constitution of the treatment plans at the shelves. CT and simulation are done independently of the therapy itself and away from it. The target lesion that is daily affected by time remains within a given confidence interval, which is not necessarily that of the fluctuation of the target and its environment, not to mention the existence of possible biases. relation to the presence in time and space of the target. Whatever is done in today's art, the topography and extent remain because of technological deficits always uncertain. Added to this, as will be discussed below, are the heterogeneity of the surrounding structures, the physiological variability of the external contours, table-top contention, immobilization, and verification of accuracy. reproducibility each time of the patient's position on the table.

<Desc / Clms Page number 72>

 2.6.4.5. - Taking into account tissue heterogeneities CT radioscan may be better than any other imaging modality for better reading intralesional and perilesional heterogeneities. In Radiation Therapy, radiation treatment bundles contain valuable information for verification and adjustment of the patient. These beams can be used to perform a CT portal reconstruction. However, the direct use of the reconstruction of the beam data can simply produce inadequate CT images because these beams only cover a part of the patient's body. Huaiqun Guan et al. used additional information from the processing beams, in addition to a series of regular beam CT projection, to reconstruct a locally enhanced gantry CT image. This approach is called adaptive gantry CT reconstruction. A computer simulation has demonstrated the benefits of this approach. The reconstruction of the image was performed by the algebraic reconstruction technique of a multilevel scheme, which takes time. The results, according to Huaiqun Guan et al. (2000), would indicate that the quality of the adaptive gantry image could not only be valid for verification of three-dimensional radiotherapy, but also applicable to diagnostic CT imaging.

 2.6.4.5.1.- Conditions of uncontrolled heterogeneity of the dose distribution (internal movements of the Target Volume and voluntary and involuntary external movements of the patient and the operator).

Three-dimensional (3-D) treatment planning has often been performed, when the patient is breathing freely, with the working hypothesis that computed tomography (CT) images represent the average position of the image when in reality those These are usually made in forced apnea. S. Shimizu et al. (2000) investigated, for example, the impact of respiratory motion on CT images of small lung tumors, in spontaneous respiration, using sequential CT scanning with the same table position. Using a single CT scan preparatory to spontaneous breathing, the patient support was fixed at a position where each tumor showed the maximum diameter of the image.

For 16 tumors, more than 20 sequential CT images were taken every 2 s, with an acquisition time of 1 s occurring during spontaneous respiration.

For each tumor, the distance between the top plate of CT CT and the posterior border of the tumor was measured to determine whether the tumor margin was, during normal breathing, sufficiently included

<Desc / Clms Page number 73>

in said intended target volume (PTV). In the sequential CT scan, the tumor itself was not visible in the examination slice in 21% of the cases (75/357). There were statistically significant differences between upper lobe tumors (39.4%) and lower lobe tumors (0%, 0/89) (p <
0.01) as well as between the tumors of the lower lobe and the tumors of the middle lobe (8.9%, 4/45) (p <0.01), in the incidence of the disappearance of the tumor of the picture. The average difference between the maximum and minimum distances between the upper surface of the CT table top and the posterior edge of the tumor was
6.4 mm (2.1 - 24.4). Three-dimensional planning for the treatment of lung carcinoma could significantly under-dose many lesions, especially those of the lower lobe, to the extent that there is a sighting bias. However, an excessive safety margin could, on the contrary, call into question any additional benefit of the 3-D treatment and consequently the technical concept itself of the conformation of the therapeutic irradiation, which requires a certain rigor in the accuracy of ballistic sighting. More research is still needed so to determine how to control not only the respiratory movement; but all voluntary and involuntary movements around the exact topography of the target lesion, over time.

2.6.4.5.2. - Gantry CT irradiation monitoring
The other aspect of use of the CT CT scanner is the in situ escalation of the dose that allows more precision compared to the methods currently in use in the ancient art. It is a fact that when Radiation Therapy (RT) is used to treat cancer patients, its therapeutic index can be significantly increased if the volume of the dose is cut to exactly equal the target volume while sparing the adjacent normal tissue . Modern conformal radiation therapy seeks to address the outcome by using modulated intensity X-ray beams to deliver high doses of radiation to the target, while protecting normal tissue. However, this method relies on strict specifications for processing beams, including the stability of the target volume.

. This is why we often stop at a single scan section for anatomical data and measurements on ghosts. Verify conformational treatment with a combination of computer-assisted megavoltage tomography and gantry imaging would tend to increase in importance. In recent years, many feasibility studies to use reconstruction
CT of megavoltage have been reported. These studies used conical beams

<Desc / Clms Page number 74>

 or fan-shaped beams associated with continuous data acquisition or a sequential, step-and-shoot approach. They also used detectors including a GO scintillator, a row of CsI crystals (TI), a xenon chamber, and portal images. However, the dosage involved in all these studies restricted their extension to clinical application. Direct use of radiation treatment beams would be desirable for portal CT reconstruction since these beams can not provide an additional exposure dose to the patient. This new approach is, however, problematic since the treatment bundles only cover a portion of the patient's cross section. Not only the truncated data can not give a complete map of the exact attenuation coefficient, but also the reconstructed local region does not have surrounding structures around it. Such non-diagnostic quality reconstruction may in practice limit the use of such images in placement verification. Even if the projection is not truncated, the number of beams may be too small (usually less than 10) to provide useful quality CT images, unless arc or cyclotherapy (rotation) is used.

To grasp this problem, Huaiqun Guan et al. (2000) propose, as we have already seen, to acquire, before radiotherapy, a small number of complete projections (i.e., covering the patient's cutaneous contour), with a low dose of radiation. The truncated projections acquired from the treatment bundles could then be added to the complete projections to locally and adaptively and synthetically enhance the quality of the image around and within the tumor. This adaptive gantry CT technique may also be applied to diagnostic radiography (CT) (local tomography or tomography of the region of interest) to achieve a dose reduction. Previous local tomography studies had previously indicated that truncated projection data altered the quantitative accuracy of image reconstruction. A general approach to improve accuracy was to extrapolate truncated projections, using either a smooth continuity of data or some prior information related to the shape of the object, by performing a computer-generated image.

However, discontinuities in derived projections can still produce artifacts characteristic of this type of synthesis. This approach can not be correctly applied, according to these authors, to clinical situations. However, a limited number of complete projections can be taken in addition to truncated projections to improve the synthetic reconstruction of the patient's anatomy.

<Desc / Clms Page number 75>

In CT, the algebraic reconstruction technique (ART = Algebraic Technical Reconstruction) can provide a better reconstruction than the conventional method of convolution of rear projection (CBP = Convolution backprojection), either from very few projections or from incomplete projections. . The disadvantage of ART is its low computation speed (which also affects the accuracy of convergence), caused by the high correlation between consecutive projections. The correlation can be reduced if the two orthogonal projections are consecutively updated. For this purpose, a new multi-level algebraic reconstruction (MLS-ART) scheme has been developed to substantially improve the computational efficiency and reconstruction accuracy of the ART algorithm (17). When MLS is used, the maximum orthogonality of projections, during reconstruction, can be maintained for ART. The MLS command, although it works best when the number is a power of two (while the MLS is the exact authorizer of the fast one-way Fourier transform), applies to any number of projections. Recently, the MLS-ART algorithm could do as well as with the adaptive gantry CT and a limited number of complete projections plus truncated projections, since it has gained an additional advantage, eg. eg, flexibility. We can simply put these pixels, which are available along the ray paths, up to date from projections. keep it truncated and leave out other image pixels. This is not a trivial task, as is already known, for the conventional conical projection technique (CBP).

The first step, in this direction of the use of the CT in Radiotherapy, was in fact made by James Winter with its so-called masking device made in material. wherein the radiant energy passing therethrough has an aperture adapted to pass an attenuated portion of the radiant energy beam for use with a 1st and 2nd generation computer-assisted tomography that uses, for the purpose of imaging, a radiant energy beam for irradiating, such as in sequential tomotherapy, therapeutically a selected region with. said beam of radiant energy is interposed between the source and the target region in such a way that the radiant energy beam has been designed (EP 0 382 560
A1 of James Winter) to make, extemporaneously to therapeutic irradiation, diagnostic imaging. In this latter embodiment, a post-masking member is preferably aligned with the opening of the masking member and is made of a material that attenuates the radiant energy passing through.

<Desc / Clms Page number 76>

 at a level of energy substantially uniform to that of the radiant energy passing through the masking member. In this way, scanner detectors aligned with the radiant energy beam can continue to provide the imaging data during the therapy procedure. This would ensure continuous visualization of the positioning of the selected region during the radiation therapy procedure and avoid any positional changes which, whether due to patient movement or false initial patient positioning, may occur prior to the commencement of the procedure itself. This post-masking member is not required for treatment. As indicated in the above description when imaging is not desired, during the therapy procedure a thicker masking member may be used so that no radiation essentially penetrates the patient except through the opening of the masking member. This could be an option that can, when desired, be enabled or disabled. These scanners can thus treat at an approximate radiation dose of about 300 - 400 rads per hour, using a conventional rotating anode scan tube, including the cooling time of the tube. This is the equivalent of two diagnostic sequential CT scans per hour, but with all sections oriented on the same target lesion (eg 120 cuts / hours x 3 rad / cut = 360 rads / hour). The treatment times are long here and the aim does not take into account the physiological behavior of the organism in which the tumor is treated.

 2. 6.5. Three-dimensional analysis of real-time motion and patient alignment.

The article by Baroni et al. (2000) describes the technology and methods involved in a system to automatically check the position of patients on radiotherapy units. This is proposed in order to improve the accuracy of the implementation geometry of irradiation, which is, according to these authors, a crucial factor in the quality control of radiotherapy. The system is based on optoelectronic photogrammetry and contiguous rows in real time and detects multiple passive markers placed on selected skin markers of patients. The alignment of the patient and the interception of the positional emissions are performed by comparing the current three-dimensional positions of the markers with those of an initial position acquired during the simulation procedure and / or at the first irradiation session. reference. The system was used for repositioning the patient to measure the accuracy of central

<Desc / Clms Page number 77>

 conventional laser geometry. Inaccuracies due to breathing and random movements were also taken into account. This is why G. Baroni et al. (2000) asked the professional technicians (were asked) to carefully reposition three volunteers, using a traditional laser centering procedure. But, what is the accuracy of such measurements, when we know nothing of the fluctuation interval with respect to a target lesion moving in the body and not visualized.

The results of this method revealed significant repositioning errors, even under highly controlled conditions, particularly affecting the body region relatively far from the skin benchmarks used for laser alignment. The result of the repositioning of the patient and the automatic detection of any errors, caused by breathing or other unpredictable movements, is that at each session one faces the same problems. Feedback from the system in real-time feedback to the patient's position instantly provides operators with the appropriate visual cues and allows them to take countermeasures in the event of significant failures. suitable measures. In addition, the use of the system output for automatic position control has been considered by G. Baroni et al. (2000). This without taking into account the possible variability of the topography of the target lesion and the fluctuation interval in which its tissue environment and their respective surface contours evolve.

 2.6.5.1. Implementation of Therapeutic Irradiation Ideally, treatment plans should be evaluated against an expected range of organ movement and anticipated error during treatment implementation. The tools of such a planning have been developed to allow, in conformal radiotherapy, a concurrent animation of the organ movement and a radiobiological analysis of the target in three dimensions (3-D). The surfaces filled with the structures silhouetted on the CT examinations are projected in the pre-treatment images or in the megavoltage images, collected during the treatment of the patient. The visual simulation of the tumor as well as normal tissue movement is then performed by applying the refined three-dimensional transformations to the selected surface. The competing recording of the surface movement of the three-dimensional distribution (3D) of the dose allows the arbitrary calculation of modification of the dose to the volume. Realistic models of movement can be applied to the structure to stimulate

<Desc / Clms Page number 78>

 the inter-fraction movement and the installation error. The effective biological dose for the structure is, according to MacKay et al. (1999), calculated as a area that, for each fraction, moves over the evolution time of the treatment and is used to calculate the normal tissue complication probability (NTCP) or the control probability tumor (TCP = tumor control probability) for the moving structure. However, it is easier to evaluate, as proposed by the present invention, all these data every day in radioscanoscopy or radioscanography on board the radiotherapy system.

Vertical load 2.6.5.1.1. - Implementation and movements in three dimensions of space
To determine the distribution of placement errors for patients treated for pelvic malignancies with and without rigid partial immobilization devices. A study of 30 patients receiving pelvic irradiation with a four-field technique was conducted with a total of 524 portal views. Patients were divided into three cohorts of 10 patients. The first cohort is processed on a table. standard treatment without immobilizers (NI); the second and third cohorts are treated with a custom-built, commercially available compression device to be used to improve the accuracy of placement: an Alpha-Cradle (AM) mattress or an Orfit casting (OC). Deviations from the set up are analyzed in the X, Y, Z directions of the coordinate system. fixed, corresponding respectively to the latero-lateral, craniocaudal, and antero-posterior directions; but not in the sense of a rotation. Rotational movements are simply not evaluated.

Considering as a measure of installation accuracy the percentage less than or equal to 5 mm of disagreement between simulation snapshots and movies. gantry cranes, immobilizers appear to increase this accuracy by:
88.5% (X), 79% (Y), and 100% (Z) with AM; 84% (XY), 97.5% (Z) with OC and only 76.5% (X), 40% (Y), and 65.5% (Z)% for NI. The distribution of systematic errors, defined as mean movements during treatment, of the three cohorts of patients had a mean and a standard deviation of. 0.7 2.7 mm in the X axis, -5.5 2.6 mm in the Y axis and -0.9 2.2 mm in the Z axis, when no 0.8 1.7 mm, 2.7 mm and 0.3 0.4 mm are added for the Alpha-Candie group; 0.3 1.4 mm, and 0.5 0.6 mm for the Orfit casting group. The random distribution of about average errors approached a normal distribution and the standard deviations were 4.4 mm (X), 4.2 mm (Y) and 4.8 mm, respectively

<Desc / Clms Page number 79>

 (Z) for the NI; 3.3; and 2.5 mm for AM; and 3.4; and 2.7 mm for OC. For these authors (C. Mitine et al., 1999), the two immobilization devices improve the reproducibility of a given pelvic field, but there is a small comparative advantage in relation to the cost and space of the devices. .

 But, these devices are not always made to measure and in optimal position of use.

2.6.5.2. - Placement, immobilization and rotation of the patient on oneself Evaluate the impact of a remodeled immobilization system according to the demands of a. buyer (modified to measure) on the accuracy of field placement, simulation and treatment delivery time, radiographer convenience and patient acceptability. Thirty people receiving radical radiotherapy for prostate cancer were randomized, in another study, using a crossover model, to have radiotherapy planning and treatment. given either in a conventional treatment position (CTP = conventional treatment position) or using an immobilization system (IMS). The randomization done by CM Nutting et al. (2000) consisted in receiving, during the initial three weeks of radiotherapy, either CTP or F IMS, from which patients were rescheduled and changed for treatment. alternative. Treatment accuracy was measured using a portal electronic imaging device (EPID). Radiographers and patients completed weekly questionnaires. While the average simulation time was 22.5 min (ranging from 20-30 min) in the CTP and 25 min (ranging from 15-40 min) for the IMS (p <0.001). The average treatment time was 9 minutes for. CTP (ranging from 8-10 min) and 10 min (ranging from 8.5-13.5 min) for IMS (p <
0.001). The mean simulated displacement from the PTC isocenter compared to that of 2.0 mm IMS (p = 0.07) compared with the isocentre of the previous fields was 1.7 mm. They were, for the values of the left lateral fields, 1.8 and 1.8 mm respectively (p = 0.98), and for the right lateral fields of 2.1 and 1.7 mm (p = 0). 06). No clinically significant reductions in systematic placement errors or random fields have otherwise been demonstrated. Radiographs reported that patients found IMS more comfortable than PTC (P-positioning) (p <0.001) and alignment with skin tattoos (p <0.001).

Although the IMS was more comfortable, the accuracy of the treatment compared to
CTP, was not improved in this department; In addition, the treatment and implementation

<Desc / Clms Page number 80>

 place of the patient took, according to C.M. Nutting et al. (2000), more time and were more difficult. Then you had to monitor the smooth running of the operations.

2.6.5.3. - Implementation Monitoring System and Electronic Imaging Dosimetry (EIS)
But, we must note here the importance, which is still not taken into account, the alignment factor, which characterizes the aiming accuracy of the object in radiotherapy ballistics. Hence the growing interest in the current art of the use of digital imaging portalised digital Radiotherapy, where gamma radiography is still the most used method, to date, to check, during an irradiation session external, the positioning of the patient. Electronic imaging systems (EIS) have been developed with the aim of having a real-time verification tool and anticipating, adapting and planning the demand for precision required for complex conformal radiotherapy treatments. The digital images of the irradiated field scroll during processing on a computer screen. The range of image-processing software can then be used to diversify and optimize contrasts or to apply filtration, to visualize images continuously and in a loop and to evaluate and / or quantify, in a statistically significant manner. , the patient's movements during the different treatment sessions. The experience gained by many teams would allow. in the short term, to consider the permanent replacement, in current art, of traditional radiography by electronic imaging systems (EIS), because (1) the quality of the electronic image, although perfectible, is practically always higher than that of gamma radiography; 2) retractable sensor systems, integrated into the stand of devices with automatic installation, make it possible to avoid. all the tedious manipulations of the installation of cassette holders and the development of photographs; 3) the quantification of the deviations always led to the improvement of the positioning, whatever the localization treated. But, against this, the image is morphometric rather than a diagnostic image.

 Electronic imaging is also considered to be an in vivo dosimetry tool in the art that needs to be exploited in clinical routine in today's art.

 Yet phenomena, of which it is clear that they have nothing in common, serve as a foundation for these models: for example, passive (2-D) and conjunction (3-D) operations are, as a transformation coordination 3-D, amalgamated (transfer of MRI images, CT, etc.). While issues of major importance are completely kept out of the discussions of the representation of reality in

<Desc / Clms Page number 81>

 its range of physiological fluctuation, voluntary and involuntary. It is also the case of the determinant of lexical elements corresponding to a given quality assurance quantization form, rather fixed of space and time, that we will discuss elsewhere, apart from a simple quest for good conscience.

2. 6.6. - Accurate location of the moving target during Radiotherapy
To describe a model and propose a tumor class treatment system that requires highly accurate target localization during external beam splitting therapy, the Phillips et al. uses image-guided techniques, in the vault of the linear accelerator, for localising the position of patients to be treated with stereotactic radiotherapy, conformal radiotherapy, and intensity-based radiotherapy techniques. modulated, for cranial tumors. The constraints of this model included, in this study by Philips et al., The flexibility of use of the processing software as well as the accuracy and precision of the repeated locations; time and human resource limits required the use of the system and facilitated its use. A stereotaxic radiotherapy system based on a system developed at the University of Florida, Gainsville, and sold commercially, has indeed been adapted for use at the Medical Center at the University of Washington. A pair of stereo cameras embedded in the vault of the linear accelerator (linac)). was used to detect the position and orientation of a series of skin markers attached to a piece of the patient block. The system has been modified either to allow the use of a treatment planning system designed for either stereostatic treatments or for the use of a general three-dimensional radiotherapy planning program. Measurements of accuracy and accuracy of target location, dose delivery, and patient positioning were made using a number of different calibers and devices. Procedures have been developed for safe and accurate clinical use of the system. The accuracy of the target's location is comparable to that of other art processing planning systems. The decline of the). gantry, which can not be improved, was measured at 1.7 mm, which was the effect of broadening the dose distribution, as confirmed by the comparison between measurement and calculation. The accuracy of the positioning of a target point in the radiotherapy field was 1.0 0.2 mm. The room calibration procedure using laser bases had an accuracy of 0.76 mm, and it was 0.73 mm using a ground-based radiosurgery system. The location error of

<Desc / Clms Page number 82>

 the target in a static phantom was 0.64 0.77 mm. Using this system, positioning errors due to bed rotation error were reduced. This, of course, does not take into account the actual range of position fluctuation, under the influence of the various voluntary and involuntary movements, of the target unvisited lesion.

 More importantly for our purposes is the emphasis of the Phillips et al. (2000) self-checking mechanisms used by a theoretical body of contemporary art dominated by the easy recourse to mathematical formalism to rectify. The system described for clinical purposes would have an accuracy and a pre-. for which it had been conceived. It would even be robust enough to detect errors, and only requires a nominal increase in time and effort to set up. Future work should, for these authors, focus on evaluation for use in the treatment of non-enclosed cancers within the cranial vault of the head and neck, and therefore more sensitive to its development. . compatibility with transmitted movements such as breathing. Microbeam therapy is also established in the Orion et al.

(2000), as a general concept of treatment of brain tumors. An X-ray source, based on the synchrotron, was chosen for experimental research in microbeam therapy, and new simulations were essential to explore, for an appropriate description of synchrotron radiation characteristics, the therapeutic parameters. To design the therapeutic parameters of tumor treatments, the LSCAT kit (Low energy
SCATtering) recently outperformed by Monte Carlo's EGS4 simulation code has been adapted to develop an exact user code written for the circumstance and des-. designed to calculate radiation dose profiles for microbeams with an accuracy of 1 micron. Because of its ability to simulate the transport of low-energy rays with detailed photon interactions (including the functions of incoherent diffusion of binding electrons, and polarized linear coherent scattering), LSCAT is highly suitable for this purpose. The properties of the synchrotron X-ray microbeams, including the polarization, the source spectrum and the darkness of the beam, were simulated by the new user codes. Two concentric spheres, including an inner sphere defined as the brain, and a surrounding sphere defined as the skull, represented the ghost. Microbeam simulation was tested using a beam of. 3 x 3 cm network for smaller treatment areas and a range of 6 x 6

<Desc / Clms Page number 83>

 cm for larger ones with different therapeutic parameters, such as beam and spacing. The results showed that the range of microbeams maintained an adequate peak-to-valley ratio at tissue-adapted depths for radiotherapy of at least five-fold. One-micron resolution 'MOSFET' dose measurements validated, according to Orion et al. (2000) the basics of the user code for microplanar radiotherapy.

 2. 7. Dosimetric optimization and total dose distribution Actual optimization can not be achieved outside the clinic; but current art proceeds remotely by a virtual simulation of the clinical act of radiotherapy, especially with the technical concept of IMRT. In order to avoid local recurrence, curative radiotherapy aims to eradicate all clonogenic cells within a target volume. This must, according to the principle of IMRT, be done without severely harming adjacent organs. Modulated intensity radiation therapy (IMRT) performed with a single external beam provides the potential to fulfill two of these requirements simultaneously. Because of the large number of parameters, such as the fluence of individual brush beams, which must be taken into account in conventional interactive optimization techniques, this is only possible in theory. This is what led to the technical concept of inverse treatment planning to aim just the action of the treatment must be described using certain objectives of the treatment, so that a mathematical optimization algorithm can find the parameters that drive of its dimensions and to an optimal distribution of the dose in a target volume which has been prejudged of its topography as a function of time. For each brush or elementary beam, the primary fluence, the position of the source, the direction of the beam in space, the energy of the beam, and the radiation modality can for example be treated as variables. In practice, the solution space is, however, usually restricted in the old art, so that only the weight of the primary fluence varies and the objective remains fixed. This, when compared to more conventional treatment techniques, is judged to result in a more therapeutic gain. Yet permanent adjustment is considered important for optimizing beam directions for only a small number of windows. Different mathematical methods were used to achieve the said

<Desc / Clms Page number 84>

optimization. If many assumptions and parameters are not well justified, it is better to use more simplistic models that require a smaller number of parameters in which authors, such as Sauer et al., Have more confidence. In addition, simple models result in a simpler and faster numerical computation and they are able to converge easily to a global optimum. Various approaches have already been used, such as, for example, maximizing the minimum dose to the target, minimizing the maximum dose to the at-risk region, the least square, or equivalent procedure for achieving a prescribed dose distribution. total energy outsourced to the target and the volumes at risk. All. these methods generally result in dose distributions that are considered to be higher than those of conventional dose distribution plans. However, the end point does not necessarily correspond to the goal for an individual patient of the treatment. In some methods it is necessary to use factors of importance without direct correspondence with the biological parameters. Others . are due to the possibility of using stringent dose constraints which, according to
Sauer et al. (1999), lead to the non-feasibility of optimization.

 Sauer et al. have indeed explored in their work a strategy where the points of the dose in the target are forced to be between the lowest and the highest limit. With the lowest limit, the probability of tumor control is adjustable.

. Different links can, if necessary, be assigned to different points of the target.

 The class value of any point is thus minimized, within a risk region. With this strategy, a weighting between the tumor control and the probability of complication of the normal tissue is thus avoided. If only one type of risk region is taken into account, none of the parameters whose responses for. specify are not usually necessary. Two distinct groups: stochastic methods and deterministic methods, such as linear or quadratic programming, are used by most teams working on this topic. Although deterministic methods are faster, there is a risk that algorithm-dependent values may only converge. to a local minimum. However, Deasy was able to show that non-convex objective functions only occurred if dose-volume constraints were explicitly taken into account.

 The most important task is the definition, through the construction of an objective function, of a sensible goal of treatment. It is obvious that this function must include radiobiological considerations of the irradiated tissue. A curve

<Desc / Clms Page number 85>

 Dose response has been suggested by many authors, beginning with Holthussen, in 1936. Brahme et al. have, in order to model the sigmoid response of the dose, applied the Poisson statistics. The tissue can be characterized by the D50 value, a dose resulting in a 50% probability of the dose effect and a gradient of the dose response at the D50 level. A method that takes into account the inhomogeneities of the dose distribution must also be applied. Kallman et al. have suggested on the other hand to define a relative seriality or its opposite, the relative parallelity of an organ. A serial organ is injured if a subvolume of that organ is injured, while a fully paralleled organ loses its full functionality if only all subvolumes are damaged. The complex physiology of living organs, however, makes it difficult to adequately model all possible effects of radiation, and clinical data are scarce and unsafe (exposed to danger) at the same time as biased. Another problem is the need to average at D50 between different patients with different radiosensitivities. The applicability of the two functions in steep slope of theoretical response and the flattened dose reached on average by the irradiations of this organ for an individual patient is equivocal. A dose as low as possible is automatically performed. The value of tolerance where the probability of complication that certainly begins to arise must be known by the clinician, in order to recognize, in current art, the plans where no curative treatment is possible. This case is demonstrated for a U-shaped target, similar to a target volume of the head and neck. For a parallel-type member, it would be desirable to apply, in order to maximize the low dose region, a seriality parameter. Alternatively, a simple relation of the dose response can be used. This is proportional to the dose for low dose values and almost flattened beyond a specified threshold dose. If more than one type of vulnerable regions are implied, relative sensitivity and seriality parameters can be applied.

Alternatively the constraints of the upper limit can be assigned to each region. This is demonstrated, for an L-shaped target surrounded by vulnerable regions that correspond to the lung and spinal cord, by Sauer et al. (1999). With these sophisticated approaches, optimization algorithms can be used together with the clinically known parameters of the tissue.

Thus multiple local minima exist in almost every problem of coplanar or non-coplanar planning. This is why X. Wu and Zhu used a global optimization method based on tomographic information.

<Desc / Clms Page number 86>

 the distribution of local minima to discover all the local minima and to select the one that is better than the global minimum. They randomly selected N points and built a topographic graph, from which initial M points (M # N) are selected to start a local search. Since each point grain is at a local minimum or near minimum, the solutions found by local search can be used as final optimization results. They verified this algorithm with those obtained by a local optimization method (sequential quadratic programming). The results show that this algorithm is feasible and efficient (X. Wu et al, 2001). Treatment planning is, according to. S. Webb (1993, 1997), a very important part of the radiation treatment system. Many researchers are devoting their efforts to this area, and many optimization methods have been used to solve the problem of dose distribution. On the basis of the functions used of the cost of Radiotherapy treatment planning, they were able to classify these methods. optimization in one of two types: linear or non-linear. The simplex planning advocated by Rosen et al. (1991) is, for example, a most frequently used linear programming method. Nonlinear methods can not be further subdivided, according to cost, into quadratic programming (Holmes and 2.5.1 Mackie, 1994) and other methods. non-convex programming (S. Weeb 1993, 1997).

On the other hand, methods can be classified, based on the behavior of research in two types: deterministic (Nocedal, 1999) or stochastic (Kan
Rinnoy and Timmer, 1984). Deterministic methods usually use gradient information and are very popular, since they are relatively simple. They are easy to use and work better and more importantly in most situations. Stochastic methods usually start from a point or multiple points in the search region drawn at random. Test points are generated, according to a certain distribution also drawn at random (usually a uniform distribution) on the region of. search, and a better point is usually accepted as ranking higher than other points. Some operations are used to allow the algorithm to escape a local minimum. The methods belonging to this class share the following disadvantages: (a) - The risk of a local minimum, with no way of knowing if the optimal solution has been found;

<Desc / Clms Page number 87>

 (b) - uncertainty about when to complete the algorithm; (c) - The long and expensive execution time.

Since in addition the genetic algorithm must maintain a large population of solutions (chromosomes), memory consumption is high, and generating an initial population is a process that also lags in length. A third classification scheme of optimization methods is based on the nature of the solutions to the problem and classifies these methods as either local (Nocedal, 1999) or global optimizations (Floudas and Pardalos, 1992). The more deterministic methods are those of local optimization. Although stochastic methods, such as techniques covering the whole (Floudas and Pardalos, 1992), random search methods (controlled or pure direct) (Price, 1983), methods based on multiple local searches (multistart (Zielinski 1981, Boender et al., 1982 and Storey 1994), and many others also belong to this class. These global methods are rarely used in optimizing planning, according to X. Wu and Zhu (2001), of Radiotherapy treatment.

2.7.1. - Calculation of Local Optima Dose For a convex problem, we can take any starting point and do a local search from that point, and we are sure to get the best absolute answer. Unfortunately most optimization problems are non-convex and there are a lot of local optima. Thus, different starting points can, for local optimization methods, result in quite different final solutions. To find a good starting point, most researchers recognized the problem and some even recommended making several attempts (trying several times, for example) to find such good starting points. X. Wu and Zhu do not believe that local optimization methods can be abandoned, but they can not, for reasons of computational efficiency, be able to find the same local minimum again and again. Lessons from earlier approaches have taught them that the coplanar orientation problem solving space is composed of a large number of local optima (Haas, 1999). We believe that almost all three-dimensional radiotherapy (3-D) treatment planning problems (including coplanar and non-coplanar problems) are at least multiple problems, except for some particularly simple and unimodal problems. Thus, the usual methods descending from the usual gradients are not well adapted to this type of problem. The authors believe, according to the classification of a given problem of Torn

<Desc / Clms Page number 88>

 et al. (1999), that most problems with Radiotherapy treatment planning are either easy or moderately difficult and that, in small problems at several hundred local minima and in moderately sized or large problems, there are many minima. local. Global optimization is well adapted, for Wu and Zhu (2001), to solve these problems, but we remain convinced as for us that the best optimization is that which removes not only any methodological bias of reasoning but also any technical and technological bias. .

Nocedal presents, for its part, a global method of a single multilevel topographic link, the so-called TMSL (= topographical multilevel singlelinkage) method, which is based on an agglomerate of pre-sampled points from the search region. The purpose of Nocedal is twofold: to use topographic information on the distribution of local minima to select, as a final solution for coplanar or non-coplanar problems, the global minimum and provide multiple solutions to the selection of radiotherapy treatment. He compares this method with the local optimization method (sequential quadratic programming, SQP = Sequential quadratic programming).

The key to calculating the photon fluence distributions of Modulated Intensity Radiation Therapy (IMRT) is the use of an appropriate objective function. The objective function should reflect the clinical objectives of tumor control and low probability of side effects. Individual radiobiological parameters for the patient's organs are not yet available today with sufficient accuracy. Some major drawbacks of some common optimization methods include a failure for arbitrary input parameters and / or a need, as a final solution for coplanar or non-coplanar problems, to converge on a solution to guide the user. optimization, intensive user input. In the work of Wu and Zhu (2001), for example, a forced optimization method closely related to the clinical objectives requested, avoiding the disadvantages mentioned above, was implemented and tested. In a prototype treatment planning at IMRT, tumor control was guaranteed by these authors by setting up a lower limit for the target dose. So that the goal of a low complication is met by minimizing the dose to organs at risk. If only one type of tissue is involved, there is no absolute necessity for radiobiological parameters. For the different organs, the threshold dose or the relative seriality of the organs or the

<Desc / Clms Page number 89>

 upper limit of dose could be placed. All parameters, however, were optimal, and could be omitted. Dose-Volume constraints were not used, avoiding, in objective function, the possibility of local minima.

 The approach was identified through the simulation of the entire head and neck and a pulmonary case. A cylindrical phantom with dose distribution of the precomputed individual brush bundles was used. The dose to risk regions could be significantly produced for only minor additional gain. The relative sterility of the organs was moderated by the use of exponent added to the dose. This approach, however, increased significantly. the calculation time. The alternative of placing an upper limit is much faster and allows direct control of the maximum dose. Forced optimization would ensure a high probability of tumor control, it would be calculated more efficiently than adding disability terms to the objective function, and the input parameters are known dose limits in clinical practice (Wu and Zhu, 2001).

2. 7.2. - Quality control, verification and documentation of the delivery of the dose
Detailed quality control (QC) protocols are now a necessity for modern radiotherapy services. QC protocols established for. Treatment planning systems (TPS) do not include recommendations on three-dimensional processing characteristics (3-D), such as volume dose histograms (DVH), highlighted. Penitsa et al. (1998) developed a test protocol on the characteristics of DVH. This protocol evaluates the computational compatibility of DVH. with the dose distribution calculated by the same TPS, by comparing the DVH parameters with the values obtained by the isodose distributions. Calculation parameters (such as the dimension of the calculation grid) applied during the tests at
GSTs are not fixed but placed, as if the test represented a typical clinical case, by the user. Six commercial TPSs were examined with this protocol,. by these authors in a framework structure of the EC project Dynarad (Biomed I).

Intercomparison results prove compliance of most results
DVH of TPS examined with isodose values. However, special attention should be drawn, according to Penitsa et al. (1998), when working under unfavorable conditions, such as in regions of high dose gradients.

 In these cases, higher errors are especially derived when a number

<Desc / Clms Page number 90>

 insufficient dose calculation points are used for the calculation of DVH (Penitsa et al, 1998). Hence the statistical problem of controlling the accuracy of dose delivery, especially in the presence of ballistic aiming biases.

2.7.2.1. - Control of the accuracy of dose delivery
Art Radiotherapy (TPS) treatment planning systems are indeed evolving rapidly towards powerful systems of graphic features and advanced calculation formalization that enable the realization of three-dimensional treatment plans (3-D) However, it does not take into account the relative mobility of the objective and its variability as a function of time. The programs. Dose Volume Histograms (DVH) are indeed being incorporated into TPS 3-D as standard plan evaluation tools. DVHs have the ability to visualize in a single graph information about doses (for example, tumor, sensitive organs), which does not integrate the temporal dimension in its reasoning. This feature would facilitate, according to Drzymala et al. (1991), the. determination of the characteristics of the plan, such as the level of the dose, the uniformity of the dose or the undesirable hot or cold spots. DVHs are, says Webb (1997) cited by Penitsa et al. (1998), also used as a basis for calculating biological parameters (notably the probability of tumor control [TCP] and the probability of complication of normal tissue [NTCP], thanks to the techniques of reduction of histograms). Although DVHs have been recognized as standard tools in the 3-D plan evaluation procedure, their performance has not been evaluated, particularly over time, in a comprehensive manner. Inaccuracies inherent in inappropriate algorithms and implementations of the calculation parameters may also introduce errors. in the treatment procedure. The need for comprehensive QC procedures, which could ensure the statistically significant acceptance of DVH programs for treatment needs, is evident. Although various QC protocols, generally for radiotherapy procedures (Kutcher et al., 1994) and TPS (Van Dyk et al., 1993; Full 3-D TPS QC res have not yet been fully established. It is especially the most advanced features such as DVH which, according to
Penitsa et al. (1998), lack documentation on its QC procedures.

 This is why Penitsa et al. have attempted to set up a series of QC tests for DVH computation programs. The tests presented were performed in one. concerted 3-year action (i.e. 1994-7) of the 32 institutions bringing together 14 countries

<Desc / Clms Page number 91>

 in the framework of the EC project Dynarad (Biomed 1 Program). The Dynarad (Development and Standardization of New Dynamic Radiotherapy Techniques) has, among other things, studied the quality control procedures of 3-D treatment planning systems (Rosenwald, 1995a, b). GSTs were evaluated according to a qualitative test protocol that evaluated the dose calculation features, such as electronic contamination, block correction, heterogeneity correction, lack of diffusion material, combination of heterogeneity and blocks (Dynarad 1996, Panitsa et al, 1997). Two complementary protocols to the above test, further evaluating non-coplanar bundles and DVH calculations have been developed (Dynarad 1997, Panitsa et al, 1996).

Thus this QC protocol evaluating all these features was developed and circulated within the institutions participating in the EC project Dynarad, whose DVH test protocol can be a part of a complete QC process evaluating the performance of TPS 3-D. The application of such QC procedures would be crucial for the overall quality of the treatment. The proposed test protocol would evaluate the compatibility of DVH calculations with the dose distribution produced by the same TPS without ballistic accuracy assessments. This is done by comparing the results of the DVH with the values obtained from the theoretical distributions of isodoses. Therefore, the DVH test protocol does not require any other measured data than that of the user's beam data library and not even updated anatomical data. It is therefore a protocol easily followed by any center wishing to evaluate the performance and the limitations of its DVH calculation program but whose accuracy of ballistic sighting would not be statistically significant. Standard procedures have been proposed to make it easier for the user to apply the tests and to allow retrieval of results when the protocol is used for intercomparisons of TPS. The application of the protocol to six commercial GSTs has otherwise proved, according to Penitsa et al. (1998), that the usability of the protocol's user is given by the opportunity to evaluate the performance and limitations of some of the most common GSTs in Europe.

The results acquired with the test protocol of the DVHs could be viewed with inaccuracy criteria accepted in DVH calculations, which would meet the criteria of the calculations of treatment planning, are even less than 5%. ICRU Report 42 (ICRU, 1987) recommends a

<Desc / Clms Page number 92>

 uncertainty less than 2% (or less than 2 mm for high dose gradient regions) in GST calculations. Other authors, such as Brahme et al.

(1988), accept higher uncertainties. For example, they accept 3% while Van Dyk et al. (1993) recommend a series of acceptability criteria depending on the situation under consideration (eg, 2% to 4% for 2-D photon beam calculations). It is also necessary to take into account the uncertainties related to the results of the measurements. The main restriction lies, according to Penitsa et al. (1998), in the accuracy with which the isodose distributions are computed, to which is added that with respect to the actual topography of the lesion. In most GST, the dose calculation characteristics for the DVH calculation differ from those used during the calculation of the isodose distribution that are high enough to allow extraction of the results. There are also uncertainties in the measurement procedures that influence the accuracy of the study. Uncertainties in the determination of values measured from DVH graphs and from the isodose map are generally considered relatively small, in the Penitsa et al.

(1998), to be considered negligible in relation to the order of magnitude of the results.

The performance of the DVH algorithms depends on the computational parameters, such as the sampling technique and the number of calculation points per volume (Drzymala et al., 1991 and Goitein, 1990 and Chin, 1993). It is well known that a high number of dose calculation points may keep the variation in DVH calculation low. It has been shown that the number of computation points is an important parameter for the variation of the high dose gradient regions. The user, who has adopted a high number of computation points (eg, Centers C, D, E), found results better than those who had used a small number of computation points (eg Centers A, B, F). On the other hand, the number of computation points does not seem to affect the variation in the low dose gradient regions. In this case, the sampling technique, as we shall see, seems to be of great statistical importance (Penitsa et al, 1998).

 In this study, the GSTs examined can not really be compared to others since the different calculation parameters (computerization) have been applied for each of them. The results of a careful examination are only indicative of possible errors that apply, in clinical conditions, to the calculation of DVH. In general, the GST examined (performed) in the Penitsa et al. Study performed in agreement with the results of the isodose distribution. of the

<Desc / Clms Page number 93>

 Significant discrepancies were observed only for the cases of the Bl (high dose gradient) test. These variations mean poor performance in high dose gradient regions, when DVHs are calculated for high dose gradient regions, either for investigation of the plan or for determination of other parameters (such as TCP and the NTCP); why the use of a large number of calculation points is recommended, as it could significantly improve accuracy. The proposed QC protocol can be used to evaluate the validity of the calculation fault parameters. It is also necessary to have, for Penitsa et al. (1998), sufficient documentation on cara. features and limitations of DVH programs, and to follow a comprehensive quality assurance process for GST.

2.7.2.2. - Patient placement control and ballistic sighting accuracy
2.7.2.2.1. - Iterative implementation of the target volume. If the lesion can not be identified, during the radiation treatment and especially its ballistic aim of different incidences of therapeutic radiation beams, with their own topographical variances additional uncertainties come from the daily errors of setting up and movements of organs that can, according to Lujan et al. (1999), lead to. differences between the dose distribution, on a single treatment plan, and the actual distribution of the dose delivered to the patient. Two primary approaches are proposed in the literature to take into account these uncertainties. The traditional approach that measures or estimates the magnitude of uncertainty of establishment as well as organ mobility and adds positive margins around a cli- target volume. (CTV) to form a forecast planning target volume (PTV). The dose is calculated on a static patient model and prescribed for PTV, with the intention that the actual dose delivered to the CTV is equivalent to the predicted dose distribution. This margin expansion (anticonformational) approach does not take into account the differences between the predicted dose distribution and the. distribution of the actual dose delivered for normal tissues close to CTV. The second approach includes error margins and directly incorporates uncertainties into dose calculations, thereby providing a more complete and accurate prediction of the dose distribution delivered to the two tissue types together, tumor volume and tissues. normal.

 Random errors of setting up, using a static model of the

<Desc / Clms Page number 94>

 distribution of the calculated dose for a patient by computed tomography (CT), can thus lead to a false prediction. Multiple recalculation of dose distribution, covering the expected range of patient positions, provides a way to estimate the course of treatment. However, due to the statistical nature of the implementation uncertainties, many of the treatment runs must be simulated to calculate a distribution of the values of the average dose delivered to a patient. Thus, direct simulation methods may be time consuming and may be impractical for routine clinical treatment planning applications. Other methods have been proposed to compute efficiently, via a dose distribution convolution (computed on a static CT model) and a probability distribution function (usually Gaussian), which describes the nature of the uncertainty and the distribution of the values of the average dose. Lujan et al. extend the computation based on this convolution to compute the standard deviation #D (x, y, z) of potential outcomes around. the distribution of the values of the average dose, and they characterize the statistical significance of this quantity, using the central limit theorem. The standard deviation also provides the confidence limits of the dose distribution, and these can be used to evaluate the stability of the treatment plan.

 The methods based on a convolution of the static distribution of the dose with. a function (usually Gaussian) representing the random distribution of uncertainties arising from the placement and movement of organs has been proposed specifically for pelvic sites (Leong, 1987, Kutcher et al, 1991, Chui et al, 1992; et al., 1993, et al., 1996). Lujan et al. described, already in 1990, a method based on a convolution to incorporate in calculations 3-. D of the dose the uncertainties resulting, because of the respiration, of the movement intra-treatment of the organs (Lujan et al, 1990). They then generalized this method, with a view to Radiotherapy, to incorporate the uncertainties due to daily errors of placement in the 3-D calculations of the dose. In doing so, they quantitatively describe the potential differences between the predictions based. on the actual distribution convolution of the delivered dose in a finite fractional spread of the treatment. The validity of this approach is confirmed, via comparisons with direct simulations, for the treatment of liver tumors.

 Thus, the treatment effects of these uncertainties are retrospectively analyzed and the dose requirements based on a protocol used for the treatment. hepatic pathology treatment at the University of Michigan (Lawrence et al,

<Desc / Clms Page number 95>

 1990 ; Lawrence et al, 1992; et al., 1998).

The basic algorithm, to convolve the uncertainties of placement with static dose distribution, has been previously described (Leong, 1987, et al, 1991, Chui et al, 1992, et al, 1993, et al. , 1996). The method of convolution affects a rigid movement of the body; without modification in the external contour of the patient, and without any deformation of organs. The study by Lujan et al., Considers random translational uncertainties, shear forces along the anteroposterior (AP), left-right, and superferior axes of placement. Based on a retrospective analysis of Schewe et al. (1996) of the placement of the patient, Lujan et al. (1999) postulated that the different translations along these primary axes are independent and that the nature of random translational uncertainties of implementation can be arbitrarily characterized by Gaussian probability distribution functions.

The mean implementation uncertainty is assumed to be zero for all three axes, and they used the standard translational derivations of the Schewe study (#LR = 7.4 mm; #AP = 4, 9 mm, #SI = 5.3 mm).

 2.7.2.2.2. - Uncertainties and importance of the marginal costs of acts of radio therapy Today, the climate of uncertainty is mainly created, in a context of budget reduction, by the marginal costs of development of radiotherapy devices and the spectacular breakthrough of numerous procedures. the need to compare not only the results of these treatment regimens with each other, but also to confront them with the quality assurance issues that arise in radiotherapy and which seem, if Judging by the emergence of more than ever items on the agenda devoted to new clinical concepts. New methods of high-precision radiation therapy, such as conformal radiation therapy, modulated intensity radiation therapy, tomographic radiotherapy or tomotherapy, for example, are credited with the potential to be able to distribute radiation at any level of radiation. accuracy, tumors and normal tissues. If such methods are justified, in the light of biological and clinical uncertainty, the topographic anatomy and the physiopathology of the tumor make them more questionable.

Doubtful are the precision methods that require more time and staff than can be expected in a clinic that runs constantly with small staff or the minimum service requirements. These uncertainties

<Desc / Clms Page number 96>

 are, moreover, related to: - the technical results of patient immobilization and the taking into account of uncontrolled movement of the organs; - the inaccuracies of historical imaging in the delayed determination of the exact tumor volume; - the length of the planning period and the timing of dose delivery; - the inhomogeneities of the dose and the exact interpretation of the dose-volume histograms (DVH); - quality assurance procedures; verification and documentation of beam delivery; - biological unknowns including carcinogenesis, and tumor kinetics; and finally, - purely economic considerations.

. It would already be easier to suppress the large part of uncertainties instead of spending the majority of one's time trying at all costs to ensure their innocence by integrating them into complicated formalisms of the process of calculating the dose. In doing so, we are already eliminating the large part related to the errors accumulated in different smoothing operations and their conjunction at the level of the overall result, making in. same time saving time. Moreover, we should not worry about the famous safety margin of what is overflowing at the level of the target, since there is still in the ballistics of the irrational area a zone of flight and penumbra corresponding to this need of marginal marginal irradiation; margin in which, as we know, the number of. Tumor cells present would not necessarily require that one uses large means to kill a fly, not to mention that the oxygenation passes by there to maintain long during spreading the radiosensitivity of the tumor.

Otherwise, we could also, with regard to the final goal of the operation, the target volume, act on new requirements. From the perspective of the budgets. or decreasing, the study by P. Dunscombe et al. (1999) was not undertaken to explore the economic effects of changes in selected operational parameters within a radiation treatment program using Northeastern Ontario Regional Cancer data.
Center and a model to staff recognized. A commercial prospectus. was used to calculate the cost of spreading the 18 radiotherapy fractions,

<Desc / Clms Page number 97>

 including all major preparatory processes such as simulation and treatment planning. Using this prospectus and on the basis of explicit and reasonable assumptions, such as the cost of radiation therapy, the size of the facility (eg, equipment complement) as well as the hours of various operations, have been calculated. Based on the assumptions used, the cost of radiation therapy in a facility treating less than approximately 1,600 patients per year starts to rise.

 At 400 patients per year, spreading costs approximately 50% more than an installation of 1,600 patients per year. Long hours of operation do not seem to generate significant savings if it is none, when assumptions. Realistic about the life of the machine and overtime pay are made. Using a flyer to simulate radiation treatment program changes may, according to P. Dunscombe et al., Be a tool for making an important decision, when the effects of changes in operating parameters can be demonstrated.

. It is also known with the notions of quality and quality management that quality is the ability of a product or service to meet the needs of users. or that, according to ISO 8402, all the characteristics of an entity that confer on it the ability to satisfy expressed and implicit needs, conformational radiotherapy is therefore felt today. now as a need. But, quality is also how to satisfy the customer?
How to ensure that the product, service or care meets the customer's expectations? The minimum expectation of a chronic patient is naturally to spend less time on the site of his care and that of an institution would be the output of a site: the number of patients per machine and per year would be in this. meaning a good indicator of a certain profitability without additional workload for the staff. And all this to do better at the lowest cost. Given the dramatic increase in health care costs over the past decades (and in a global health care reimbursement system), the concept of cost-effectiveness is one of the decisive elements in the political choice of one. given strategy. In addition to health, the concepts of efficiency, effectiveness, safety and risk management must be added. It is true that quality assurance imposes a rigorous organization and methodology to make a priori proof that the product, service or act of imaging or irradiation will conform to its specification. This is, in other words, all actions. put in place in order to have confidence in obtaining quality internally

<Desc / Clms Page number 98>

 (within the service) and externally. The perspective and time horizon of the present invention should therefore be clearly specified in the description below. Cost effectiveness is a synthetic measure requiring a microeconomic approach, to determine actual costs (numerator), and a measure of effectiveness (utility), taking into account patient or community preferences (denominator), to allow comparisons between interventions of different natures.

The background to the realization of this invention is therefore to make a counter-fire to a design of simulation of the treatment that uses artifices inexorably making the sum of the methodological errors, through the notion of isolated virtual simulation, based on a delay, and seeks to account in theory only for a very precise definition of the target volume. But the rest of the practice does not gain either in accuracy or in ballistic precision; This is because the aim continues to be blind and remote from the virtual simulation and the CT scan of treatment planning, in conditions that are far from those of the therapeutic clinic. The most important thing for ballistics is, regardless of the gateway or firing angle of the radiating particles and their respective positional variances, to always be on the ball. Having exactly the same entrance door marked to the skin does not prevent the imperceptible movement, such as rotation on oneself, modifications under the influence of voluntary or involuntary movements, topography of the lesion as well as even imperceptible translations. in the three axes of space no precision, serves, strictly speaking, ballistic aiming. This is all the more true when, as in stereoradio- therapy, two synchronous beams of therapeutic irradiation are used whose precise objective is none other than the electromagnetic interaction which must occur exactly within a target volume, the only object of interest evaluated daily by the statistical method, with each movement with its exact amplitude of variation not only closer to the physiological movement but also at each moment and at each fraction of treatment, all at along the sprawl.

2. 7.3. - Principle of stereoradiotherapy The decisive step, from this point of view, has been taken in the use of several simultaneous beams of radiation in therapy by the Strasbourg School of Radiology (ERS, Rector: Dr. Christophe Mwanza Chabunda). In fact, in terms of major events in CT-guided dynamic image radiotherapy, in this beginning

<Desc / Clms Page number 99>

 In the 21st century, nothing can compete with the launch of this true concentrate of what is best, this really new technology of our invention comes in several simultaneous steps involving megavoltage irradiation, instantaneous and permanent kilovoltage imaging and real-time dosimetry with constant tracking of the target volume on a CT display screen associated with instantaneous control and optimization of said therapeutic irradiation. These procedures make it possible to instantly identify any problem relating to the placement of the different beams together with the optimal targeting conditions of the precisely defined target in the patient's body and lead, if necessary, immediately and without loss of time at the subsequent ballistic correction and dose delivery hashes. More than amais, this new invention embodies the pinnacle of radioscanography: imultané, or even cinéradioscanographie, associated with a real-time dosimetry applied during radiation therapy by multiple synchronous external beams. But more than ever, this invention also meets one of the most fundamental aspirations of conformational radiotherapy: the hitherto utopian combination of the instantaneousness of the image visualizing the target volume in its organic environment and this allowing the optimization of dose delivery, concretized into a device combining in the same complex stand two rotating megavoltage irradiation heads to a fixed electron beam computed tomography system producing kilovoltage X-rays to explore in slices multiple and 3-D the whole of the target volume irradiated on the basis of images not morphometric synthesis but diagnostic images, as they are made in three dimensions of space by any Imatron scanner Fastrac with electron beam. Taking note of this evolution, we wish to offer this time to the profession 'an opportunity of a real control of scientific excellence of the practice that the situation exposed above imposes on art and this to the extent of the means that for example, we have made available not only in our patented invention No. 01/01133, which combines simultaneous irradiation with two simultaneous beams, n mirror to permanent diagnostic means of control of the target and external contours, that of the implementation with statistical estimation of their variability associated with online and real-time dosimetry; but also as we will see later in the present invention. Solve all problems on a case-by-case basis, while spreading the radiation treatment, and find

<Desc / Clms Page number 100>

 solutions relevant to the problems of the moment t, at time t ,, with the pledge of the simultaneity of the response of the radiotherapist and the physicist with respect to a situation that the simulation of planning must certainly be able to solve with the bank of the data put in memory at the control console of the entire treatment system, not only to anticipate but to adjust and adjust by forcing a little on the different variable parameters so as to optimally optimize the treatment with clinical simulation, in progress treatment, which anticipates any current problems,. Once in the patient's bed, it is necessary to realize that the conditions change ipso facto and that it is necessary to adapt exactly to the new conditions exactly at each moment, by aligning for example the optical axes of the beams of therapeutic irradiation with the isocenter of the system and the center of the lesion.

This technical contract could not therefore be developed without considering a fixed imaging medium, such as the electron beam scanner, integrated as such and without any other additional adaptations than an improvement, in the device of invention, of the solid-state X-ray detection network. It is easy to see that the constant identification by means of diagnostic examination methods that are as precise, as sensitive and as specific as possible to any problem that arises during therapeutic irradiation, is d a great contribution to the safety and the precision of the irradiation, thus initiating the evolution towards the intelligent CT scan. These are the statistical evaluation methods associated with very reliable real-time imaging means. We realized that it should be possible to build an automatic computer capable of recognizing simple numbers (statistics) and calculation situations. This is the principle of a CT estimator applied to the present invention. Although it had only a few computer clock mechanisms for any technology, we built a differential machine capable of developing processing plans and executing them using expert systems, and finally developed in the essential ultimate phase, the analytic machine capable of accepting instructions and then processing them according to a series of instructions and logical sequences of assembly. Such a machine, capable of resolving various specific problems by itself, could have already performed functions identical to those of any modern computer, but the present art has mainly been oriented towards the protocols of intervention on mechanistic bases. computer-assisted rather than automated and intelligent electronic methods.

<Desc / Clms Page number 101>

2.7.2.1. - Stereoradiotherapy device
This device, the complete description of which has been made in the patent No. 01/01133, relates to a first intelligent device for instantaneous on-line measurement of the dose on the basis, according to one of the modes of its implementation, of the CT and control system for the delivery of a double synchronous beam of particles (photons, electrons, positrons, neutrons, protons) and methods of implementation and permanent control of the device in working order, characterized in that, according to the embodiment: linear accelerator (linac), microtron, 60Co, etc. collisions between photons, between electrons, between electrons and positrons, and even between neutrons and neutrons. between protons and protons, between protons and neutrons, etc., are electromagnetic interactions produced to increase the intratissular energy deposition, in an even more precise target volume, thus realizing in the target volume the new concept of two-beam stereoradiotherapy external synchronous radiating in a plane of mirror symmetry. It is also characterized in this. that it uses both the phenomenon of braking (bremsstrahlung) and that of collision of the particles with the aim of a substantial increase in the intra-tissue energy deposition. The Bilinac-Imatron device thus relates to a dual system and methods of its implementation for radiating therapeutically associated with a device for intercepting emissions of all parameters,. whose invariability is required to meet the requirement of very high precision. This dual device is embodied by a scanner, Imatron Fastrac, with electron beam. itself connected to a dosimetry console in real time to exactly stick to the object to be irradiated by a statistically significant permanent estimate and the underlying reality is not, in this case, an ima-. age of composition seen offline but is the snapshot of the moment; this for a better control of scientific excellence and the quality of radiotherapy acts, especially during the IMSRT.

2.7.3.2. - the IMSRT
The device called Bilinac-Imatron concerns, in one of the modes of its realization, that. we have described for the simplification of the presentation in its only form of linear collider (type JLC), characterized in that a linear accelerator (microtron) synchronized double waveguide to accelerate and debit particles that collide at within a given organic tissue volume said electromagnetic interaction zone. In the center of the stand a tunnel occupied by a scanner and irradiation examination bed, by means of two external beams

<Desc / Clms Page number 102>

 synchronous X-photons, or electrons, or electrons on the one hand and positrons on the other. Depending on the needs of the dosimetry, these beams can be of the same energy or spectral intensity or have energy asymmetries weighted at will to influence the shape of the isodoses, in the target volume; and finally provide a new type of modulation in the time and space of the intensity of the two beams, Intensity-Modulated Stereo Radiation Therapy (IMSRT), including IMRT, IMAT and a mixture of these two. fixed field and arc therapy techniques. The electromagnetic interaction of the beams is carried out with mirror symmetry of the two therapeutic irradiation heads. The symmetry . by mirror plan used here is the most natural. It is also the most fundamental, because by playing with two mirrors we obtain all the other spatial symmetry operations: the rotation (intersection of the two mirror planes), the translation that can be described by two successive reflections, etc., especially in view dose escalation.

. 2.7.3.3. - Principles of dose escalation, in stereoradiotherapy
Let us postulate that we are in the presence of n micro-dosimetric molecular targets. Combinatorial analysis and so-called multiple-target theory n can, depending on the density of energy deposition, apply to our device for two characters: destruction (D) and cellular repair (R).

> For an increase of P1, probability of tumor control: D must prevail over R> For a decrease in P2, probability of non-tumor tissue complication: R must prevail over D
2. 7.3.3.1. - Energy deposition density and simultaneous irradiation quantum yield with multiple external beams
1. - Electromagnetic interaction by convergence of two simultaneous beams
The case of using two simultaneous beams, such as stereoradiotherapy, with mirror symmetry leads to an increased probability of destruction, due to the high density of energy deposition, in the areas of electromagnetic interactions. of the two mirrored beams within the lesion of the order of P1 = ((n) 2) 1. D / RF + Diffusion of Coulomb, if F represents the number of fractions or the spread duration in days. Effects caused by nuclear reactions are evaluated, including increased dose due to neutron production and induced radioactivity resulting in a relative biological effectiveness factor (relative biological

<Desc / Clms Page number 103>

 effectiveness, RBE) increased by a factor greater than 1.03. Always knowing that diffusion represents the flow of secondary particles.

* 2. Electromagnetic Interaction by Convergence of Simultaneous Beams In the case of eleven simultaneous beams which have, two by two, ten possible combinations of the mirror symmetry planes of the present invention.

 Which gives P1 = ((n) 2) 10. D / R.F + Diffusion of Coulomb.

The correct ballistics is carried out in such a way as to leave completely all the surrounding structures belonging to the healthy tissues completely outside the electromagnetic interaction zone, where a single beam is encountered each time, whose potential and density of energy deposition are , as in the traditional irradiation exposed hereafter, relatively very weak.

 Hence P2 = n. D / R.F + Diffusion of Coulomb (weak).

3. - Transaction of a single conventional beam with a crossed tissue medium Coulomb diffusion is negligible here, except at very high radiant energies <20 MeV. Dose participation is here, unlike Electromagnetic Interaction, discussed above, an essential element of the traditional electromagnetic trans-action process. By not seeing in the transaction, as we have done until now, a current of radiant energies which passes, on the one hand, the elements according to well-established internal mechanisms (deposition by deep attenuation and unidirectional diffusion) in the same sense as the fluence of the beam.

In contrast, in cerebral radiosurgery, for example, the term F = 1 in StereoRadiotherapy, the term F = ~ 12 - 15 days max (two weeks). The maximum dose effects occur within the lesion rather than the periphery.

 While in PolysteroRadiotherapy, the term F = 3 - 5 days (one week of max treatment). Where conventional radiotherapy proceeds with F = 22 - 30 sessions (3-4 weeks) or more the number of fractions naturally tends to be reduced. In the case of conventional irradiation, the maximum of the dose effects occurs at the entrance gate of the single incident beam and decreases with depth. This has the consequence of being able to hinder enormously the tumor control for the benefit of cellular repair and particularly complicates the ballistic optimization, as will be seen hereafter in the example of the thoracic irradiation.

<Desc / Clms Page number 104>

2.7.3.3.2. - Example of optimization of thoracic irradiation
In order to optimize the probability of tumor control and to evaluate the use of inhomogeneous distributions, with or without intensity modulation of a single beam, of the target dose obtained with conformational radiotherapy plans. -
D, WR de Gersem et al. (1999) planned the treatment of 10 patients with stage III NSCLC cancer, to provide the opportunity to increase normal tissue toxicity indices and / or to increase the tumor control indices of non-cancerous lung cancer. small stage III cells (NSCLC = non-small cell lung cancer), using a conventional 3-D technique and a technique involving. the intensity modulation of the non-coplanar beam (noncoplanar beam intensity modulation = BIM). Two target planning volumes (PTVs) were defined for this purpose: PTV1 included macroscopic tumor volume and PTV2 included tumor volume, both macroscopic and microscopic. The virtual simulation then defined the beam shapes as well as the inciden- tions. both for the orientation of the corner filter (3-D) and for the contours of the segment (BIM). The weights of the filtered beams, the unfiltered beams and the segments were determined by optimization, using an objective function having a biological component and a physical component. The biological component included only the tumor control probability (TCP) for PTV1 (TCP1), whereas for PTV2 (TCP2), it included the probability of normal tissue complication (NTCP) of the lung, those of the spinal cord and heart; but did not evaluate the variability over time of these movable members as advocated by the present invention. Regardless of the potential bias of aiming, the physical component included, as the present advocates. invention, both the maximum and minimum dose as well as the standard deviation of the dose to PTV1. The most inhomogeneous target dose distributions were obtained using only the biological component of the objective function (biological optimization). By allowing the physical component, in addition to the biological component, the inhomogeneity of PTV1 was reduced (this is biophysical optimization). NTCP values as well as maximum doses or dose levels to relevant functions of organ volume were used as indices of toxicity to normal tissues.

 When optimization was performed, relative to the objective biological function and the biophysical function, the inhomogeneity of PTV1 in this example decreased for 3D- from 13 (8-23)% to 4 (2-9). )% (p = 0.00009) and for BIM plans

<Desc / Clms Page number 105>

from 44 (35-56)% to 20 (9-34)% (p <0.00001). The minimum doses of PTV1 (expressed as the lowest dose-voxel) were similar for both objective functions. Mean and maximum target doses were significantly higher, with biological optimization for both 3-D and BIM (p <
0.001, for all values). The probability of controlling the tumor (estimated by
TCP1 x TCP2) was, for example, 4.7% (3-D) and 6.2% (BIM) higher in the WER study by Gersem et al. (1999), for biological optimization (p = 0.01 and p = 0.00002 respectively). The NTCP (lung) and the percentage of lung volume exceeding 20 Gy were higher with the use of the objective biophysical function. NTCP (heart) was also higher, with the use of the only objective biophysical function. The percentage of heart volume exceeding 40 Gy tended to be higher, but the difference was not significant. For the spinal cord, the maximum dose as well as the NTCP (cord) were similar for the 3-D planes (D (max): p = 0.04, NTCP: p = 0.2), but they were . significantly lower for BIM (D (max): p = 0.002, NTCP: p = 0.008), when the objective biophysical function was used alone. Using conventional 3-D techniques, inhomogeneous dose distributions seem to offer these authors the potential to further increase the probability of uncomplicated local tumor control. They seem to be indicated to use techniques, such. as the BIM, which would lead to the escalation of maximum target doses, to limit the inhomogeneity of the target dose in order to avoid dose levels that are so high with equally high variances that safety becomes problematic (WR of Gersem et al 1999), without mastery of the phenomenon of statistical bias.

 These authors nevertheless considered that this type of optimization improves homogeneity. 3-D shots and BIM of the target dose as well as the minimum dose to PTV2 which decreased especially thanks to the optimization in the 3-D planes. After optimization, the dose-volume histograms (DVHs) of the lung and heart were shifted (changed) to lower doses for 80-90% of the body's volume. Since the lung is the dose limiting organ in stage III NSCLC,. a minimum dose of PTV1 increased at the same time as the dose decreased to the main lung volume; which suggests in these conditions an improved therapeutic ratio. Optimization allows 10% dose escalation for 3-D planes and 20% for BIM plans at the levels of pulmonary isotoxicity and spinal cord. The esophagus can become for escalation of the dose the structure. limiting dose when PTV1 extends close to the dose. The opfimisa-

<Desc / Clms Page number 106>

 The use of an objective biophysical function would, according to Gersem et al. (2000), an increase in the therapeutic ratio of radiotherapy planning, for stage III NSCLC. Once again we observe that a mathematical formalism is used to solve technical questions, taking into account real anatomical elements in presence with their physiological and biophysical variability.

2. 7.4. - Irradiation symmetry, organic symmetry of representation and
Therapeutic mechanism of anticancer transformation
The best choice of a therapeutic strategy, in Radio-oncology, must put in. • works with knowledge of the evolutionary modalities of the disease, and those of the technical means put into practice. The Radiophysicist is usually confined to the execution of calculations, and this is, from our point of view, a serious error of consequence.

 That is why this invention tackles the problematic of what should be done: a real-time control by the statistical methods of precise evaluation of the. delivery, as it is true that tomotherapy, for example, things evolve at the slightest involuntary movement of the patient or the treatment bed, every day and throughout the spread of treatment, this considering, among other things, regression under treatment of the tumor volume and its fold in the surrounding healthy tissues, according to the therapeutic response over time. At this sign. if, in a voluminal lesion / healthy tissue ratio of 1/3, we start from a given tumor volume surrounded by a target volume three times larger, delimited taking into account the inevitable minimum safety margin, at Jo day, we must expect time to a reversal of this report, given the melting tumor, disfavor healthy tissue gradually slipping into the forecast target volume.

. Maintaining a constant target volume (VC), over time, does not therefore seem logical and therefore does not exactly fit, in this case, the desire for accurate conformational irradiation, but rather approximate and harmful to healthy tissue . We must therefore approach a little more the reality, over time, of the target volume. It is this new approach that the Bilinac- device already proposed. Imatron, essentially related and for the purpose of the description with the embodiment of the two synchronous beams of irradiation. Otherwise, it is also possible to design other embodiments including an irradiating apparatus with a single single source of radiation. In this latter embodiment the two side windows, described in the aforementioned patent, of the joint space of irradiation and image communicates at the level of the vault of the stand to do

<Desc / Clms Page number 107>

 more than a single window with a 320-angle beat amplitude allowing the source to run around the patient support, with a semi-permanent CT imaging system, since the detector array is here retractable.

2.7.4.1. - Electromagnetic interaction dosimetry
But, the dosimetry which interests us here is that of the interaction of the two synchronous irradiation beams, which we know that this irradiation can be with energy of symmetrical or asymmetrical spectral intensity, according to the needs of the modulation.
IMSRT. But, the mirror symmetry of this irradiation modality brings one. privileged relationship at the molecular level and advances science in this area. It is a fact that the discovery of chirality by Louis Pasteur showed us that two chiral objects are images of one another through a mirror. We take the example of the left hand and the right hand to illustrate it, but it is the case for many objects around us: the heli. DNA, shells, elementary particles, etc., and here the two beams of simultaneous therapeutic irradiation Pasteur has patiently sorted tartaric acid crystals into two groups having symmetrical properties, those which turned to the left and those who turned to the right, before realizing that they did not have the same properties on the light, that is to say the electromagnetic radiation which include X-rays and other radiation. The ordering, the symmetrization of his object allowed him this important discovery of organic levogyric and dextrorotatory molecules. Thanks to this discovery, Pasteur linked right / left crystalline symmetry to symmetry of the same kind at the molecular level. There are therefore straight molecules and left molecules,. we should not continue to ignore it any longer in Radiotherapy. Symmetry then played the role of bridge between the macroscopic and the microscopic.

 Microdosimetry could not ignore this knowledge for long. Especially when we know that the discovery of X-rays in 1895 had brought a similar example to the extent that scientists quickly realized, as the figure lb shows, that the diffraction of these rays by crystals formed circles. , highly symmetrical structures. These images of matter came to support the thesis of crystals having the same properties and all these notions are at the very foundation of the therapeutic action of the rays and the Physics of therapeutic radiations integrating the new concept of isotropic diffusion of energy. ra-. diante deposited. Therapeutic irradiation with mirror symmetry on

<Desc / Clms Page number 108>

 tissue molecules arranged in a symmetrical structure could undoubtedly improve the theory of a single target to move to the theory of n molecular targets to increase the effects of radiation in irradiated tissues and therefore a massive destruction of molecules of target tissues and allowing to marry by spatio-temporal arrangement of the beams to the morphological anisotropy of the tumor form modeling the object.

We can already distinguish two embodiments of the same effect at the operator-machine interface. One is direct as when it comes to producing, for example, in the recipient of the speech, the understanding of its meaning, the grasping of its validity. This is the necessary representation of what happens in the target volume. The second is indirect. It consists in producing at first a certain one in order to produce later or to induce the production of another effect which this expressive (operator) seeks by this detour; adaptation feedback, first degree effect (diagnostic image), some receiver disposition, situation adjustment to actual and non-historical therapeutic irradiation conditions with two or more simultaneous beams.

2.7.4.2. - embodiment with two beams MV CT simulation on a dedicated Imatron Fastrac CT, according to one embodiment, which we will describe here the most modern of them: the Bilinac.

 The generator used is, in the state of the ancient art, generally a linear accelerator of the Sagittarius type, for example, delivering electrons whose available energy range extends from 7 to 32 MeV per step of 3 MeV, or photons with a maximum energy of 25 MeV. At the end of the vacuum tube, the electrons are monoenergetic to within 0.5%. They go through the exit window and the monitor room. The flow rate is adjustable from 5 Gy / min to 40 Gy / min at 1 meter from the source at all energies. The single beam is then reduced to the desired dimensions by a collimation system allowing fields of 4 x 4 cm to 30 x 30 cm at the axis of rotation of the device, which is located 105 cm from the source. The primary collimation of the electron beam consists of 2 pairs of movable lead jaws 20 cm thick, which serve as collimation to the photon beams. The additional collimator for the electrons is, like the MLC, attached to the X-ray main collimator and opens therewith; it consists of 4 blades 7 mm thick whose outer edge is located 33 mm within the geometrical limits of the beam defined by the photon collimator.

The phantom used for the measurement of beam depth efficiency is gener-

<Desc / Clms Page number 109>

30x40 cm polystyrene plate (density 1.03) stacked on top of each other, allowing thicknesses of 0.5 cm
40 cm. Two of these plans are dug cavities adjusted to the dimensions of the ionization chambers. The influence of the DSP on the depth of ray yield is mainly felt at the level of the dose at the entrance and the depth of the maximum.

The dose at the entrance is indeed higher as the dimensions of the beam are smaller. At 10 MeV, the dose at the entrance, for example, which is 85% for one. field from 10 x 10 cm to 90 cm becomes equal to 99% for a field of 1.5 x 1.5 cm to
90 cm. The maximum dose is closer to the skin as the dimensions of the beam are smaller. At 10 MeV, the maximum dose that is 2.5 cm deep for a field of 10 x 10 to 90 cm, is 0.5 cm deep for a field of 1.5 x 1.5 cm to 90 cm. The slope of the decreasing part of the. The yield curve in depth is even higher than the dimensions of the field to the skin are larger. The proximal portion of tissue between the skin and the lesion is where the energy deposition by trans-action is most absorbed.

 The variation of the dose as a function of the distance to the source is usually done according to the law of the square inverse of the distances. Measurements by the chambers were made at all energies for source-skin distances of fields ranging from 5 x 5 cm to 20 x 20 cm, measured at the axis of rotation of the apparatus. The dose rates are measured at different distances for constant opening of the collimator. For dual simultaneous irradiation it can be expected that the inverse square law of distance is not satisfied, but. rather, that of the primary and secondary diffusion of radiant energy.

 The variation of the dose rate at a point is also related to the opening of the collimator (and not to the dimensions of the field to the skin), in the case of a single beam. This relative variation depends on the energy of the beam and is all the more important as the energy is lower. At 8 MeV the dose rate varies, for. fields ranging from 5 × 5 cm to 25 × 25 cm, for example between 50% and 120% of its value corresponding to a field of 10 × 10 (these fields being measured at 90 cm from the source). At 28 MeV, this variation does not exceed 2%.

<Desc / Clms Page number 110>

3. DESCRIPTION OF THE INVENTION The invention relates to a CT-guided radiotherapy device as well as to several pre-and intra-operative 3-D clinical planning and treatment planning methods, characterized in that it comprises, according to the preferred embodiment of icosahedral symmetry, on its rotating gantry: five polyanodic X-ray tubes placed equi-angularly, in a single orbit rotating at 0.5 s / cycle inside a stand, configured so each projecting several distinct simultaneous fans of the concentric conical bundles from said tubes, provided with a collimation system which restricts the lateral range of the irradiation range of the beams of the X-ray tubes intended, such as in CT multi-phase helical (to reduce the width of the fan and to decrease the secondary diffusion ... [cf our patent 00/17333]) as well as to irradiate simultanémen t one and the same object in cyclotherapy; six fixed tubes placed in pairs on half hops, intended to radiate the same target synchronously to the cyclotherapy, by beams outside the stand and fixed fields, and able to generate at the same time dynamic diagnostic images computer-aided tomography (CT) system with methods and means for dynamic management of diagnostic quality digital radiological imaging and tissue tissue statistical evaluation of the balance position of the human body around which oscillates various movements, voluntary and involuntary, together with a method and system of physiological induction of radiotherapy. The delivery of the beam (s), consisting of radiation source (s), emitting as required in isolation or in a joint manner, thus comprises a stationary system and a rotary system mounted on a gantry that can turn around the patient on a 360 arc, is simulated beforehand to plan the treatment. In one aspect, the present invention relates to a method for scanning an object with a multi-layered CT imaging system of detectors each having an isocenter. The method comprises the means for helically balancing an object with the multilayer CT imaging system to obtain data segments.The unique CT means facing, in the preferred embodiment with icosahedral symmetry, an opening by which the transmitted rays reach the fourth generation detection array disposed around the rotation orbit of the tubes of the rotating gantry and a system

<Desc / Clms Page number 111>

 Computed tomography data acquisition system (DAS), as a central system server, to an online dosimetry console.

So that, when the CT scan is not necessary, during the therapy procedure, the CT tube rises from 120 to 240 kV and the access to the detector network is closed by an imaging means for characteristics of 240-250 kV orthovoltage therapeutic radiation rays.

Orthovoltage uses all X-ray tubes in the system, including the CT, at the same time, which can be used at higher kilovoltage to perform therapy and achieve higher dose rates and therefore higher electron scattering efficiencies with X-ray conversion efficiency having a lower percentage of tube heating (eg, 250 kVP) as well as lower kilovoltage (80-140 kVP) for contrast more optimal image. In front of each therapeutic irradiation tube there are imaging means collecting the different characteristics of orthovoltage rays for dosimetry. It is the way of combining these different simultaneous beams of therapeutic irradiation and not on the conformation of the exit slit of the tube of the irradiation beam that a resulting morphology is determined which must conform to the anisotropy of the irregular form. or not the target lesion. Said exit slot generally has a simple hexagonal shape which can be more or less elongated, such as in commercial scanning, and to narrow the beam strictly to the target lesion as needed. A multimodal optical system, such as PCT / US00 / 28879, for tissue diagnostics, determines the clean tissue characteristics taking into account tissue heterogeneity.

OBJECTS OF THE PRESENT INVENTION AND THEIR IMPLEMENTATION By asserting that multi-phase polyconic projection radioscanography with helical scanning marks the beginning of a new CT age, the Strasbourg School of Radiology (ERS) does not hesitate not to insist on the innovation that accompanies the opening of a new fourth dimension in multi phase helical CT, as in the 01/16812 patent and a particularly daring style of 4-D polyconic projection tomography. , used here for therapeutic purposes, not only for the hyper-sophistication but especially for: - Objectively materialize the target volume, to follow the evolution, at the

<Desc / Clms Page number 112>

 the course of treatment from beginning to end, depending on its extent and precise topography with all detectable spatio-temporal variability.

 To materialize all the technical parameters by a double imaging system and to adjust the ballistics in a flexible and extremely precise way, in order to more precisely deliver the dose in the target volume.

 Clinically simulate, by the statistical method to do this on the patient himself, boundary situations with all the uncertainties of positioning, repositioning, angulation or incidence of the beams, etc., the topography of the isocenter in relation to the contours of the lesion itself and its variability as a function of time and not in relation to the cutaneous landmarks of the fields (beam entrance gates), this having regard to voluntary and involuntary movements, etc., as well as their aptitudes respective, and a statistically significant permanent ballistic optimization and a very user-friendly machine user interface.

. The optimization of the treatment is not done by virtual simulation but by a clinical act, simulating the bed of the patient on the patient himself, based on statistical methods, much more than on a virtual act conceived with a computer, the instantaneous 3-D acquisition data in real time from the patient's images, as well as any error conditions inherent to the eventual establishment. the alignment of the fields on the single objective and the ballistic aiming accuracy of the target lesion with its statistical variability, in time and in space. Unlike other simulation methods, ballistics does not remain fixed here, it is rather characterized by its flexibility and adaptability in close relationship with the actual conditions of the current session, bringing it each time. on a statistical basis, contemporary dynamic CT images of each gesture and an iterative test of the microscopic target volume. In fact, with the help of this system all important operating parameters become continuously variable and modeled from statistical data: this results in increased availability of torque and power, whose values are increased, at the same time. time that harmful side effects are reduced. These statistical methods, which we explain below, are therefore better than compensating for the inevitable, whose safety, comfort and wealth of equipment have been designed with the aim of setting new standards.

<Desc / Clms Page number 113>

3. Computer Based Machine User Interface Computer Systems on the Present Invention
This orchestration of the present invention is characterized in that various avant-garde technological subsystems, serving precision and quality but also a perfect simplicity of use, perform different but interdependent tasks. are implemented. A method for using a computer-assisted tomography scanner that uses a radiant energy beam for imaging purposes to therapeutically irradiate a selected interaction (target) region with the radiant energy beam in which the step of scanning the re-. The selected step through multiple cycles includes the steps of: obtaining a gating signal representing these times when the selected region is within a radiant energy beam; and irradiating the selected region only during such times that such gating signals select (eg ECG: or cardiac diastole) or respiratory synchronization.

. In the computer system, this internal arrangement is relatively constant (it is the computer function); in physiopathological systems, the disposition of the elements may vary infinitely, and it is frequently modified by what is called evolution in time and space; it is the biophylic function. The object of the present invention is to calculate, besides a ballistic aim. precise, dose of radiation at any time in the body, using for example the three-dimensional Monte-Carlo transport or better dosimetry by kernels.

 The invention also relates, as in EP 0 945 103 A1 (Mertelmeier und Scholtz, 1999) to a device for the examination of living tissues by X-rays, and measures the electrical impedance with the aid of a device. an active X-ray source and radiation receiving device and means for influencing the tissue to means for measuring the induced potentials or currents, so that the production of an X-ray image and a measurement of the induced potentials or currents take place either. simultaneously, or sequentially. A fusion of the images thus produced by one and the same system is also possible.

<Desc / Clms Page number 114>

4. Brief description of the drawings FIG. 1 schematically illustrates the ballistic principle in Radiotherapy with the different aiming possibilities on a given number of shots aiming, according to the dynamic imagery of guidance or not, presenting according to the statistical data of the sighting impact four different scenarios. Knowing that CE = external contour; MS = safety margin; VC = target volume; and the lesion or goal. It is a question of exactly aligning the objective with the geometrical place of passage of the rays (impacts of the shots), to be able to produce very precisely an interaction within the objective, L. Figure 1b shows the diffraction spectrum of an X-ray beam irradiation of a single source.

FIG. 2 is a diagrammatic representation illustrating the major symmetries in the therapeutic irradiation space of the former essentially cubic (canned) paradigms, the different faces of which represent the coplanar fields of so-called cross-fire irradiation performed with unidirectional irradiation at successive beam firing or simultaneous non-coplanar tetrahedral incidences (only four tubes are in switched off mode) converging at the isocenter of the system and at icosahedral incidences (all thirteen tubes are operational) converging at the isocenter of the system and whose vertices represent the ray sources of the present invention. The Bilinac Imatron, in its simultaneous four-beam embodiment, based on a linear collider at the energies used in Radiotherapy, coupled to an electron beam scanner, systematically performs tetrahedral symmetry of therapeutic irradiation to achieve a flow rate. high dose and a consequent energy deposit.

Figures 3a and 3b illustrate the two gantry systems: the stationary arranged in pairs: 15, 16; 17, 18; 19, 20 with possibility of lateral rocking of the pair on either side of the central position, and independently of each other, so that a pair can shift at most its two half hops of 30 and at the end of each half hoop an X-ray tube: 8, 9; 10, 11; and 12, 13; in the different views of the device of invention with its six hemi-arches 15, 16, 17, 18,
19, and 20 at the ends of which are arranged six tubes 8,9, 10,11, 12,13 stationary anodes preferably fixed on either side of the plane (x, y) of the stand inside which turn five tubes with rotating anodes 1, 2, 3, 4, and 5. The assembly rotates around a single isocenter 1 and a patient support 24.

 FIG. 4 illustrates the preferred embodiment of the five-tube rotary gantry

<Desc / Clms Page number 115>

 equidistant and equiangular 1, 2,3, 4, and 5 placed in orbit, of which only one is associated in front of the fan-shaped beam of optical axis 7 in polyconic helical scanographic scanning projection, which through the window 15 impresses in the multilayer ring of the detectors DAS 14, constructed with a system of simultaneous multi-cutting acquisition of the CT image data transmitted by the examined image 34 for the purpose of radiotherapy guidance imaging; while the other four are Radiotherapy tubes alone. The source 4 is able to increase in voltage up to 250 kV for Radiotherapy use or remain at 80 to 120 kV for CT use. The other sources 1, 2, 3 and 5 of the X-rays each emit a 250 kV orthovoltage beam (axis 6 ', 6 ", 6"', 6IV) which is scalable in intensity and in diameter for exactly match the geometrical form of interaction included in the target lesion 20, always with negative margins of safety, knowing that because of the isotropic diffusion of the radiant energy, there is as well through electromagnetic interaction of the beams between them that between the beams and the medium diffusing a peak of energy deposition, widening in depth the energy deposition zone to the whole of the target lesion, margins taken, while the surrounding tissues are spared and not are interested only in transit deposition where trans-action in which the rules of deep attenuation and cross-checking of fields remain de rigueur. Opposite these sources are separate X-ray detecting means for each detachable beam as well as 16, 17, 18, and 19 fixed shielding the orthovoltage bundles, for detecting the dosimetric characteristics thereof and for producing a imaging materializing these beams themselves and their impact.

 Figure 5 illustrates the different incidences of therapeutic radiation with fluence mirror symmetry for therapeutically treating with two synchronous beams a target lesion placed at the isocenter of an electron beam scanner in coincidence with the isocenter of rotation of a linear collider in a device called Bilinac-Imatron, which must be referred to the description and the detailed drawings included in the patent 01/01133. It schematizes the new forms of intensity modulation by intratissile collision of the particles of the two distinct beams emanating from the two different sources 22 and 27 of irradiation, in order to increase the intratissular deposit of radiant energy by electromagnetic interaction in the collision region VC of the radiating particles as well as their secondary diffusion whose dosimetric characteristics are captured by the CRX.

 Fig. 6 is a diagrammatic illustration of the CT scan of a source

<Desc / Clms Page number 116>

 X-ray projection of the present invention according to the description of the stereographic tube of our patent n 00/17335 and its use in multisection scan scanning helical described in Patent No. 00/17333, intended to provide, during radiation therapy using eleven synchronous beams of radiation, according to the icosahedral symmetry (11 vertices = 11sources) in which eleven beams of photons are produced simultaneously from eleven orthovoltage X-ray tubes 4, 5, 6, 7, 8, 9, 10, 11, 12, 13 and 14. It illustrates the arrangement of the rotating anodes of the tube as well as the parallel orientation of their respective optical axes, intended for dynamic 4-D CT imaging during image radiotherapy. -CT-guided, an unbiased estimator based on statistical methods. It schematizes the arrangement in the center of the stand of the irradiation space at the same time as that of radioscanography, in the translational tunnel of the patient support. This production system, in accordance with our invention, of the five X-ray beams of orthovoltage by means of the tubes rotating on their orbit inside a rotary rack or gantry, to process so, and in the mode of performed at 11tubes, including six external fixed, a target lesion placed at the rotation isocenter of a CT scanner can radiate from a tetrahedral symmetry (mode complete) to an icosahedral symmetry (complete mode) to achieve a high dose rate and increase the intratissular deposition of radiant energy by isotropic diffusion of the radiating particles.

FIG. 7 illustrates the convergent arrangement at the isocenter of the same stereographic tube of the patent n 00/17335 with, contrary to FIG. 6, optical axes of the rotating anodes converging at the receiver in a triconic tube intended, like the ten tubes on eleven of the preferred embodiment of the present invention, to the only therapeutic irradiation by means of the X-ray multiple beam tubes. The receiver here being the scattering medium of electromagnetic interaction of said beams of therapeutic irradiation.

Fig. 8 is a diagram illustrating a modification, according to another embodiment of the present invention, in which the orbit of the rotating gantry of the apparatus of Fig. 6, this time has nine orthovoltage X-ray tubes 20,21, 22,23, 24, 25, 26, 27 and 28 placed inside the stand, to therapeutically treat a target lesion 19 placed at the rotation isocenter of a CT CT scanner, in one embodiment with 19 to 23 tubes, ten of which are arranged in pairs on five pairs of outer half-hoops. The difficulty of such technical prowess on a 4th generation CT CT is obvious, illustrating at the same time

<Desc / Clms Page number 117>

 the congestion that may represent as well as the limits of the control of the primary and secondary diffusion when one moves away from the preferred embodiment described above, in Figure 6.

FIGS. 9 and 10, taken from our patent No. 01/16812, schematically illustrate the orbits of multi-phase helical scan radioscanners inspired by the present invention and illustrate the principle of their adaptation to achieve a high dose rate (FIG. orthovoltage) for Radiotherapy.

5. Detailed statement and embodiment
Little used, in science until the beginning of the XXth century, the symmetry became thanks to Pierre Curie a real tool of forecasting of new phenomena, as the isotropic diffusion in tissue medium. Thanks to him, symmetry has assumed the status of great physical law: when certain causes produce certain effects, the elements of symmetry of causes must be found in the effects produced, or conversely, when certain effects reveal a certain dissymmetry, this dissymmetry must occur. find in the causes that gave birth to him. This is the case with the tumoral anarchy to which the arsenal of the therapeutic irradiation of Radio-Oncology is addressed. The diffraction of X-rays (see Figure lb) from a single beam shows, from the point of view of strict symmetry, a very particular aspect which gives an idea of the way in which the diffusion with more or less angle in the direction of the incident electromagnetic trans action, depending only on the spectrum characteristic of a given beam, and the electromagnetic interaction, where there is no specificity in equilibrium and where the spectrum can widen in harmonics and under multiple harmonics and densify at the same time, in a given volume while diffusing all azimuth in condition of symmetric structure of perfect isotropic organization. Since spectroscopy is a method of detecting the presence of different chemical components in an object, by analyzing the light it emits when it has been brought to high temperatures, irradiating with a scattering of energy. azimuth and a wide range possible can touch all sensitive organic strings. This is why it has always been the heart of the present invention to reduce the importance of the entrance door or even the heterogeneity of the scattering medium, to promote in-situ escalation of the dose. A broader spectrum includes as many elements as possible in its sphere of influence.

<Desc / Clms Page number 118>

It can now be thought that successful experiments on mirrored therapeutic irradiation models, such as that of Bilinac, on which the clinical simulation is carried out, in one embodiment, by a Imatron Festrac CT CT scan beam. electrons, as in the invention No. 01/01133, have allowed to acquire this necessary assurance before daring to attack the efficient integration in a set of therapeutic irradiation CT CT scanners of 3rd or 4th generation to one or more helical scan phases, for parallel processing planning and monitoring on an on-line dosimetry console coupled to the CT scanner data acquisition system (DAS).

The invention thus relates to a device and several methods which are characterized in that it is a light and compact device which makes it possible to irradiate simultaneously one and the same object and to generate at the same time cineccanographic diagnostic images of said object with their physiological or non-physiological movements of computer-assisted tomography, in a particular symmetry of irradiation satisfying at the same time the three requirements: simulation of the treatment, dosimetry and actual treatment, and consisting of a dynamic radiotherapy imageguidée whose radiation is effected systematically In this case, the objective of the present invention is to provide a low albedo, a transaction, a high albedo and very high accuracy, the interaction, and finally to collect statistical data of variability over time of said dynamic images. In this dynamic image-CT-guided radiotherapy, the interaction in a volume that is representative of the anisotropy of the target lesion and of the entire swept portion of the patient's body; choose an anatomical reference fixed point in the image represented with its involuntary or even voluntary physiological variances; and selecting at least one three-dimensional image (3D) from a sequence of such images with respect to a function of the relative positions of the moving body part and the reference fixed point exactly at the common isocenter 1 of rotation of the converging multiple 3-D beams and the target volume, in the statistically significant sequence of images, to correctly deliver the dose, under conditions close to interstitial irradiation by intratumoral or intracavitary implants, substantially minimizing the dose to the patient. all surrounding tissues.

The fact that a three-dimensional (3D) image is displayed by a medical computer system without reference to any anatomical landmark is a problem because the images can not then be compared with each other by

<Desc / Clms Page number 119>

 With regard to the same reference surface and to the same reference axis according to the patent WO 01/03065, a three-dimensional image is structured as a function of multisection tomography image data. An anatomical landmark of the displayed three-dimensional image is specified and a reference surface is created based on this landmark. This brief overview outlines the three-pronged conceptual framework: preparatory, therapeutic, as well as a dosimetric monitoring tool of the new invention, and henceforth proposes to open a reflection on the resulting allocation decisions. as part of the preparation of dose treatment for tumor and healthy tissue. This is why the object of the new invention is to construct on the basis of the particular icosahedral symmetry, which is very flexible and non-permanent, to simultaneously irradiate a target volume permanently visualized within the body with a multitude of mirror symmetries alternating in time, with varying degrees in a perfect (unsymmetrical) sinusoidal motion of dissymmetry and a method for. seize, as a function of time, with the cinescanography of topographical modifications by statistical methods, but also tissue heterogeneities, such as in patent PCT / US00 / 28879. When the system is at the position of perfect icosahedral geometry shown in Figure 2, the secondary scattering is isotropic. The latter is then oriented in space by deformation prog-. resurrects the icosahedron to fit exactly the more or less irregular shape of the target lesion reconstructed in 3D by cinecanography with all its asperities morphological contours. This last diffusion anisotropy voluntarily put in place during the oriented deposit of the simultaneous radiant energy constitutes the new conformational radiotherapy of electromagnetic-tick interaction, in one. single and same target volume, simultaneous multiple beams of therapeutic irradiation. The determining element of energy deposition is thus no longer the attenuation at depth and that of the conformation is not at all the iris system placed around the source, but a spatio-temporal construction by composition at the source. 3-D geometric objective of incidences of said beams. We . conclude, for our purposes, that the results are, from a spectral point of view, in a more complete range of radiant energy ranging from the value of incident radiating energies to the lowest diffusing secondary intratissive energy.

 The bulk of the dose is that achieved by the collision of the radiating particles with each other and between them the scattering medium in which said collisions take place, which we will examine below.

<Desc / Clms Page number 120>

 5. 1. - Isotropic Dif fusion of convergent radiating energies The use of a CT X-ray CT scanner, with several X-ray polyconic projection tubes, attached to the scanner stand to reduce the required processing time In one of the embodiments of the device, the radiotherapy can be radiographed by means of a CT radioscanner with three X-ray tubes with the tubes located at 120 degrees from each other and placed in the orbit of the rotating gantry inside the stand, outside of which are arranged in triangle on either side six tubes on stationary gantries. For the sake of the description of the present invention, it is true with reference to the diagram of FIG. 4, which is the mode of our preference, and FIG. 'odd (three, five, nine) x-ray tubes are placed inside a 3rd (G-3) or 4th generation (G-4) multi-array scanner (helical) scanner stand and has two gantry systems, stationary and rotary designs different from those fixed, illustrated by Figures 3a and 3b, placed on arches 15, 16, 17, 18, 19, and 20 attached to the bonnet of the stand The present invention comprises on the one hand a radiation source which could be a rotary anode or anode X-ray tube fixed as shown in FIGS. 6 and 7 or a stationary anode X-ray tube with oil cooling of the anode which delivers generally slower than anode tube rotating its radiation dose, but little t operate continuously and without interruptions for cooling the tube. For this reason, fixed anode x-ray tubes are generally preferred in orthovoltage radiation therapy, since the processing time is less, but the multi-mode system is much more efficient.

The apparatus is designed as a CT-guided irradiation apparatus so that it can produce, for the purposes of simulation, design and development of computer aided processing of any radiation sources. , by CT-scan dynamic 4-D imaging, and is characterized in that delivery of the radiotherapy dose by means of multiple synchronous sources of orthovoltage photons 1, 2, 3, 4, 5 and 8,9, 10, 11, 12, 13, for example in FIG. 4, with a view to radiotherapy procedures with a very high probability of tumor control without the complication counterpart and consequently with enormous possibilities of dose escalation not without interactive control of the position of the rays
X and exposure using the data of the multilayer instantaneous diagnostic images as well as the dosimetric characteristics of said irradiation beams, as reference and optimization information, being

<Desc / Clms Page number 121>

 treatment and at any time, to prevent the occurrence of possible complications, and predict the accuracy of ballistic sighting and dosimetry online and in real time. The X-ray scanning system comprises a display device, six stationary gantries illustrated in Figures 3a, 3b and a rotating gantry or stand with tubes mounted on the rotating orbit around the subject with or without a generator. tube, an examination table provided with an X-ray detector and an X-ray control system connected to the display device or touch screen, the gantry and the examination table and means for directing said photon beams simultaneously produced by said X-ray tubes in the material to be irradiated not only from the antipode situations of the sources but also according to other spatio-temporal geometric convergence compositions. X-ray control has user input to indicate the next X-ray exposure.

Referring to the drawings of FIGS. 4 and 8, a respective odd number of five and nine of the X-ray tubes angularly equivalently spaced around the stand, such that they provide a weight balance and, since the beams are not collinear, avoids the alignment of the beams on one and the same axis; a model of five rotating X-ray tubes could, for example, be used in places and places of the three-tube model. As shown in FIGS. 4 and 8, tubes with several rotating anodes, schematically illustrated in FIGS. 6 and 7, of which only one 4 and 20 respectively are intended for imaging, are used inside the stand, while that the tubes with several stationary anodes (a stationary anode dissipating the heat, cooled with the oil) or rotating, according to the embodiments, and with beams converging at the isocentre, illustrated in Figure 7, arranged outside of the stand cover, as in Figure 3a and 3b are used only for radiation therapy. The simplest embodiment of the present invention is, as we have pointed out above, that having three X-ray tubes only embarked on the rotating gantry and six others on stationary gantries, including three on each side of the as well as means for directing the photon beams produced by said X-ray tubes. It is also the latter not represented by a drawing which is the simplest and cheapest way to achieve, including in a set having a embedded X-ray tube generator. Indeed, in a combination of at least three (five, nine) tubes with four (three) anodes converging at the isocenter in the same stand, type Tri-Quatri-X, for example a 4-D CT scan is carried out. multiple scanning phases, such as

<Desc / Clms Page number 122>

 in FIGS. 9 and 10, of which only one phase is used for the CT imaging at the same time as for the radiotherapy, for the other phases and for the tubes external to the stand, by arranging the tubes in the following way: the mode, five helical scan phases and four anode focal sources on each tube) four-anode (rotating) tubes in the orbit of the tetraconic projection stand in three (five) phases over a 100 mm grating width x 3 (x 5) and a translational speed of the patient support, during a CT scan, of 60 cm / s for dynamic 3-D reconstructions. In this mode of operation where, unlike James Winter's device, data from a number of rotations would be added by iteration, to form together dynamic CT 4-D imaging data for real-time monitoring of the radiation therapy procedure.

On the other hand, it is known that in X-ray patient treatment facilities, the x-ray machine does not operate for the time each patient is properly installed in the treatment room. Generally the preparation time of the patient is about twice the actual irradiation time. From this, it can be deduced that a given installation is only really functional on about a third of the available time. Another object of the present invention is to reduce this dead time by a multiplex system electronic management and appropriate automation of the manipulations concentrated in a moment to reduce the duration, knowing that in the case of irradiation by two simultaneous beams of the Arrangement schematized in Figure 5 and described in the patent No. 01/01133, includes the use of a single unit of high frequency energy to drive alternately at least two waveguides concentrically, as shown schematically In Figure 5, the patient-treating capacity of such a microwave linear accelerator facility is multiplied by eight for an increase in cost of capital invested which is only a small fraction of the total cost of a new installation. complete. Rigorous descriptions of the fluence of photons or other radiating particles in the interaction diffusing medium are given in more detail in the aforementioned patent No. 01/01133, as a simplest type of interaction involving two beams. simultaneous therapeutic irradiation in mirror collision symmetry. The present invention further extends this technical concept to several synchronous beams with several mirror symmetries and further automation.

<Desc / Clms Page number 123>

The absence of bias means, in figure la, that the ballistic impacts of the shots obtained with the different impacts as well as the alignments of the beams with respect to the objective do not deviate, during the successive fractions (or samplings). of treatment, of the true topography of the lesion undergoing influences in space and time involuntary voluntary and physiological movements. A low variance of these fire impacts indicates that the estimates of the impacts of the ray shots are little dispersed around the coordinates of the target lesion and that there is little difference between the values resulting from the estimation and the actual topography and its variability over time. Thus any non-optimal control of the tumor exposes, because of the existence of bias, to the risk of recurrence or evolutionary recovery as soon as the therapeutic irradiation imperfectly carried out ceases.

In daily practice and according to the studies, 1/3 of the treated patients do not respond.

These figures must be interpreted in the light of the different possibilities of impact on the target by fraction, illustrated schematically in Figure la. While FIG. 1b shows by diffraction of a single beam of X-rays in a diffusing medium of quasi-crystalline structure, the consistency of the projectile used and gives an idea not only of the process of possible deposition of radiant energy in the medium. and how the fluence of the beam diffuse and decomposes in the direction of incidence, but also allows the interest of the technical concept of isotropic diffusion of primary and secondary radiant energy by interaction. electromagnetic of several simultaneous beams directed on the same target. Several simultaneous beams with multiple and convergent effects would therefore be more enriching, from the point of view of the collision diffraction and the density of energy deposition in the geomagnetic locus of electromagnetic interaction. Which from the point of view of the target theory increases the number and the probability of tumor control as well as that of highly targeted tissue destruction, in less time than it is necessary for the safeguarding of healthy tissue to come online account. This idea is not altogether new, and the basic technique had already emerged in 1861, when the German physicist Gustav Kirchhoff, who analyzed solar light, discovered, thanks to the very broad solar spectrum, the presence in the sun many known elements. Each line or group of lines in the spectrum is the signature of one or more chemical elements. Most of the known elements therefore have a well-typed spectral characteristic, the corollary of which is their irradiation with an albedo and a consequent relative spectrum. Since it is possible to identify

<Desc / Clms Page number 124>

 (groups of atoms of different elements) by groups of spectral lines that they produce, producing on the contrary a beam which can comprise a very wide spectrum makes it suitable to interact, in the field of interaction of the simultaneous convergent radiations, with a large possible number of elements of the scattering medium, provided they radiate with a wide fluence active spectrum over this wide range of elements. This is the fundamental reasoning underlying Pierre Curie's rule of symmetry.

The problem, for a single beam, associated with the theory of diffusion to increase the energy absorption is related to the anisotropy of this mode of irradiation and can be quantified by comparing the attenuation coefficients in a medium given for an albedo of high to low with the attenuation coefficient derived from the rigorous solution of the transport equation (van de Hulst, 1980). Since the scattering angles are oriented in the single direction of incidence or fluence of the single beam, it is therefore necessary to necessarily widen the field to include a certain margin of safety. It is this anisotropic diffusion of radiant energy in the only direction of incidence (unidirectional diffusion) or fluence of the beam and the albedo of the diffusing medium which constitute its technical limits. Spectroscopy is undoubtedly an important tool today for studying the internal structure of atoms and molecules. Irradiating atoms and molecules properly, however, requires the use of the widest possible range of irradiation energies of fluences, incidental and secondary, which can be modulated with the different spectral intensities of the simultaneous beams in interaction. And the IMSRT is based precisely on the modulation of the fluence intensity of the different beams relative to each other and to the conformation of the resultant of their respective angles of incidence on the morphology of the lesion. At the beginning of the 20th century, it was largely thanks to spectroscopic identifications that the first modern theories on the atomic structure were indeed verified. Today, the mechanisms that preside over the emission by atoms of a light of characteristic wavelengths are indeed well understood. Conversely, a detailed analysis of changes in a spectrum caused by corresponding electrical or magnetic fields of atoms and molecules is the theoretical basis of the present invention. If the most common use of spectroscopy relates only to the analysis of unknown substances and the determination of their composition; in most cases, the indicator spectrum is produced by the light radiation that causes

<Desc / Clms Page number 125>

 heating the unknown substance or spraying a small sample into a flame or arc. Modern techniques have reached such a degree of precision that traces of an element only entering the sample for a few millionths can be detected. The expected symmetrical operation of spectrometry is the lethal irradiation of a substance with the widest possible spectrum to which the various components would be sensitive to scan the most diverse and diversified elements of its intrinsic structure. Producing such a spectrum is possible only by having several simultaneous beams directed on one and the same objective reproduce an infinite number of mirrored planes.

5. 3. - Integration of the 4D multi-cut CT scan into a Radiotherapy device
Since the introduction in 2001 of sub-millimillimetric multiple-scan radioscanners with multiple scanning phases (Figures 9 and 10) capable of producing 4-D imaging. cinéscanographique or instantaneous scanoscopique, new perspectives have opened. To reduce, for example, the treatment time required for radiotherapy, a five-tube synchronous model instead of the multi-phase X-ray tube model could be used in a scanner CT scanner with three X - ray tubes, for example,. with the tubes located 120 degrees apart from each other on the circumference of the stand. Indeed, when multiple X-ray tubes are placed in orbit in a scanner stand, only one set of detectors corresponding to the polyconic CT tube and a multiplex system electronic management are required. Additional filtration of the beam by the use of copper or aluminum in the fab-. The opening of the tube field can be used, in order to spare the surface skin, to harden the radiation therapy beams. This device is able to confirm with a submillimetric precision by unbiased statistical methods the correct positioning, during the radiotherapy procedure, of the chosen target region while minimizing the dose of radiation outside the region. chosen target. The power is multiplied by two when we go from three to five helical scan phases with in the latter case the technical need for a more suitable fourth generation CT scan (G-4) of the X-ray radiotherapy X-ray system, capable of exploration inframillimetric CT scan of 100-150 mm length per rotation of the tube, this in less time than for traditional thicker sections (5 mm, 10 mm)

<Desc / Clms Page number 126>

art. While CT scanners of the present art offer a limited range of imaging, and generally cut thicknesses of 1.5 mm up to 10 mm thick, the present invention modifies this aspect of imaging by proposing 4-D isotropic multi-phase imaging, which is more practical for tissue characterization and the physiological determination of surface contours as well as infra-millimetric slices, from which the intralesional volume of electromagnetic interaction is accurately defined. without necessarily resorting to other modes of complementary imaging or image fusions of very different modalities. The exact reconstruction software is de rigueur. 4-D CT scan. Reasoning on the dynamics of movement thus highlighted by the statistical methods and their respective physiological amplitudes provides more precision to the actions taken. The appropriate terminology would be here those of CT and kinescanography, or even kinescano-scopie, which is capable of 4-D CT to produce rather physiological imaging. than sticking to strictly multi-phase exploration, whose understanding is confusing with the multiple phases of 4-D CT scanning. The present invention also provides a 4-D viewing geometry of the selected target region and conforms it by lateral collimation of the much more flexible therapeutic irradiation beams to restrict diffusion. lateral and unnecessary exposure of the patient. Then, the monitoring of the treatment can be done, on demand, by a visual observation of the simultaneous imaging. The monitoring of the therapy is automated by computer comparison of successive images of radiation treatment with an audible alarm that triggers when the image changes or shifts. A change in the image can. signify a problem with the treatment, such as the movement of the patient moving the target lesion out of the firing range. The image change detection can be done by known image processing techniques, such as the integral of the mean squared difference between the images, or the comparison of the image centroids, or the motion detection. especially volunteers. cross correlation. This is especially so since the present invention is equipped with the system designed by Masatake, Nukui, and Akihiko Nishide or Computed tomography method, a method for detecting the central position of a subject and a CT scanner. WO
01/49177 (12 July 2001) PCT / IB00 / 020467 (28 Dec. 2000) (7 January 2000 JP) (GE
Yokogawa Medical Systems, Ltd). This system makes it possible to detect if the cen- position. of a subject is offset from a scanning center and to notify a

<Desc / Clms Page number 127>

 human operator The result of this detection procedure, the central position of the subject is detected on the basis of the profile of a parallel view projected at a viewing angle of 90. If the distance separating the central position of the subject from the scanning center exceeds the limits of a predetermined range (50 cm for example), a warning is displayed on a cathode ray tube, which warns the operator to adjust the position of the subject and to repeat the imaging process. The dose to the regions at risk can be significantly reduced by using at least seven current beam incidence input gates, in which to increase the number of gateways beyond seven only a minor additional gain and this approach significantly increases the calculation time. However, it is not the same with the present invention using several beams and therefore several simultaneous fields. The accuracy of all these operations is rigorously guaranteed by an internal electronic clock.

. According to embodiment G-3, the present invention relates to a computer tomography system characterized in that it comprises a stationary section and a rotary section, such as in patent EP 1.086.653 A1 of
Marconi, electrically bonded by at least one intercon- nection slip ring or slip ring. On one side, the slip ring is configured under. form of a series of electrically conductive segments separated by non-conductive interruptions. On the other hand, the same number of emitters are in selective contact with the conductive segments depending on the position of the rotating frame. Also included is a demultiplexer that takes image data from a plurality of multilayer receivers and rearranges the image data into. a predetermined sequence. There is also provided an angular encoder to the demultiplexer for assisting the reconstruction of the image channels in the predetermined sequence. In the latter conditions, a single multi-array array of DAS detectors is required for CT imaging in front of the imaging tube 4, whether in the G-3 compliant embodiment (Fig. 4) or. in embodiment according to G-4 (FIG. 8). The device of invention is characterized in that the movements of the patient are, in so-called multi-phase exploration of the polyconic helical CT, better taken in images: the shake, the breathing, the cardiovascular beats, the peristalsis, the swallowing , etc., which will not have any impact on the sharpness of a 4-D CT scan, but the so-called multi-phase diagnostic

<Desc / Clms Page number 128>

 by statistical methods the amplitudes or the movements transmitted gradually. The solution generally envisaged elsewhere is, on the other hand, the reduction below 0.5 s of the scanning time, which, while not being, as in the 4-D mode of stereoscanographic X, instantaneous gives an impression of dynamic blur rather than assess the range of voluntary and involuntary movements. The gain achievable from the point of view of the temporal resolution results not only in a great precision of ballistic aiming, taking into account not only the physiology, but also in the conjunction, during the same session of therapeutic irradiation, effects of several irradiated fields simultaneously. not to mention the economy of caches and conformation of the fields by sophistication of the collimation; the economics of contouring and the reproduction of the drawings for the ballistic study, this insofar as several detailed cartographies are acquired in computed tomographic sections in 3-D of all these elements as well as their variabilities over time. From the standpoint . of the investment, it is undoubtedly gaining in the cumulation in a single device and on the same site of the CT simulator, of the 15 MeV linear accelerator and in the reduction on the various positions of the personal time, especially around the placement of the patient and the bundles. And from the theoretical point of view, taking into account the temporal dimension of radiotherapy is not at all. negligible and introduces a new conceptualization of the therapeutic strategy.

5. 3.1. - Historical antecedents
The first step in this direction has undoubtedly been carried out, with a view to an image-guided therapeutic irradiation, by James Winter (patent EP 0 382 560 AI), in the form of a device called masking and post-masking members, intended to be used. in sequential tomotherapy or in three-dimensional computer-assisted radiotherapy embedded on first and second generation X-ray scanners. But the object of the present invention does not bother tomotherapy. The technical concept of J. Winter was already revolutionary and relied mainly on the post-masking member, preferably aligned with the opening of the masking member. and both made of a material that attenuates the radiant energy, at a level of energy substantially uniform to that of the radiant energy passing through the masking member, to impress the data acquisition system. images for image-guided Radiotherapy purposes. In such a way that the scanner sensors, aligned with the radiant energy beam, can continue to provide the necessary imaging during the therapy procedure. This should ensure

<Desc / Clms Page number 129>

 the continuous positioning of the selected region to be irradiated, during the radiation therapy procedure itself and thereby avoiding any positional changes that may occur either because of patient movement or because of the patient's initial false positioning, before the beginning of the procedure. These scanners can deliver, under these conditions using a single rotating anode tube, an approximate radiation dose of 300 - 400 rads per hour, including the cooling time of the tube. This is the equivalent of two sequential diagnostic CT scans per hour, but with all sections oriented on the same target lesion (eg 120 cuts / hours x 3 rad / cut = 360 rad / hour).

The present invention is distinguished, in turn, by the absence of masking and post-masking members described above and proceeds on the one hand exclusively by a mixed cyclotherapy (associated with fixed fields), using a number of odd multi-anode tubes also having a dynamic anti-scattering system with side collimation jaws of the polyconic fan-shaped beam, such as in the patent 00/17333, and is, on the other hand, intended as a system estimator of all current irradiation parameters, clinical CT simulation of treatment planning, on-line dosimetry and real-time monitoring of dose delivery as well as permanent aiming monitoring ballistic. An aperture at the irradiation slot of the tube is selectively varied in relation to the location (location) of the selected electromagnetic interaction region and the source of the radiant energy beam, according to the method of Hirofumi Yanagita and Masatake Nukui of GE Yokogawa Medical Systems Ltd. (Process for Radiation Impact Alignment
X and method and device for X-ray tomography EP 1 121 899 A1 (08. 08.2001), so as to align the X-ray impact position with a fixed position from the start of observation of a patient to be scanned with an X-ray emission / detection apparatus, an X-ray focusing position is predicted based on the statistical data of successive scans, and the position of a lateral collimator and an X-ray detector placed facing the beam and adjusted so that X-rays reach a fixed position on the X-ray detector and the target lesion control image.

Next, the present invention geometrically calculates the aperture required at the tubes of the therapeutic irradiation slot according to the different orientation angles of the scanner space, to irradiate the target lesion while the scanner

<Desc / Clms Page number 130>

 turned. The computer can then control the processing process by adjusting the X-ray output, the opening of the beam exit window, the table position, and the stand rocker as required to perform the prescribed treatment on the computer. targeted target lesion. Monitoring of the treatment can be done by visual observation on simultaneous imaging screen. The issue of low complication is not only fulfilled by minimizing the only dose to organs at risk but this is valid for all non-incriminated tissues in the tumor process. The monitoring of the therapy can also be automated compared to successive images of the processing computer with audible alarm if the image changes. The image change detection can be done by image processing techniques such as the integral of the mean squared difference between the images, or the comparison of the image centroids, or the cross-correlation motion detection. etc. The other pair of jaws having a spacing that provides the axial width of the fan radiation beam. Each jaw is selectively adjustable as to leave a given slot between its respective jaws. The invention also includes a system and method of interactive control of X-ray position and exposure using image data as reference information, such as in EP 1 092 391 Al de General Electric Company (US Patent 418167 - 13. 10.1999), intended for a medical application to supplement or replace fluoroscopy.

Similarly, it has been confirmed that the radiation sources of the existing radiation therapy units are in spite of the higher developed powers that perform less well than the x-ray tubes currently used in most computed tomography scanners which provide a higher dose rate. high that neither cobalt nor other sources can do it. The stereoscopic tube of the present invention goes further and provides with three times as many anodes a flow rate three times higher, in a preferred embodiment with a single triconic trianodic tube, or four times more with a quadriconic tetranodic tube. of which there are five in the orbit of the rotating gantry inside a same CT scanner, destined for one to the location being treated of the lesion by diagnostic CT and for the other four to therapeutic irradiation. The more accurate instantaneous polyconic CT localization of isotropic effects on the target of simultaneous beams of radiant energy, while the lower standard kilovoltage X-ray irradiation energy level (80-120 kV) allows enhancement by their interaction in the fabric

<Desc / Clms Page number 131>

 target of the target region's iodine dose and a higher GWR, by switching to orthovoltage energies. The patient resting on a generally fixed support.

 A series of 4-D diagnostic slices and a repositioning of the patient is performed if necessary, and adjusting the table position in width, depth, height as well as several degrees of freedom in cranio-podal and lateral-lateral inclination. plateau of said table, to thereby move the target lesion to the isocenter of rotation, at the desired distance from the isocenter as determined by the conventional use of CT scanner image display measurement software described in this patent.

. The gantry and the patient support revolve around a single I isocenter. The CT Scanner stand tilting capacity of the present invention associated with the degrees of freedom of the patient support is geometrically much more flexible than that available with the usual linear accelerators ( 6 to 15 MeV) used for radiation therapy. The present invention finally provides an improved profile of. Radiation dose compared to conventional radiotherapy units due to a continuous rotation of five sources of convergent beams associated with six sources of static beams, realizing an increased number of fields simultaneously providing beams of therapeutic irradiation converging on the same target . The prior art includes the use of variable collimators to shape the. external radiotherapy beam, but not in simultaneous connection, as in the present invention, with diagnostic CT scanners, where only the peak scattering shape of the radiant energy interacting in the targeted scattering medium is counted.

5. 3.2. - Irradiation with several mirrored planes of symmetry
Indeed one hour of exposure of a standard scanographic tube of 120-140 kV. would give 360 rads; it takes 20 minutes of time if, for example, three conventional tubes (120-140 kV x 3) are used simultaneously in orbit to irradiate the same target; and about 7-10 minutes with 3 conventional orthovoltage tubes (250 kV) + 30% dose increase due to the combined effect of the three, in the target volume. It takes less time for the polyanodic tubes for acco. perform the same work by irradiation fraction because the device of the present invention has a number of odd rotating anode tubes and are of high thermal capacity and advanced design to produce said beams of radiation). X-ray tubes angularly spaced equivalently around the stand, as shown in Figures 4 and 6, provide a weight balance. The present invention is otherwise conceived, at the level of its CPU,

<Desc / Clms Page number 132>

 as a multifunctional estimator, which is characterized in that it is capable of combining several types of complex mirror symmetries for irradiating synchronous beams, ranging from tetrahedral symmetry to icosahedral symmetry.

5. 4. - Icosahedral Symmetry Orthovoltage Cyclotherapy In cyclopolysterotherapy the effects increase tenfold and the time is still reduced, one can easily reach 300-400 rads to the whole target volume in a session of 5 to 10 minutes of time. irradiation including the cooling time of the tube, which begins during the various physiological synchronization and is prolonged during the change of patients on the irradiation table. The present invention is based on a high voltage CT scanner using an orthovoltage therapy X-ray tube capable of operating at higher kilovoltages or orthovoltage for therapy and achieving higher dose rates and higher energies. Electrons with high X-ray conversion efficiency with a lower percentage of tube heating (eg, 250 kVP) for a more optimal image contrast at lower kilovoltages (80-140 kVP). In the preferred embodiment, the tubes of the stationary gantries include either multiple stationary heat-dissipating and oil-cooled anodes or multi-anode tubes of high thermal capacity and advanced design, such as the stereographic (projection) tube. in biconical, triconic or tetraconic mode) embedded on the rotating gantry, to produce an interaction of the multiple radiation beams, with several X-ray tubes and not to perform a tomotherapy (sequential or helical) but rather a phase cyclotherapy. multiple irradiation, when associated with a simultaneous movement, such as in helical CT, continuous translation of the table, limited to the target volume.

It is ultra-rapid orthovoltage and high precision cyclotherapy in the context of CT-guided multiple symmetry symmetry or symmetry-type symmetry over time that combines waves, when the cycle of passage of the Perfect symmetry at the relative dissymmetry is five times faster than the rate of rotation of the stand which is one cycle per 0.5 s, and particles (X photons) of therapeutic irradiation in its operating mode where the choice of a certain symmetry of spatial organization of the sources has an influence on the effectiveness of this

<Desc / Clms Page number 133>

 mode of allocation of radiating energies diffusing in the target as well as the functional organization of the device is adapted to the internal problems of conformational dose allocation and interactive intraoperative estimation of the exact conditions of delivery of said dose, suitable to reduce the error both human and instrumental by statistical methods. Does the review of the literature not indicate high levels of error in the radiotherapy technique, whose etiology is certainly multifactorial. But, the effort of resolution is mainly focused on the improvement of human performance by a culture of safety and intervention in the form of protocols.

 5.4.1. The general principle of a dynamic image-guided imaging CT scanner, for Radiotherapy purposes, of the present invention based on statistical methods or radioscanometry.

The invention relates to a method and a device for obtaining by radioscanometry a series of three-dimensional image data of an organ that is generally in voluntary and / or involuntary movement and that moves periodically in the body of a patient 34, thanks to a source X-rays 4 and an X-ray device comprising a multilayer X-ray detector 14 or radiocinescanography or radiocinescanoscopy, and means for directing the photon beams produced by the different synchronous sources. With this device is obtained at the same time as the series of projection data, a motion signal which are in connection with the periodic movement of the body organ or radiocinescannometry. Cross-sectional imaging of the entire region may be periodically repeated to verify patient movement and patient misposition and the effects of different radiation beams.

If a complex experiment of multi-phase physiological exploration results from the realization in a certain order, of a first simple experiment which can lead during the clinical simulation to no different results of measurement, then to nor different measurement results, during the first session or fraction of irradiation, followed by a second simple experiment in the second session that can lead to n2 different measurement results, then a third experiment in the second session and so on throughout the protraction of X-ray treatment; number of possible distinct results of the global experiment is equal to n = ni x n2 x n3 x ... x nk for, in addition to the planning simulation, k therapeutic irradiation sessions. The major challenge is to determine exactly

<Desc / Clms Page number 134>

 the amplitude of the errors made, by first eliminating any new opportunities for error and any measurement subjectivity, in: a) the implementation on the simulator of the patients with respect to their predictive setting, in view of the alignment of the planned treatment beams on the intended target; and b) in the treatment unit repetition of placement at each fraction of the patients and to explore how the components of the various systematic errors can be reduced not with a generally off-line decision protocol for corrections implementation, but using an image-guided estimator such as that advocated in the present invention, evaluating by statistical methods involuntary and involuntary external mobility and internal mobility (movements of organs and movements transmitted remotely ) over a sufficient period of time to obtain a statistically significant and representative sample of the different mobilities in question. These data, sampled differently at each session and thus belonging to separate samples, are not influenced by the censoring of inter-fractional intervals and guarantee the scientific excellence of the results obtained.

Each beam is in conformation Radiotherapy increasingly shaped today by a rotating collimator to match the morphological irregularities of the targeted lesions. Tomotherapy also has an evolutionary modality which, it seems, advocates the promise of conformation of the dose to tumor volumes the best with a concomitant reduction in radiation-induced lesion surrounding structures. The present invention, for its part, shapes the spatio-temporal geometry of the peak of energy deposition, as we have already seen in the electromagnetic interaction volume, by playing on the respective incidences of the synchronous simultaneous beams of radiation as well as on the number of X-ray sources simultaneously switched. Thus, a method for using a radiant energy beam for external beam radiotherapy purposes in which the fan irradiation beams are laterally limited by anti-scattering flaps which delimit the beam width at the region of the beam. and includes the step of restricting radiant energy bundles of the scanner below the entire diameter of the target lesion (negative safety margin). The shape of the cross section of the field aperture of the tube does not matter, except to diaphragm to limit unnecessary irradiation. The switching of different tubes may, according to the principle of multi-phase helical CT, be carried out simultaneously or each source igniting independently and the resultant of the

<Desc / Clms Page number 135>

 teraction must eventually marry the shape of the target lesion. So that for simulation-only operation, for example, the scanner tube is switched on alone and the device operates at this time as a simple helical scan multi-cut single-phase scanner, with two opposite directions of translation during processing, the patient support in the tunnel of the stand.

5. 4.2. Automation of the device of invention Today, automation affects an infinite number of domains, and among its most interesting applications is the relationship between computers, doctors and patients in the context of clinical medicine. In the past, doctors or specialists used their knowledge (reminiscences rather) and their clinical experience (more or less long) to know if they had to neglect the description of the patient's symptoms and until to what extent objective tests, in this case diagnostic imaging, were required. Today, on the threshold of a new era in which computers can study a patient's history, conduct a physical examination faster and more accurately than by traditional methods, request compliance testing, examine results and prescribe an adjustment, it must be content rigorously its action on a more or less relevant intervention protocol. We still stick, in ancient art, the protocols of intervention to avoid errors inherent to both man and machine. Convinced that we are that the production and conditioning of quality care would be impossible in Radiotherapy without a very high level of automation of this practice; in ballistics in particular, where we know that certain elements are better to be fixed, in this case on a physiological rather than historical imagery, in a rigid support allowing to increase the range to which an adjusted shooting is possible: a pistol by The example does not usually lose its effectiveness at about 100 meters away, yet the bullets can still be deadly to several hundred meters, provided to rely only on chance. It is not the same for Radiotherapy where the chance of a single ballistic sighting session does not allow to overcome all the cancer cells and to cut down certainly a tumor lesion. This leads us to redefine, in optimal conditions, the ballistic aim in Radiotherapy.

Short processing times may also require the use of Artificial Intelligence today (LA.) Artificial intelligence is the discipline of building computer programs that mimic intelligence.

<Desc / Clms Page number 136>

human. While the computer formalism of ancient art mechanically manipulates information, Artificial Intelligence (AI) formalizes human knowledge and reasoning to produce expert systems. The physiological behavioral modeling that the present invention carries out consists in reasoning on the dynamics of movement thus evidenced by the proof of the image and by instantaneous traces of isodoses. CT scan and kinescanography, which is capable of 4-D CT incorporated in the present invention certainly makes this type of service. An expert system (SE) is a software or program that is characterized by its ability to reason by logical inference from a problem posed by the user. The system uses a knowledge base and a set of so-called production rules, made up of a human expert. Some researchers were already advancing the idea, in the early years
1960, that the laws of reasoning, combined with the power of the computer, could produce systems beyond the capabilities of human experts. And there. physiological behavioral modeling already performed in radiotherapy the present invention is part of it. To do this the device must include the
Operating systems and the following interface software: 5. 4.2.1. - Characterization system by artificial intelligence of simultaneous beams and physiological behavioral modeling. This method comprises the exact positioning of the patient on the patient support of the irradiation and imaging space inside a tunnel of a control scanner associated with means for imaging the characteristics of the photon beams. .

 The patient is normally placed in a known position on a platform that is surrounded by a stand, seat of the rotating gantry beams of radiant energy. embedded so that the opposing detectors 15, 16, 17, 18, and 19 receive the transmitted beams as they pass through the patient's body. The preferred method includes the steps of positioning the patient in the scanner tunnel using the positioning constraints of a restraint apparatus to prevent movement of the patient relative to the scanner. Then a multi- series. 4-D sub-4-D diagnostic polyconical CT scan of the patient is performed and a region corresponding to the seat of the lesion is fully explored, as is normally done in the art in diagnostic 4-D CT, for the purpose of therapeutic irradiation of the patient. After the preliminary cuts, the coordinates of the lesion or target region are determined using the CT scanner computer display with their statistical variability.

<Desc / Clms Page number 137>

 relative to the rotation isocenter of the scanner and the patient is moved, if necessary, the distance and in the corresponding direction to bring the target lesion to the isocenter of rotation of the system. The patient is, if necessary, automatically repositioned to move the chosen region to a predetermined location, because the device of the invention is characterized in that it still has a remote control system for positioning the patient associated with an observation / intervention device , such as that of (Philippe Cinquin, and Jocelyne Troccaz (WO 01/05319 of January 25, 2001 - PCT / FROO / 02042 of July 13).

2000) which consists of a remote control system for patient positioning of an observation and / or intervention device comprising a frame to which the device is linked according to a certain number of degrees of freedom, flexible connection means of which each is disposed between the frame and a point integral with the patient support or the patient himself, remotely controlled means for changing the length / voltage of the connecting means, and means for remote observation of the behavior of the device. The spatial dimensions of the chosen region are measured.

The selected region is scanned several times to statistically measure the range of voluntary and involuntary movements to irradiate the selected region to a truly chosen level of exposure. But also to vary the CT scanner tilt angles according to computer-assisted repeated tomography sections to therapeutically follow and radiate a selected three-dimensional region, and then apply the selected translation movement of the CT table. computer-assisted tomography scanner, to move it within and outside the slot opening to accompany the tilting movement of the stand above and below the physical rocking axis at the location of the the chosen region. In any case, after the series of preliminary tumor localization cuts, the size and location of the target lesion to be treated is determined manually or automatically by silhouetting the target lesion, notoriously identified on the display screen. with all means of radiographic diagnosis, then including said lesion in a three-dimensional geometrical shape internal to the surface contours of the lesion, or by other conventional means, such as the system proposed by Shabbir B. Bambot, Tim Harrel, and Anant Agrawal [Multimodal Optical Tissue Diagnostic System. WO 01/34031 (May 17, 2001) - PCT / US00 / 28879 (May 11, 1999) (SPECTRX, Inc.)], which consists of an apparatus and method that utilizes a radiation source as well as a processor for combining more of an optical modality (spectroscopic

<Desc / Clms Page number 138>

 that), namely and non-exhaustively, fluorescence, absorption, reflectance, polarization anisotropy as well as phase modulation, to decouple the morphological and biochemical changes associated with tissue changes due to diseases (cancer ), and thus form by esional tissue characterization an accurate diagnosis as the irradiation of the state of tissues.

 5.4.2.2. The present invention is also characterized in that it includes user interface elements, which it is important to know: ## EQU1 ## A touch screen (1985), such as the one presented in FIG. 1985 by Zenith (USA) as the first touch screen system, based on the technology of surface waves: just touch a sector of the screen with your finger to give an order. Knowing that this menu represents the control panel of the dynamic operating system with graphical interface.

 # 2) An optical pencil (1963), which is an accessory with which the user can draw on his screen, as he would with a real pen. The first optical pencil, associated with the Sketchpad conversational graphic system, was presented in 1963 at the Massachusetts Institute of Technology (MIT) by LE. Sutherland.

 Today, this accessory is mainly used, as in the present invention, on CAD consoles (computer-aided design).

# 3) But most importantly the CPU of the present invention has been designed to work as a very accurate multifactorial estimator and database of a computer-aided design (CAD) system with expert systems typical of CAD and software. Artificial Intelligence (AI). This control system is called closed loop, because a feedback indicates at the input the conditions prevailing at the output. Any external disturbance is automatically countered. An open-loop system does not have this feedback device. This is the feedback, which, by closing the loop, allows the output to always oscillate around the value actually desired. A system of
Computer Assisted Design (CAD) (1968) - CAD invented by Pierre Bé
Bezier (1910-). Computer-aided design originated in the 1960s, in the major American aeronautical military programs, on one side of the Atlantic, and on the other, at Renault, with the Bezier curves, developed by Pierre Bézier (see Biography at the end of the book). Presented in 1968 at the Detroit Automotive Engineers Congress

<Desc / Clms Page number 139>

 (USA), the Bézier system was quickly adopted by competitors and then by other industrial sectors. Today, the CAO has penetrated all areas.

It makes it possible to draw an object in three dimensions and to examine it in a very large number of fictitious conditions, even before having begun to build it ... no loss of time, rigor of the gesture and yield assured in the analysis , as in the present invention, of any movement (In 1888,
Marey, thanks to a new device, obtains independent images on a film Celluloïd, which, by projection, recomposes the movement by cinéscanographie): # Spotting of the scanographic sections containing the target lesion and recomposition of the possible movements with their amplitudes, possible correction of set-up - installation at the isocentre of the lesion center: about 5 minutes.

 > Planning and clinical simulation or conception and development of computer-assisted treatment: 7:00 pm to 10:00 pm (9 patients) + Weekends.

 > Therapeutic irradiation, surveillance and optimization: 10-15 minutes per session.

> Beat for change of patients: entrance and exit: 5 minutes> Every thirty minutes a new patient from 7:30 to 18:30: sick / day
Let us define, within the framework of the present invention, what we mean by dynamic estimation and unbiased estimator image-guided.

 5.4.2.3. - Estimation and estimator in dynamic image-guided multi-beam radiotherapy.

The present invention is equipped with a central computer (CPU) which functions as an accurate multi-function estimator and a data bank (Launched in 1980 by the American publisher Ashton-Tate, dBase II is the archetype of database management software It is a microcomputer-based data system that classifies, sorts and selects information according to many criteria, and was developed in October 1979 by C. Wayne Ratcliff, who marketed it under the name of Vulcan. However, this is a decisive aspect of the problem: errors due to poor selection of the sample have even more serious consequences because they are often much more difficult to detect and correct than Formula errors: Without always being necessary, this is enough to avoid most errors due to the selection of the sample.). It provides the most accurate stereotactic positioning and positioning of the patient, due to the fact that one and the same device

<Desc / Clms Page number 140>

having a diagnostic imaging capability is used for imaging and therapy purposes and on the periphery of which there are online servers and dedicated consoles. The present invention is itself an unbiased estimator that statistically estimates the topography of the lesion, relative patient motion, and repositioning variability at each fraction, continuously providing real-time imaging and so-called multi-dynamic exploration. -phases of the patient; Extemporaneous and always contemporary physiological imaging with radiation therapy, which is carried out at the same time as therapeutic irradiation to verify the accuracy of the localization and measure the position of the target area. as well as the accuracy of ballistic sighting, with regard to the different effects of therapeutic radiation beams. It is therefore a therapeutic radioscanometry. It can also be used as a pretest simulator. The point estimate of the ancient art consists in attributing a single point value to the parameters studied from observations made on a single approximate measure. and not a sampling of the correct limbs, an approximation often made at a distance from the therapeutic irradiation itself. This last numerical value can be called approximated point estimate and should in principle be ideally the result of several measurements or sampling of the same parameters made during the different fractions of therapeutic irradiation on the volume. target me and its immediate organic environment. In addition to its calculating role, the
CPU of the CT scanner plays an additional role of multiparametric estimator and data bank. It is more specifically the formula or the mathematical procedure used to obtain it which is called estimator.

 When moving from one sample to another, the estimator (the formula) remains the same. but the estimate (the result obtained from the formula) varies according to the observations. The true value (coordinates of the lesion and possible modification of its topography and its dimensions as a function of time) varies each day differently (It is therefore necessary to constantly renew the sampling).

 Each establishment of therapeutic irradiation is therefore a new opportunity. to sample, according to the induced mobility that this new tool detects, the various positioning and lesion topography parameters, that we need, as we will see, to precisely determine thanks to a CT-guided multi-beam collider, in the case of two simultaneous beams or another system of perioperative scanning capable of being operated continuously,. as a control tool and instant verification on the touch screen of the voucher

<Desc / Clms Page number 141>

 flow and quality of delivery of synchronous multi-beam radiotherapy. In the new invention the distribution of a characteristic random variable X associated with an event A corresponds to a binomial distribution, (3 (n, p).

n n ( 1 ) E P(k) = 2 Enk pk qn-k = (p + qt = 1 ; A partir duquel on peut estimer : k=0 k=0 - Calcul de l'espérance mathématique : la définition de l'espérance mathématique est m = k = #nk=0 k.Cnkpkqn-k ... m = np - Calcul de la variance : V(X) = E(X2) - [E(X)]2 - En résumé, pour une loi binomiale p(n,p), on a : E (X) = np et V(X) = np(l-p), permettant une modélisation comportementale physiologique. The operation therefore consists, at the level of the CPU, in calculating the expectation and the variance of X. Otherwise, after n tests of a multiphasic dynamic exploration, the probability for X to take the value k (0,1, .. ., n) is: P (X = k) = Cnkpk.qn-k (k is the number of times event A has occurred); while p = probability of performing event A for an event, knowing that q = 1 - p.

nn (1) EP (k) = 2 Enk pk qn-k = (p + qt = 1) From which we can estimate: k = 0 k = 0 - Calculation of the mathematical expectation: the definition of expectancy is m = k = # nk = 0 k.Cnkpkqn-k ... m = np - Calculation of the variance: V (X) = E (X2) - [E (X)] 2 - In summary, for a binomial law p (n, p), we have: E (X) = np and V (X) = np (lp), allowing physiological behavioral modeling.

 It is then possible to operate each time an adjustment of the model. Adjustment is a statistical technique that makes it possible to take into account, in a single test, a factor that increases the variability of the judgment criterion and the pledge of scientific excellence of the procedures used in the present invention. Where in ancient art the notion of correction has something arbitrary and proceeds by way of proto- cols to avoid methodological errors and errors inherent in the machine.

 The new invention is aimed at an appropriate technology and technique, removing all opportunities for error and arbitrariness. Would not the recidivism be, in the conditions of the ancient art, the fact not only of the firing biased in a large proportion, as illustrated in the figure la, but also of a insufficiency of the spectrum of the single beam of irradiation therapeutic, whose spectral component would not interest all the wavelengths to which the irradiated elements of the tumors would be sensitive? It is, within the laboratories of the School of Radiology of Strasbourg, a serious track of research that we develop.

Otherwise, the computer of the invention device is characterized in that it uses combinatorial analysis comprising a set of methods that determine the number of all the possible results of measurement of a particular experiment. Knowledge of these methods of enumeration is essential

<Desc / Clms Page number 142>

 the calculation of the probabilities of control and complication of Radiotherapy acts, which constitutes the statistical basis of therapeutic ballistics, placing itself firmly in a configuration of a design and development of computer-assisted treatment with a base radio-kino- metry and online dosimetry data.

If the random variable X, illustrated by a firing impact point of Figure la and represented at the microdosimetric level by x which is a function of the number x of the cellular targets: number of damaged cells can not be repaired by the fractionation; x also follows a binomial distribution (x, p) of parameter x (x = 2 ... k). If the study is done on k cell targets and p = destroyed cells (D) minus repaired cells (R), (D - R), divided by the number of tumor cells, assuming that each cell is irradiated equiprobably , that is to say without aiming bias and at an effectively tumoricidal dose. A practical way to illustrate this formula and to enumerate the possible results of a series of experiments during which a series of therapeutic irradiation with several mirror-symmetry planes is experimented, which consists in using a diagram of tree arrangements of the different spatio-temporal coordinates of mirror symmetry planes with respect to the anatomical coordinates obtained with the device of the present invention. Knowing moreover that the arrangement of n elements p to p (p # n) is obtained with any ordered set of p of these elements, all distinct. An arrangement is therefore characterized by the nature of the elements or by their order. A total number of distinct arrangements of n elements p to p is denoted. Any arrangement of p objects can be constructed in the following way: we consider p boxes, numbered from 1 to p (p <n) ... Apn = n (nl) (n-2) ... (n-p + 2) (np + 1). Example of a number of mirror symmetries, two by two, in an 11-tube device: A211 = 11.10. = 110. By the way, Ap = n! / (n-p)! This gives an order of magnitude of the dose multiplying factor in a multiple synchronous beam interaction system compared to a simple depth attenuation proportional to 1 / R of an electromagnetic trans-action with the transit medium.

The same relation can be expressed otherwise: a) Arrangements with repetition of mirrored planes of symmetry: An arrangement of n objects p to p with repetition is an arrangement where each object can be repeated up to p times. The preceding reasoning shows that for each box, we then have n possible choices. The total number of such arrangements of mirrored symmetry planes is therefore: [alpha] pn = nP, in the case of the new invention.

<Desc / Clms Page number 143>

) Combinations Cpnds mirrored symmetry planes. We call combination of p elements taken from n (n # p), any set that can be formed by choosing p from these elements, without consideration of order. Two distinct combinations therefore differ in the nature of at least one element. We denote by Cpn the total number of combinations of n objects p to p. Noting that the number of arrangements of n objects p to p is none other than the product of the number of permutations of the p elements of each combination, ie Apn = Apn / p! = n. (n-1) (n-2) ... (n-p + 1) / p! = n! / (n - p)! But also, in terms of permutations with repetitions and combinations of mirror symmetry planes. A permutation of n objects with repetitions r1 = p and r2 = n-p (where p # n) is a partition of these objects in two sets, one of p elements, the other of n - p elements. To give oneself such a permutation of mirrored planes thus amounts to the same as to give a part of p elements among n, that is to say a combination of n elements, taken p to p. We thus have: Cpn = Pn (p, n-p).

 Since Pn = (p, np) = Pn (np, p) by definition, we deduce that Cpn = Cn-pn., A permutation of n repetitive objects (ri, r2, ..., r) is called also a generalized combination. Just as CPn = (np), we denote Pn (ri, r2, ..., @ k) = (ri, r2, ..., rk).

 And in other words, Newton's Binomial.

 Newton's binomial is the product of n factors equal to (a + b), ie (a + b) n # np = 0 Cpn.an-p.bp ... Cpn = Cpn-1 + Cp-1.permet a practical determination of the different coefficients Cpnau by means of Pascal's triangle.

Combinations of repeating symmetry planes in the present invention
Suppose we study the distribution of n objects according to r criteria (or theory of targets), and that we look for the number of such possible distributions.

Such a distribution is called combination with order repeat (number of targets). The number of these combinations with repetition is: [nr] = Crn + r-1 Indeed x1, X2, ..., xn the objects (or planes of symmetry in mirror). Indeed, x1, X2, ..., xn the objects. A distribution of these objects (planes of symmetry) according to the criteria (number of the theoretical targets) can be represented thus: x1 X2 X3 / X4: x5 x6 / / X7 / ... / xn-1 xn /
The number of combinations with repetition of the planes of symmetry is therefore equal to the number of ways of separating the xi by r borders. It is thus the number of ways to choose r objects among n + r - 1 without taking into account the order. This makes it possible to estimate the effect of the interaction of the irradiation beams in the different mirror symmetry planes and leads to statistical dosimetry. Before

<Desc / Clms Page number 144>

 to see the applications of such a sophisticated system, let us see the least complex configuration, consisting of the two synchronous beams of irradiation in symmetry with a single plane in mirror.

 # 5.4.2.4. A device, such as in the General Electric Co patent, for managing data relating to peripheral devices and subsystems in a multiple imaging system including a memory circuit, and a signal processing circuit placed thereon. , where desired, in peripheral devices.

Data fabrication, data identification, service record information, calibration data, and other data can be stored directly in the peripheral devices. Peripheral circuits and subsystems may also include sensors, encryption circuits, and circuitry for interfacing memory and processing circuitry with other components, particularly with a controller for imaging systems. The peripherals may include DVD players and drivers for different on-board imaging systems (tissue characterization, X-ray beam characterization and scanning), table, monitors, and other ancillary devices. An initialization sequence is performed when connecting the device to the system, identifying the device and transferring the information needed to examine the sequences (Paul E. Licato, Peter C. Bosch, Suriyanarayana, Robert Steven Stormont: Method and apparatus for managing peripheral devices in a medical imaging system EP 1 110 504 A2;
469999, 22.12.1999 [General Electric Company]).

 5. 4.3. BACKGROUND OF THE INVENTION In one aspect, the present invention relates to a method for scanning an object with a multilayer multi-layer CT imaging system of detectors each having an isocenter. The method comprises the steps of helically scanning an object with the multilayer CT imaging system to obtain data segments including peripheral data segments combining data from a first peripheral data segment with a second opposite segment. to form a data set for reconstructing an image layer, and reconstructing the combined data into image layers. Data are recorded during a period of a few seconds of inactivity on the part of the subject during the time when there is still no therapeutic irradiation. Body movements due to respiratory activity are also

<Desc / Clms Page number 145>

recorded. The feedback device controlled from the desk starts a
A physiological triggering method and system predictive of radiotherapy.

Cardiac Synchronization: With respect to, for example, involuntary movements, the present invention utilizes the arrangement of a multi-phase computer assisted helical computed tomography scanner that utilizes a beam of radiant energy for the purpose of imaging for therapeutically irradiating a selected region with the radiant energy beam wherein the step of scanning the selected region through multiple cycles comprises the steps of: a gating signal representing these times when the selected region. is located inside a beam of radiant energy; and irradiating the selected region only during such times that such timing or gating signals select (e.g., ECG: systole or cardiac diastole). Better Hybrid Reconstruction System for high-pitch, multilayer, helical cardiac imaging.
Respiratory Synchronization: Breath synchronized radiotherapy (BSRT) has also been jointly developed between the University of California Davis Cancer Center and the University of California.
Varian Associates and complements the physiological behavioral modeling as well as the interactivity of the present invention as well as its adaptability to any. what clinical situation. BSRT describes an emerging procedure of Radiation Oncology, where simulation, CT series, treatment planning, and radiation therapy are synchronized with voluntary apnea, forced apnea, or synchronization (gating ) of breathing. The BSRT system consists of a breath emission interception system (BMOS) and a packet of. hardware of the computer gating of the linear accelerator and software. Two methods, a video camera-based method and the use of panoramic inductive plethysmography (RespiTrace), which generates BMOS signals. BMOS signals and synchronized fluoroscopic images are recorded simultaneously in the simulation room and analyzed later. to define the ideal point of treatment (ITP = ideal treatment point) where the movement of the organ is stationary. BMOS signals at the ITP can be used to synchronize a CT scanner or linear accelerator to maintain the same organ configuration as in the simulation. Unlike the BSRT system proposed by HD Kubo, PM Len, S. Minohara, and H.
Mostafavi (Breathing-synchronized radiotherapy program at the University of

<Desc / Clms Page number 146>

California Davis Cancer Center. Med Phys 2000 Feb; 27 (2): 346-53. ) authorizes apnea or synchronization, the present invention best matches the irradiation moment intermittently and the optimal phase of the physiological cycle or radiates continuously to the point where oscillation of physiological movements. This dual role allows the system to be applicable to a variety of patients, ie, apnea method for patients who can maintain and reproduce ITP, and forced apnea or synchronization method for those who, like children or uncooperative people are not able to reproduce exactly the same voluntary apnea. This is the sequence of operations allows a). continuous progress # 5. 4.4. - Dosimetry online and in real time by the device
5.4.4.1. - Geographical maps (1975) dosimetric
Since the mid-1970s, geographic information systems (GIS) have revolutionized this sector: on a traditional map, a geographer could only add a limited amount of information. With GIS, it is possible to superimpose several layers instantly: a base map, an irrigation network map, a map of soil types, a cadastral map, a satellite image of crops in progress ... This technology finds multiple applications, particularly in terms of spatial planning. With it the present invention can super-). ask to compare several dosimetric maps on top of each other on the same screen image monitor.

- 5.4.4.2. - Technical file of the device
Finally, a dosimetric record DVD (1997) per treated patient is made for the entire treatment. This CD disc of 12 cm works, like a CD, according to 5. the principle of optical reading. It uses the MPEG II compression standards, offers an excellent quality image, digital stereo sound and above all has a very large storage capacity (4.7 GB per side) and is suitable for the individual Radiotherapy technical file. In the consumer version, it offers movies with excellent resolution and laser sound. For IT applications, it). significantly improves the storage qualities available.

- 5.4.4.3. - 3D printer for custom molded parts
In 1987, a Texas engineer, Paul McLure, took a close interest in the work of Joseph J. Beaman and his pupil Carl R. Deckard at the University of Texas. he

<Desc / Clms Page number 147>

imagined the commercial potential of their discoveries, obtained a government grant and founded DTM Corp, to which he associated the university. This process - {Selective Laser Sintering Process) makes it possible to quickly create models in three dimensions. A computer, such as the scanner CPU of the invention device, cuts the object to be built in successive slices, which it memorizes the contours. A laser beam coupled to the computer scans the surface of a resin that polymerizes at the point of contact, reproducing the object. This digitized information can be transmitted to any compatible computer that, on the spot, can control a machine built by TDM and reproduce the object. In 1996, the American firm
3D Systems has come up with a machine that makes a plastic model that exactly matches the end object. The latter can be used to design a thermoformed mold for restraining and immobilizing a given region.

 5.4.4.4. - A helmet-screen (1965) on the dosimetry server, called dosimetric console.

To arrive at the concept of the virtual world and the total immersion of the user, it lacked an appropriate visualization means. The principle of a stereoscopic vision helmet was designed by Ivan Sutherland in 1965 on behalf of the US military, but it was only in 1986 that was born, under the aegis of the
NASA, the first HMD (Helmet Mounted Display). Its inventor is Scott Fisher who will impose himself, soon like the other guru of the virtual one. Intended for the simulation of missions in space, the HMD displays binocular information on two LCD screens and transmits to the computer the movements of the head: the displayed image shifts according to the direction of the gaze .

 This is the best way to better understand in real time the statistical effectiveness of the different therapeutic irradiation beams.

 Virtual reality is precisely used in this invention by the use of a helmet-screen to the dosimetry console coupled to the radiotherapy scanner. The device is previously or per operatively able to perform a clinical simulation together with kernel dosimetry of the interaction points in real time along with a control of the movement on the repetitiveness and the maintenance of the implementation. in place and especially of a ballistic aim without bias. While saving time on the overall performance of Radiotherapy acts. Since the clinical simulation is reduced, for example, to a simple so-called multiphasic exploration of a given region of interest; it thus occurs in spontaneous respiration to measure the amplitude of the different movements, without and with contrast injection.

<Desc / Clms Page number 148>

 for about 5-10 minutes for the acquisition of images as well as anatomical landmarks on coordinate axes that take into account said anatomical landmarks. Cineangiographic vascular mapping of the pedicle as well as the perfusion of the tumor to make the exact part of the peritoneum # and the target lesion itself. It will therefore not be necessary to repeat each session a contrast injection because the relay is taken, as described above, by the electronic tissue characterization to account for possible tumor damage over the whole of the therapeutic spreading. This brings us straight to clarify the Ballistic Principle implemented in the present invention.

5. 4.5. - Ballistic Principle in Dynamic Image-Guided Radiotherapy Allowing shots on a mu-moving target by intrinsic and extrinsic periodic movements, such as the ENT sphere, chest and abdomen, despite the movement and even more so when there is no movement. There is no movement internally, as for the skull, for example, without ever missing the aim of aiming. This unbiased ballistic principle is very important in the conceptualization of the new image-guided ballistics. Indeed, before executing an action or a command with the help of the mouse or the optical pencil, it is necessary to place the pointer on a well determined position of the extemporaneous 3D image. the shot is thus tightly adjusted, as shown in Figure la, without enlarging the target volume indefinitely and unnecessarily under the pretext of safety margins, which are known to be toxic to the surrounding healthy tissues. Knowing the displacement of the target organ and its timing we deduce a systematic synchronism (contextual menu): to coincide exactly with the place of passage and interaction of the rays is the primary objective of an image-guided ballistics. Otherwise things happen blindly despite the help of virtual reality, as in ancient art and the results are as illustrated in the figure the characterized by a great variability of the points of impact of the shots by fraction therapeutic.

In some situations, it may be desirable not to irradiate from the particular directions in order to spare specific anatomical structures such as the marrow. This can be achieved by switching off by a grid-controlled system in the x-ray tube. In the back and forth the tubes switch off at the end of the stroke and come on at a specific moment coinciding with the starting point of the target volume, as soon as the inertia of the table disappears and that the movement is started in the opposite direction at the speed prescribed in advance of the table, to avoid

<Desc / Clms Page number 149>

 overheating at the periphery of the target lesion. Need still a negative margin of safety (field joining phenomenon) insofar as the different beams of the beams must not necessarily be tangent to the edges of the target lesion. The conformation takes place here by perfect coincidence of the target lesion and the volume of electromagnetic interaction of the different simultaneous beams, role devolved to the optimization that of concretizing this perfect coincidence exactly matching the morphology and contours of the target lesion and taking into account its variability over time; Negative safety margins = Periodic oscillation dose fluctuation buffer, taking into account mobility. intrinsic, of the lesional topography.

 Hence the requirement in this invention of a permanent and necessarily interactive visualization of the target in order to target only the target by evaluating the dose in real time and optimal ballistics during the act of Radiotherapy to increase the probability tumor control and decrease that of complication. The goal . is undoubtedly to explore by image the possible displacements of a target volume during the treatment compared to the planned situation, in order to define the planning margins appropriate to the morphology of the target lesion as well as to identify the methods efficient to ensure or even better reduce these margins in negative margins, given the peak of radiant energy diffusion. These. margins should take into account all internal displacements of the target volume and external errors of implementation, at each fraction. Internal displacement for thoracic tumors is mainly due to respiration and cardiac activity. An estimate of the internal displacements of the tumor can for example be obtained from the study by Ekberg et al. (1998). who studied fluoroscopy for tumor mobility in all three directions of space in 20 patients. They found that for tumor localizations covering the left and right lungs as well as all lobes, the mean standard deviation (SD) of tumor displacement in the X-, Y-, and Z- directions were 1.4, respectively; 2.6 and 1.3 mm, during the. normal breathing. However, Ross et al. (1990) described how the extent of motion measured with ultra-fast CT was found to be highly dependent on the proximity of the heart / aorta or diaphragm. There is therefore a bias in this way of doing things since the comparison used here is historical. In fact, they describe a displacement from the to the respiration of the order of 1. cm for lower lobe lesions. Indeed, Ekberg et al. also observed

<Desc / Clms Page number 150>

 large variations in the range of motion, up to 12 mm. Apart from these large variations in amplitude and frequency with tumor localization, the distribution function of the probability of these movements strongly deviates from a Gaussian distribution (A. E. Lujan et al, 1999). Therefore, the standard deviation (SD) of these displacements is considered to be a poor representation of the distribution function and, when in such a case the expression given above for safety margins can be applied, it is unclear . This is why it must be measured exactly in each patient and at every moment. In addition, in this expression, the systematic and random movements are separated, but this separation is not trivial in the case of internal movements. In particular, the systematic errors introduced by the internal movement during the CT scan are not to be described directly since these movements occur on a time scale of the order of the cutting times of the scan of a CT scanner according to JM Balter and al. (1996) (cited by Hans C. J. de Boer, John R. van Somsen of Koste et al, 2001).

For an estimator to be appropriate, it must be verified, as in Figure 1a, that it has two main qualities: absence of bias and low variance. They can be understood, with figure la, taking the analogy of shots on an animated target; the center of the target represents the true value of a parameter (that is, the coordinates of the center of the lesion, which one wants to estimate and superimpose on the isocentre) and the different impacts of the shots represent the estimates obtained on successive samples or fractions of therapeutic irradiation or even better irradiation fields in which: # The absence of bias means that the estimates obtained on successive samples do not deviate systematically from the true value. In other words, their average is equal to the true value.

 # A low variance indicates that the estimates are not widely dispersed and therefore there is little difference between the values from two separate samples or irradiation functions.

 But, the instant ignorance of the modalities of extension of tumors, the imprecise delineation of tumor volume certainly before the advent of computed tomography and nuclear magnetic resonance imaging had initially led, in the name of margins, security, to the systematic irradiation of the whole body. This attitude is questioned,

<Desc / Clms Page number 151>

 particularly by publications having studied the site of recurrences after irradiation (A. S. Garden et al, 1991, L. Gaspar et al, 1992, K. Nakagawa et al, 1998). These studies showed that 70 to 100% of recurrences were within 2 cm around the volume taking contrast or resection margins. As a result, the target volume could now be adapted by taking safety margins of 2 to 3 cm beyond the contrast enhancement or imaging defined hypodensity. Moreover, studies comparing the results of biopsies performed under stereotaxic conditions with CT or magnetic resonance images have shown that the hypodensities could correspond to a single peritumoral # (P.J. Kelly Kelly et al., 1987). Spatial type 1 corresponds to well-circumscribed tumors, whereas spatial type III corresponds to infiltration by isolated tumor cells. This classification has the advantage of contributing to a more rigorous definition of the limits of the irradiation fields, but at the cost of several stereotactic biopsies according to embodiments that are not always easily applicable in practice. Clearly, this is a region of low tumor cell density that should be considered as such in the future in therapeutic irradiation strategies, opting for negative margins and thus irradiation. of little importance.

# 5.4.5.1. - Need for an increase of the total dose in
Radiotherapy of conformation ...

The existence of a tumor dose-control relationship has not, for example, been clearly established for all histological types of malignant brain tumors, particularly for low-grade gliomas irradiated at a dose of 45 to 60 Gy (A.B.

Karim et al, 1996). The team of Lee et al. (1999) analyzed failures after conformational irradiation in 71 patients with high-grade glioma and found among the 29 single-site relapses that 26 were central and three marginal. This despite the increase of the dose 70-80 Gy, this led the authors to continue their dose escalation up to 90 Gy. In the latter group, the failures occurred mainly outside the irradiated volume, and in particular bind in the neuraxis. In the group that received 90 Gy, however, two patients died of radionecrosis. In view of these results concerning relatively few patients, the interest of an increase of the total dose is certainly not established and the clinical researches are oriented towards the conservation of the healthy brain tissue (C. Haie-Meder et al., 1999), which is the major concern of the present invention

<Desc / Clms Page number 152>

 without giving up the dose escalation but removing any bias that may explain the poor results of radiation therapy.

 # 5.4.5.2. - ... and its correlation: Decreased dose in healthy tissues.

Long-term toxicity studies have identified the factors responsible for the complications of cerebral irradiation. As such, data on patients treated with protons are well documented. Of 367 patients treated between 1974 and 1995 for basal skull tumors, brainstem lesions were observed in 17 cases. The risk increased significantly with the maximal dose received by the brainstem and the volume of the brainstem receiving more than 50 Gy Cobalt equivalent. The number of surgical interventions prior to irradiation and the existence of diabetes were also associated with the risk of complications (J. Debus et al 1997). The responsibility for irradiation must therefore be assessed taking into account other factors that favor the occurrence of complications. The development of the non-coaxial technique allowed comparative studies with opposite parallel classical field techniques. For example, A.F. Thornton et al. (1991) compared the two approaches in 50 patients with high grade glioma. Dosevolume histogram analysis showed that, for identical dosimetric coverage of tumor volume, conformal radiation therapy was associated with a reduction (up to 30%) in healthy brain tissue included in 95% isodose. Over time, it seems clear that with the increased precision of the delineation of organs at risk through appropriate technology, such as that of the present invention, which will allow a better knowledge of doses tolerable by healthy tissue and relationships between volume irradiated and effective tolerance.

* 5.4.3.3. - Closer examination of tumor metabolism In addition, clinical applications of magnetic resonance spectroscopic imaging (MRSI) in the study of brain and prostate cancers have developed significantly during the last decade. Brain and prostate studies using proton NMR spectroscopic imaging (MRSI), based on metabolite levels before and after therapy, have demonstrated in a clinically reasonable time the feasibility for non-invasive assessment human cancers. NMR spectroscopy (MRSI) thus provides a unique biochemical window for non-invasively studying cellular metabolism.

<Desc / Clms Page number 153>

NMR spectroscopic studies (MRSI) demonstrated dramatic spectral differences between normal brain tissue (low choline and high N-acetyl aspartate) and prostate (low choline and high citrate) compared to brain tumors (low NAA, choline high) and those of the prostate (low citrate, high choline). The presence of edema and necrosis in the prostate and brain is both reflected by a reduction in the intensity of all resonances due to reduced cell density. NMR spectroscopy (MRSI) was able to discriminate necrosis (absence of all metabolites except lipids and lactate) from viable normal tissue and cancer as a result of therapy. Results from current spectroscopic (MRSI) studies also provide evidence that the extent of metabolic changes in cancer regions prior to therapy as well as magnitude as well as metabolic changes over time after therapy may improve our understanding of cancer aggressiveness and mechanisms of therapeutic response. Clinically, NMR imaging. combined with NMR spectroscopy (MRI / MRSI) has already demonstrated the potential for improved diagnosis, staging and treatment planning for brain and prostate cancers. Additionally, studies are in progress to determine the accuracy of anatomical and metabolic parameters by providing an objective quantitative basis for assessing disease progression and response. therapy (Kurhanewicz et al, 2000). This is a sufficient demonstration of the need to take a closer look, through a daily tissue characterization, to the fate of a tumor process undergoing treatment to achieve satisfactory control.

 However, the evaluation of the interest of the ballistic modifications authorized by a. The conformational approach should not be limited to the observation of a better distribution of doses received by the tumor and the healthy tissues located in its vicinity, especially when the irradiation is motivated by a benign lesion (pituitary adenoma). In fact, the dose received by the gonads can reach 3 to 4% of the contribution of a beam whose gateway is the vertex (I.J. Das et al., 1997). The use of non-coplanar fields also imposes an accurate evaluation of the doses received by the thyroid, doses which must be taken into account when defining the constraints of optimization of the ballistics of treatment Moreover, the tumors of favorable prognosis lend themselves better at studying the complications of conformational radiotherapy. Thus, the dogma of the irradiation of the whole of the. Posterior fossa is questioned in patients with medulloblastoma.

<Desc / Clms Page number 154>

 Due to the low recurrence rate observed in the posterior fossa, a more focused irradiation, guided by modern imaging means, could be discussed to limit the dose received by healthy tissues, including the cochlea (N .

Fukunaga-Johnson et al., 1998a, N. Fukunaga-Johnson et al., 1998b). From now on, we must think of systematically reducing the importance of therapeutic irradiation fields at the entrance to reduce unnecessary collateral irradiation.

 5. 4.6. Perspectives opened by the present invention of conformational radiotherapy In brain tumors, the use of multiple fields (coplanar or non-coplanar clusters) has already been proposed as a possible alternative to radiotherapy under stereotactic conditions (LB Marks et al, 1995). ). Technological developments, which have made conformational radiotherapy possible, are likely to further reduce the distance between these two techniques, which in fact have the same objectives. Beam intensity modulation is likely to allow conformational radiation therapy to compete with the most sophisticated approaches, such as stereotactic radiotherapy (RM Cardinale et al, 1998, SL Meeks et al, 1998) or proton therapy (RM Cardinale et al 1998, SL Meeks et al, 1998). The study of S.L.

Meeks et al. (1998), evaluating the potential clinical efficacy of intensity modulation versus radiosurgery, found only minimal differences between these techniques for the treatment of small intracranial tumors. However, stereotaxic radiotherapy retains the advantage of a more pronounced dose gradient outside the target volume and may therefore be associated with better saving of healthy tissue. All the more so, that of an isotropic diffusion of energy deposition with an entry gate with negative margins.

Standard clinical trial designs may lead to restrictive conclusions: the best treatments recommended on the basis of the test results, although generally applicable to patient populations, do not necessarily apply to individual patients. In theory, radiobiological modeling, coupled with reliable predictable assays, can be used to rationalize patient selection for particular radiotherapy regimens (B. Jones and RG Dale: Radiobiological modeling and clinical trials. 1; 48 (1): 259-65). But, linear quadratic modeling of radiotherapy can be used to simulate a clinical trial. This is achieved by random sampling techniques in which the parameters

<Desc / Clms Page number 155>

 Key radiobiological factors (alpha, beta, T (pot) and clonogenic number) are selected from the known or expected series. Radiotherapy development of the clinical trial can be enhanced by radiobiological evaluation designed to estimate potential changes in the tumor cure probability (TCP) and the biologically effective dose of probable normal tissue (BED). = biologically effective dose). Modeling can also be used to rationalize the assignment of patients to a list of tests or standards for individual optimization of a treatment regimen. Such approaches depend on the existence of reliable predictive dosages of radiobiological parameters in individual patients. The influence of variations in predictive dosage accuracy on the improved results is thus evaluated. Clinical trials, which were preceded by modeling simulation, potentially offer substantial improvements in cancer treatment outcomes by radiotherapy. These exceed the development of the preclinical trial and could include modeling assessments that indicate how to better structure the trial? It is clear from such modeling that what should be done is accurate and real-time evaluation by an appropriate estimator, as it is true that things evolve every day at the slightest involuntary movement or that of the treatment bed in tomotherapy, for example, and this also taking into account the regression under the effect of the treatment of the tumor itself and its decline in the target volume, as and when this evolution with respect to different reference contours and skin markers. Once the planning is complete, the centering data is then stored in the dosimetry console computer for comparison, optimization, and real-time intra-fractional dosimetry, as in our invention. 01.01133. So that if we start for example with a tumor volume surrounded by a target volume in a volume density ratio of 1/3, given the inevitable minimal safety margin, we should expect an inversion of this report as a function of time to the detriment of the healthy tissue slowly sliding in the target volume with regard to the melting tumor. It is by interactively repeating the test over time, at least once a week, that these essential elements can be objectified. Maintaining a constant volume therefore does not exactly match, in this case, the desire for precise conformational irradiation.

<Desc / Clms Page number 156>

 5.4.6.1. Verification of Virtual Simulation in the Art and in the Present Invention Several approaches have been suggested to improve finite width problems of the blades. The first approach involves superimposing the fields ... The control of the transition from the virtual patient to the real patient can be done at the simulator and / or under the device itself. It immediately raises the problem of the objective reference: the object or its shadow? The expected reference image for each unique beam of the art comes from digitally reconstructed radiographs (DRRs or digitally reconstructed radiographs) from historical CT scans. It will be compared with the radiological or electronic images obtained during the simulation and under the treatment device. Some image analysis software automatically calculates offsets between reference image and treatment image from previously identified boundaries of fields or anatomical structures. On the other hand, the setting up of certain non-coplanar beams can not be verified by this type of imagery. in these cases, the simulation of anterior and lateral fields centered on the isocenter is necessary to control the position. The implementation of a conformational irradiation is thus accompanied by a significant increase in the duration of the preparation of the treatment. The main source of variation comes from the survey of the macroscopic target volume and organs at risk, the difficulty of which may justify the use of other specialists (neurosurgeons, neuroradiologists, etc.), which is not the job of determining the volume target of a radiotherapy act. Moreover, the irradiation technique may require, in order to adapt permanently to the actual irradiation conditions, the determination of several successive macroscopic target volumes, which lengthens the duration of the target volumes. We can not honestly afford this luxury of pulling things in length, especially since the response to irradiation can be quickly, fraction by fraction, evaluated by statistical methods on extemporaneous imaging with the possibility of adjustment. Ballistics, reliably and predictively on a CT integrated into the linear accelerator (linac), such as in Patent 01.01133, or much more difficult at the cost of many trips by a diagnostic CT or MRI, network or not. In the technique of electromagnetic interaction, it is necessary to define by permanent visualization of the lesion not each time any target volume but an intralesional electromagnetic interaction volume adapted to the present circumstances of irradiation, surrounded by a region of action of

<Desc / Clms Page number 157>

 different beams of therapeutic irradiation and a lower dosimetric effect.

The present invention is characterized in that it comprises a 3D image data reconstruction system for a Ching-Ming Lai region (Three-dimensional computed tomography image reconstruction WO 01/06931 of 01 Feb 2001 - PCT / US00 / 18899 of 12 July 2000); Analogic Corporation) having a radiation source and a detector array on the opposite side of the region used to generate scan data of the target region from a plurality of diverging beams of radiation, it is to say a cone of radiation. One or more image slices are defined for the region. Each slice of image is divided into a plurality of slices of image, each segment giving rise to the generation of image data. The combination of image data for the segment generates image data for the slice. The image is even thinner as these segments are.

 5. 4.6.2. - Deposition of energies by trans-action and electromagnetic interaction At the beginning of the 21st century, we can hardly deny that technology and its evolution have a profound influence on the form and evolution of technical concepts and the rare allusions we have been able to do so far do not reflect the importance that should be attached to it. The nuclear reactions can be produced in the irradiated material, in particular by electromagnetic interaction and by interaction with the tissue medium. The energy above which these reactions can occur, in the current single-beam irradiation art, varies depending on the particular material with which one is irradiating and also the intensity of the incident beam. particles (photons, electrons), but this intensity is usually of the order of about 20 MeV for the accelerated electrons.

Induced nuclear reactions can occur at much lower energies than that in simultaneous irradiation with multiple beams. It is the challenge of the present invention to obtain such an effect with orthovoltage spokes.

James Winter had already reasoned with his masking device that it was to be expected that the unsealed therapeutic bundles used for the radiation therapy procedure would be several times more intense than the imaging portion of the radiation beam. which reaches the detectors of the CT unit.

The specific choice, as for the energy of the imaging beam, will be done in such a way that to achieve adequate radiation at the detectors, taking

<Desc / Clms Page number 158>

 in Winter's sequential tomotherapy system the ability of the CT unit to sum the radiation from a number of rotations of the tube to obtain the data of a single image, image having an acceptable noise level .

For example, if a typical unique diagnostic section represented 3 rads at the target region, a treatment could consist of 100 sections of the target region, producing a treatment dose of approximately 300 rads at the target region. If the radiation passing through the body of the masking limb, and not through the opening of the masking limb, is 1/20 of the radiation passing through the aperture, then an image quality equivalent to a usual diagnostic cut can be obtained. by combining 20 therapeutic sections of the target area. By displacement averaging, combining the previous number n cuts in an image (n =
In this example), the image can then be refreshed, and reconstructed again or refreshed, after each cut to include in the image projection data the n most recent therapeutic sections. By varying the number of. n cuts, a tradeoff can be made between image noise and how to actualize this image is. However, in the present invention, which relies instead on icosahedral symmetry cyclotherapy, we have, for example, four anodes per 250 kV stereoscopic tube. Which gives four times more fluence of radiation. The dose is no longer counted by the number of cuts made. but in terms of dose kernels depending on the different flow rates of the tubes and their respective contributions to the electromagnetic interaction. This was already our way of seeing things already with the Bilinac-Imatron whereas in the current art prevails essentially the following methods: 5.4.6.2.1. Point Estimation and Maximum Likelihood Method a) Qualitative variable, likelihood function; but the values used remain punctual. b) Quantitative variable absent, insofar as the lesion is visualized only historically.

5.4.6.2.2. - Estimate by confidence interval (2 #, 95%). No% or average and small samples used in point estimates made of sums of these so-called point values, regardless of whether their distribution is normal or not?
The technical systems of art constitute in themselves completely depersonalized systems, that is to say schemes of mechanisms operating independently of the human element. The choice of a confidence interval and not

<Desc / Clms Page number 159>

 that of a fluctuation interval does not go in the direction of more precision in the different manipulations programmed to contribute to more precision.

The confidence interval and the fluctuation interval must be calculated from the true values in the population (, # 2, fixed / observed). Not to be confused with the range of variation or range, which is the difference between the extreme values of the examined trait. Experience has taught us, however, that technical systems mean nothing on their own - they only start making sense when they relate to the realities of patient and bundle placement and they interest and prolong systems of resolution of the temporal problems, related to the change of variable (between the different implementations and the movements), namely the case of the change of origin or scale, where both at the same time, in the ballistic sighting coordinates.

 The average observed in ballistics is a random variable since its value varies, in a predictable way, from one sample to another and over time, in the various studies of the literature. But still it is necessary to measure it and to approach it very precisely. To give the fluctuation interval of a variable is to indicate the interval in which the observations made on a sample, or a parameter calculated from these observations, such as an average or a percentage, must be placed. knowing that in Radiotherapy each fraction is in itself a sample different from that of another fraction and requires a lot of statistically significant measurements. To access a series of measurements at each fraction, a permanent estimator associated with the therapeutic irradiation system is required. This is the optimal system that can lead to an optimal radiotherapy strategy.

 In some cases, it makes it possible to considerably reduce the intervention of man and the multiplicity of measurement systems of the current art by means of automation. This is what is called, in the present invention, the moment of a statistical series.

5.4.6.2.3. - Moments of a statistical series
Definition: we call moment of order q with respect to xo, the arithmetic mean of the powers qlème of the deviations of the values of the character with respect to xo, the value of reference. The presence of measurement biases must then be detected by an instantaneous representation of the reference.

 Presence of bias: The more scattered the variable, the larger the differences, and the greater the variance. Conversely, in a population where all subjects have the

<Desc / Clms Page number 160>

 the same value of X (which corresponds to a total absence of variability), the average is equal to this common value and the variance is zero. This is the optimal system.

 5. 5. - Optimal system, in conformal radiotherapy An optimal system is integrally interconnected in time and space, so as to allow a free and rapid exchange of information, services and means between all the subsystems. essential systems of which it is composed, is extremely cohesive, and at all times takes active attention to present needs.

To be optimal, the clinical simulation must, for example, be able to be done with the same estimator also at the planning stage of the treatment as in the procedure, as well as interactive online and real-time dosimetry. To be optimal, all these operations must be carried out on one and the same estimator in order to eliminate the accumulation of errors of several task-specific estimators, differed in time and space, with respect to the actual processing and between which patients must move from one site to another. Measuring instruments or estimators of current art still lack precision from this point of view. Rather than sticking to meaningless concepts that none of them determine on their own, why not simply adopt the accurate 3-D spatio-temporal CT definition, whether in the simulation or in the treatment, realizing what we call CT-guided radiotherapy or conformational radiotherapy guided by 4-D X-ray CT? Optimal system sticks to a schema of internal mechanisms that allows for a continuous and efficient process of solving problems. We were inspired by the new invention of what we were already proposing in our earlier inventions,
Patent Nos. 01/01133 and 01/16812 which can classify and control the effectiveness of its main products and use this information to reorient both the objectives it has in mind and the means at its disposal, to broaden the scope of electromagnetic interaction to a more complex mirror symmetry system.

<Desc / Clms Page number 161>

5. 5.1. - Image-guided electromagnetic interaction and trans-action in an irradiated volume 551.1. - The case of two synchronous beams converging on the objective or stereoradiotherapy Stereoradiotherapy concretizes the therapeutic irradiation by two simultaneous external beams, double synchronous acceleration and collision of the particles in the target volume, such as in the patent n 01.01133, which uses both the phenomenon of braking (bremstrahlung), characteristic of the radiation-tissue trans-action energy deposition phenomenon, and that of collision of the particles with each other and with the scattering medium or electromagnetic interaction, with consequent increase by secondary diffusion. essentially of intra-tissue energy deposition and intended, according to the embodiment, not only to the preparation of the treatment (centering, simulation, and treatment planning) as well as to the bicéphalique therapeutic mirroring mirror by double spontaneous beam converge at the isocentre towards the target lesion, may s also to real-time verification of the delivery of the dose. And to meet the requirement of a very high accuracy this device is coupled to an electron beam scanner, Imatron Fastrac, itself connected to a dosimetry console online and in real time and has a test time for the tissue characterization determination of the microscopic target volume and a mobile ultrasound volumetric targeting and contouring system for instantaneously adhering exactly to the object. Where the ancient art uses more or less pifometric morphometric means realizing images by means of the electronic imaging system (EIS or EPID) of synthesis, in which the imagined real is simply approached without any means of effective control of excellence scientific, the referent itself being a constructed image and not a true faithful representation, such as in Radiodiagnosis, of the object by the image. Which, when one knows the diagnostic quality of a constructed image, does not conform to the radiological semiotics that has been kept up to use, according to the well-understood triangle of symmetry of Sanders Pierce's transformation: object - image - representation, where reality holds a predominant place and not in the form of fractals. The philosophy based on coarse imagery, that is to say without any diagnostic quality, of the electronic imaging system (EPID or EIS), and, without being a criterion of scientific excellence, its guarantee of quality remains nevertheless in a quest for good conscience for the operator and a meritorious approximation to the requirement of quality assurance.

<Desc / Clms Page number 162>

Stereoradiotherapy by electromagnetic interaction of the two external beams, with synchronous acceleration and collision of the particles in the target volume, uses, as described above, a simple mirror symmetry of therapeutic irradiation both for the braking phenomenon (bremsstrahlung ) that for the collision of the particles whose consequence is indeed an increase in intratissue energy deposition. As in this new-type device, the desired effect was already the electromagnetic interaction of the two irradiation beams, which must be delimited within the target volume, it is clear that the frontal collision is the one which produces the strongest interaction between the two radiant energies and a maximum secondary diffusion. In the latter case, the volume of electromagnetic interaction integrates more than necessary into the surrounding tissues. This is why angle 180 represents in phototherapy a critical convergence angle of this new technique: 180 between the axes of two beams is an angle with maximum electromagnetic interactions, that will be necessary. reserve as needed for irradiation with charged particles or not, types of electrons, protons, neutrons, etc. rather than photons and avoid to return to the trans-action of conventional beams alternated in time rather than synchronous so that the electromagnetic interaction volume does not overflow the target volume. And on the opposite side there are angles with weak electromagnetic interactions between the '. two beams for angles less than 45 between the axes of the two beams. Moreover, the more energy particles will be better the energy deposition by collision of the particles of irradiation. Hence the obvious interest to embark on this device a double waveguide of a microtron, whose energy ranges from 10 to 50
MeV per milder of 5 MeV, rather than another dual-energy linac only.

. It is essential to accurately determine the dose delivered by electromagnetic interaction to the patient during therapeutic irradiation. Different factors intervene in the determination of this dose and for a given generator, when passing from one patient to another, the changes in the dose received are due essentially to the geometrical conditions which are different for each. that treatment. The dose received is influenced, inter alia, by the source-skin distance; field size, monitor response, etc. If the relative biological effectiveness (RBE) of two radiations of different quality is the ratio between the radiation doses (expressed in grays, Gy), delivered with these two radiations, which are necessary to cause the same biological effect.

. If a biological effect is obtained for an absorbed dose of n Gy, with a radia-

<Desc / Clms Page number 163>

 Type A and m Gy with type B radiation, the relative biological efficacy of type A radiation compared to type B radiation is m / n for this biological effect. The concept of EBR is an experimental concept, it has meaning only if it is related to a specific biological phenomenon. Efficiency increases with TEL in an intuitive way. This same hypothesis of morphological lesion necessitates a local density of sufficient energy. This is why the collision of the irradiation particles of the two simultaneous beams, in a given interaction volume, can only be greater by its local energy density than the only loss by trans- action with the medium traversed. the kinetic energy of the particles of a single beam of irradiation. The irradiation by a double synchronous beam is not always, from the point of view of the energy intensity, necessarily symmetrical or equivalent, insofar as it is modulated in energy according to the dosimetric needs and the shape of the target. We must therefore consider here the asymmetry of the irradiation energies of the two synchronous beams, as a possibility of modula- tion. intensity.

And to meet the requirement of greater precision than that achieved by
EPIDs this device incorporates within it a scanner C-150 Imatron Fastrac, electron beam. This set is characterized in that it defines a space of irradiation at the same time as that of image, whose system of acquisition of. Data (DAS) is itself connected to a real-time dosimetry console. We can not imagine what is real, we observe it with diagnostic imaging means of real control of excellence, which are also proven by long experience in Radiodiagnostics. This is likely to reconcile us with the true radiological semiotics, where the real holds a very preponderant place. Now, the phi-. The philosophy of the morphometric SIE gives pride of place to the composition of the image, from the various ingredients that are as clever as each other. The margin of safety to be taken in any volume being here negative, ie set back relative to the actual target volume where the electromagnetic interaction between the two convergent beams of therapeutic irradiation occurs.

. A whole revolution, which holds in so little thing already described in the patent
01/01133, whose diagrams are not included in this description but the references that relate to it!
It is characterized in that the delineation of the two beams of irradiation is made from the two sources, housed in two respective enclosures 22 and 27, respective ends of the two acceleration guides of the Bilinac Imatron, in

<Desc / Clms Page number 164>

 which, as we have just seen at their terminations, can distinguish, according to the embodiment, the sheath of the X-ray generator or, in another embodiment, the head of the telescopic therapy apparatus, which stops the radiation in directions that are not under any circumstances used for therapeutic irradiation; the shutter, which is essential in the equipment of radiotherapy, makes it possible to stop the irradiation without interrupting the emission and is also used in certain generators of X-rays; a pair 24 and 26 of multi-leaf dynamic collimators, embedded in the irradiation window 44, and with mirror symmetry of the fields forming a channel of variable dimensions, as a function of the fields (fixed, mobile, mixed, etc.) and the irradiation technique (conformational or not), defining the shape and the dimensions of the beam. A patient support is fixed inside a conduit of the electron beam scanner, for the purpose of producing at any time and in particular during a radiotherapy session of a computer-assisted dynamic tomography. .

 A three-dimensional image of diagnostic quality is generated in three dimensions in a diagnostic image space and stored in a memory.

 The patient receives at the same time a double beam of synchronous megavoltage, fixed fields or arcs or a combination of both, from the irradiation device called Ri-linac, whose two heads 22 and 27 are arranged on both windows lateral members 44 of the stand opposite and aligned in coincidence with the central axes of the irradiation beams and with at least one of the scanographic sectional planes of the diagnostic image space. FIG. 5 illustrates the different possibilities offered by the synchronous double beam device: fixed fields or arcs or simultaneous fixed-field combination. The strong electromagnetic interactions are, in the case of the two simultaneous beams, those which occur in the frontal collisions of the radiating particles and the weak ones are those which occur between two convergent beams whose axes form an acute angle; while between the two an obtuse angle gives intermediary interactions. The object of interest here being the target volume (VC), whose representation with all its tissue environment is obtained, such as on an ordinary scanner device, by dynamic imaging in real time.

The perfect execution, thanks to exact calculations of treatment planning, of a technique has never been guarantor of the effectiveness of the result. That's why Bilinac
Imatron can interactively produce from this 3-D acquisition, three-dimensional task-dependent images, such as in WO 00/04830

<Desc / Clms Page number 165>

 Wake Forest University (US), characterized in that a first and a second series of projected images of a selected object are recorded in a first and a second projection plane. The first and second series of projected images are rendered at a common enlargement, and then integrated into first and second three-dimensional volumes. The three-dimensional representation of the object is then produced by combining a projected image from the first set of projected images and an image from the second set of projected images, or by merging the first and second volumes. dimensional. From there, it is easy to call on the same monitor images made during the treatment planning; those relating to the optimization of the ballistics by the virtual simulation as well as those relating to the session of irradiation in progress and thus to check their correspondence and their perfect coincidence, knowing that the volume tumoral, the volume target, the various fields and their respective beams are materialized and allow to control the proper installation of the patient on the examination table and irradiation. Radiotherapist and Radiophysicist can thus communicate online at the patient's bedside, not only before the therapeutic choice is made, but also during the execution of the patient until its conclusion with the maximum of efficiency. The only subjective subjective inter-fractional clinical examination of the patient, which, in the present art, remains in the process of being treated roughly in the case of deep tumors, and must be completed or even replaced within the framework of the present invention by a 3-D objective and dynamic CT scan.

Several methods have also been developed for the dosimetry of asymmetric radiation fields formed by moving jaws independently of the collimator. Based on different principles, modified to conform to a series of data from Tsalafoutas et al. (2000), three of these methods are used for asymmetric field dose profiles. These three methods use flow rate factors and percentages of the depth doses obtained during the trans-action with the crossed tissues or even maximum tissue ratios of the opposite symmetric fields. In the first modality, the off-center ratio (OCR) calculation of the asymmetric field from which the asymmetry is deduced, by installing one of the jaws in an asymmetric position.

In the second modality, the OCR of the symmetric field is used for the OCR calculation of the asymmetric field of the same size; while the third modality does not allow the asymmetric calculation of OCR. The results obtained, using the

<Desc / Clms Page number 166>

trans-action data for the 6 MV photon beam of a Philips SL-20 linear accelerator, indicate that both the first and the second modality can, from the symmetric field data, accurately reproduce the field profiles asymmetrical; the third does not allow the reproduction of penumbra, but it is at the exact central part of the asymmetric field. The problems encountered and taken into account in the application of the three methods are reported and, in this analysis quoted above from Tsalafoutas et al. (2000), their accuracy is compared for single beam irradiation. The same experiments carried out, in the context of electromagnetic interaction of the two synchronous beams of phantom irradiation, would give a deposit of the higher dose of a certain percentage in 15% to 30% in the tissues in the state of the current art as well as a different in-depth performance, from the deep attenuation of current art. The exact measurement of the darkness of the radiation beam is essential to conventional conformal radiotherapy but
Stereoradiotherapy alone accounts for the knowledge of the extent of diffusion peak in the tissue medium as a function of the respective incidences of the electromagnetically interacting beams. For this purpose a detailed knowledge of the spatial response of the dosimeter is required in the ancient art, where the determination of the spatial response of the detector is however troublesome and is restricted to the type of specific detector and the spectrum of the beam used. This is why a model was developed by van't Veld et al. (2000) to calculate two dose profiles in the beam gap geometry as well as the detector response profiles. The summations of the field profiles of the spectra are considered representative of the photon beam for the polyenergetic beams. The model described in the van't Veld article and the resulting verified dose profiles combine with the multiple scattering theory of small Fermi angles.
Eyges and the transport of functions at the limit of electron transport the diffusion
Compton incident photons, the resulting electron transport. This analytical model thus gives the kernels of the diffusion line of the primary dose in. a phantom in water and can, with some adjustments, be perfectly transposed to the diffusion of a synchronous double electromagnetic interaction Bilinac-Imatron, knowing that the margin of safety is in the latter case negative. It has been shown that the spatial response of an ideal detector at a point of a primary photon beam can be well described by this model, the calculations are. verified by measurements with a diamond detector, in a telescopic geometry

<Desc / Clms Page number 167>

 peck, in which all dose distributions excepted for the primary dose may be excluded. Effects of photon detector behavior, linear accelerator source size (linac), and detector size are investigated in the same van't Veld article (2000). Measurements show that the dose profiles of the crack calculated using the kernel are accurate to. within an interval of 0.1 mm from full width to mid-maximum. The model is considered adequate for calculating local dose effects, which are dominated by the approximately single-directional fluence of primary photons and the diffusion they cause. Knowing that the analytic model provides more directional electron fluence information and is designed to be applied to various detectors and spectra of the trans-action beam of a single-beam Veld Veld linear accelerator, we generalize this concept. multi-beam technique in which the radiating particles collide in the irradiated target medium.

In standard teletherapy, a treatment plan is generated through a treatment planning system, but it is common to run a verification calculation on an independent monitor unit (MUVC = Monitor Unit verification Calculation). In exact analogy, Kung's paper (2000) proposes and demonstrates that a simple and accurate verification by the monitor unit (MUVC) is possible in modulated intensity radiotherapy (IMRT), introducing at the same time the concept Clarkson Modified Integration Program (MCI). To simplify the dose calculation, the modified Clarkson integration (MCI) is considered to have rotational diffusion symmetry, and for computation along a central axis (CAX), the incident fluence of IMRT is first replaced by an averaged azimuthal fluence. The integration of Clarkson is carried out secondly on annular sectors instead of sectors on pie. Doses calculated with the Clarkson Modified Integration Algorithm (MCI) are within a range of 3% in agreement with the doses calculated by CORVUS, which uses the trans-action of a beam in a brush. 1 cm x 1 cm, in the calculation of the dose (JH Kung et al, 2000). The latter method is characterized in that another quality assurance procedure in IMRT and IMAT or the combination of both, that of the patient's specific dosimetry and a verification calculation (MUVC) are performed on a console ad hoc. The monitoring unit verification calculation (MUVC) includes the scattered dose, the MLC transmission, and an offset of therapeutic irradiation fields.

<Desc / Clms Page number 168>

5.5.1.2. - 3-D conformation of the dose
The primary goal of three-dimensional conformal radiotherapy is to increase the degree to which the delivered dose conforms to the target volume while reducing the dose to the surrounding healthy tissue. It can be considered, in many cases, as a natural evolutionary development emerging from previous efforts; but, remains a theoretical concept not yet completely realized in the technology of the current art. Otherwise the intensity modulation seems to be an effective method of conformation of the dose. The simplest form of modulation of intensity is, perhaps, the introduction of a static wedge filter, which modulates the fluence in one direction. Two-dimensional intensity modulation has traditionally been practiced using holus blocks and retractable compensators. The development of computer-controlled advanced linear accelerators and multi-blade dynamic collimators has placed intensity modulation as a modality in its own right, at the emergence stage. Make . reacting between them, in a strongly symmetrical or asymmetrical manner, two megavoltage synchronous radiations, ranging from 10 MV to 50 MV, constitute the new form of modulation in time and in the intensity space of the beams, which is added to the modulation based on multi-blade collimation, resounding by electromagnetic interaction certainly on the intratis- deposition. sutary energy. And hence the need for a new extemporaneous dosimetry adapted to this new form of in silu escalation of energy deposition, which could perfectly adapt to the prognosis related to histology and radioresistance by a perfectly controlled escalation of the dose. This process indeed requires the use, without major risk of complication, radiation energies more and more. higher than electromagnetic interaction can be achieved by the collisional scattering of radiating particles as well as the high simultaneous multi-beam deposition density of radiating energies for effective control at doses deemed optimal of the tumor process. This device is also able to support the clinical planning simulation or, during treatment, the ballistic optimization per-. procedure. The present invention multiplies, as we will see, exponentially this potential of high density radiation radiant energy deposited within the interaction volume. It also guarantees the accuracy of the simulation, or even the reproducibility of the acts on similar devices, that is to say having, according to the embodiments, the same specifications of the source tubes of the various devices in use.

<Desc / Clms Page number 169>

It is easily understood from this example that an estimator as defined in this new invention is all the better that it has a minimum bias and variance. Hence the specific properties required for such a multifunctional estimator able to reduce for the comfort of patients the time of measurements. The literature suggests that to improve the provision of information to radiotherapy patients, it is necessary to examine whether a given amount of written information has an effect on anxiety and patient satisfaction. When the step is carried out in a very short time and concentrates in a CAD system several operations, whose individual execution time is spread over time over several hours or even days requiring, at each operation, the recovery specific information by different operators. At the same time, there is a decrease in the number of opportunities for errors, as well as a decrease in the number of interventions. Providing information to the patient in a step-by-step manner has led to less anxiety, among radiotherapy patients undergoing simulation, related to treatment and greater patient satisfaction. The anxiety is therefore proportional to the time during which a lot of things happen when, during the time to do it, there is little information and that this varies according to the speakers. It is therefore necessary for the comfort of the patient absolutely to control the duration of the acts of Radiotherapy.

Less time will last better this will be worth it! 5.5.1.3. - Intervention Protocol More and more articles in the literature talk about interventions to reduce the share of medical error (to do this lack of gold standard) by an intervention, which is either a deferred imaging or a test laboratory intended to improve the diagnostic technology in question. Several approaches have been suggested to improve only the problems inherent to man. Some non-randomized studies were also evaluated for comparison of relevant data. In addition, the report by J. B. Owen and L. R.

 Coia presents the results of surveys (careful reviews, surveys, questionnaires) of facilities facing the challenges of care management and quality, which are useful in the practice of Radiation Oncology. These surveys collected data from the entire census of megavoltage radiotherapy facilities. Data includes equipment load, staff load, and patient load. The data presented show that most facilities in different parts of the United States are but not all adequately equipped in terms of very high energy processing machine,

<Desc / Clms Page number 170>

computer type of treatment planning, simulation, and quality assurance programs. The data also show the percentage change in new cases receiving radiation therapy and repeat patients as a percentage of new radiotherapy cases by census region. The data show not only trends in increasing patient load by staff type for academic, hospital-based, and private facilities in 1994; but also that academic facilities used more machine therapists than other facilities (JB Owen and LR Coia: The Changing Structure of Radiation Oncology: Implications for the
Era of Managed Care. Semin Radiat Oncol 1997 Apr; 7 (2): 108-113.).

5.2. - Mastery of the problem of the evolution in the time of the data and Exploration multi-phasic
Therefore, the present invention is characterized in that it proceeds rather so that after a series of preliminary scanographic sections, on which are estimated, in multi-phasic radiocine-scan exploration, the coordinates of the lesion or the target region as well as the time-determined amplitudes of their variability with respect to the rotational isocenter of the scanner, at each fraction and by means of the screen display of the CT scanner computer and the patient is moved from the distance and in the corresponding direction to bring, through a fine adjustment of the patient support according to the proper anatomical landmarks, the target lesion at the isocenter of rotation. Then, the size and location of the target lesion to be treated is determined manually, by silhouetting the target lesion enhanced in contrast on the display screen, or automatically by an embedded tissue characterization electronic system encompassing in 3D and Next . its mobility in time the lesion and the perilesional # deme in a geometric form, or by other conventional means. The present invention is characterized in that it calculates geometrically the lateral opening (x) and that of the required field of view, in the axis (z), for different tubes and in each orientation of the scanner in order to correctly irradiate the target lesion while turning on them. common orbit inside the x-ray tube stand, together with the movement of the table and the necessary tilting of the stand, without overflowing the electromagnetic interaction volume beyond the target lesion into healthy tissue surrounding areas, where only a simple energetic trans-action must occur.

 The computer can then control the processing process by adjusting the X-ray output, the field of view aperture of each x-ray tube, the position of the

<Desc / Clms Page number 171>

 compared to the CT sections containing the target lesion, and the stand rocker when required to perform the prescribed treatment at the target lesion area. Because these questions are essential, because they are at the center of all the techniques of preparation for treatment, that of simulation and dosimetry, preparatory to the treatment as well as that of the ballistic aiming and the verification of the issuance The dose we illustrate below using examples from the literature and whose daily practice is modified by the present invention will contribute.

5.5.2.1. - References and placement In placing the reference (during which the isocentric treatment is marked on the skin of the patient and the reference images are obtained from these benchmarks), the position of the Anatomy of the patient versus the isocenter would be for many institutions in accordance with the corresponding position in the CT treatment plan. It is customary in these institutions to mark in the simulator after the planning has been carried out, the establishment of the final beam. The final definition of the isocenter is based on visual inspection; it can therefore deviate from the isocenter intentionally on a CT scan. This error in setting up the simulator necessarily leads to a systematic error in the treatment of the patient. For the irradiation of the prostate (A. Bel et al., 1994) and that of the head and neck (F. Lohr et al, 1997), it was found that the errors of setting up the simulator were comparable to systematic errors in setting up the treatment unit. But this is without counting the physiological movements transmitted at a distance, such as the cardiovascular beat, intestinal peristalsis, etc., which are never evaluated in clinical practice. In these two cases, the bone structures relevant to the placement of the patient can certainly be clearly identified, if not slightly in contrast to the thoracic region. It should therefore be expected that, for lung cancer patients, errors in simulator placement will be at least equal to or greater than systematic set-up errors at the treatment unit. The errors of implementation in the simulator are however not well in the literature as to the accuracy of setting up the patients. In fact, among seven implementation studies to which reference is made here (D. Yan et al, 1997, J. van de Steene et al, 1998, L. Ekberg et al, 1998, MJ Samson et al, 1999; Halperin et al., 1999, JE Schewe et al., 1996, GC Bentel et al., 1997), there are six that explicitly define what type of reference frame of placement was used. Of these six studies, five used

<Desc / Clms Page number 172>

 the simulator images based on a conventional 2-D simulation (J. van de Steene et al., 1998, L. Ekberg et al., 1998, MJ Samson et al., 1999, Halperin et al., 1999, JE Schewe) ( eg, no record of digitally reconstructed radiographs [DRRs]) and only one study (GC Bentel et al, 1997) applied both the simulator and the DRRs (Hans CJ de Boer, John R. van Sômsen Koste et al., 2001). This review by H. C.J. Boer et al. Systematically identifies and evaluates published research on interventions to reduce medical errors, and recommendations to caregivers to reduce errors via 3-D protocols, using intervention means in the form of doctor's orders and other intervention protocols, computer assisted.

The errors in the implementation processing unit compared to the reference implementation always have a systematic component and a random component.

According to experience, the systematic component could be reduced by gantry imaging, which is not diagnostic imaging, and the proposed implementation corrections are based on an off-line decision protocol (A Bel et al, 1993, S. Shaley et al, 1995). Both components were evaluated by Hans C. J. de Boer et al. (2001) and results reported according to an off-line decision protocol. However, in multi-beam interaction radiotherapy implemented in the present invention the risks are such that approximations of this type, made without any diagnostic update of the target image, can not be used without the counterpart. an inevitable technical sanction: the complication and / or its corollary recurrence where the bias has not allowed, at each fraction and every minute of irradiation, to correctly target the same lesion to produce exactly the interaction electromagnetic research. But the fact remains that the computer input of the doctor's order, the simulation, the training programs of the team, analysis of the root cause, even the coding by bar and other systems relating to the error did not prove to be easily reducing the error. Their discoveries would in fact only be an incentive to adapt these very expensive and time-consuming methods and to draw attention to a question not yet solved in the prior art and which can in addition lead to the responsibility of caregivers. The present invention takes a resolutely different perspective: that of bringing the patient back into a central position of the entire radiotherapeutic equipment and of organizing the rest around him, rather of transferring him from one point to another. an infrastructure exploded, despite the facilities offered by the network organization

<Desc / Clms Page number 173>

 of PACS and as many opportunities, for each isolated act, of error rather than looking for practices based on the only anatomical evidence, more targeted and more focused, or unbiased image-guided estimation of the various irradiation parameters, which in doing so improves the safety of the patient rather than a quest, of something related to good conscience, in the form of intervention protocols to prevent specific errors.

5.5.2.2. - Measurement sampling fluctuations
The existence in the present art of misplaced measurement measurements of the implementation of the implementation measure does not allow estimators used to date to fully fulfill their objective and to be precise at the same time. Data is also extracted with study implementation and inappropriate sample sizes, with study population characteristics and quantitative components, quality components, insignificant; the definition itself of the errors does not know whether reducing the error was a primary or secondary result of the proposed intervention, nor the number of errors per arm of the comparative statistics. It is known, for example, that a single measurement is taken where only a series of repeated multi-phase radiocinecroscopic explorations with samples of different composition must be observed, within the same population of measurements of the parameters of therapeutic irradiation or therapeutic fraction. Moreover, samples of measurements are from one fraction to another of more or less biased composition, with each fraction, with each voluntary or involuntary movement, with each ray-induced melting or with each incremental increase. spontaneous evolution always different and can not lead, in the presence of sighting bias, to an estimate in sight. an increase in the probability of tumor control of the true value of the expected parameter.

 The methods of describing implementation errors published in the literature have for the most part adopted, for systematic and random errors, the definitions introduced by Bijhold et al (1993). The error of systematic placement of a patient is understood as the implementation error averaged over all the fractions of the dose. If the average of this error is denoted by u and its standard deviation (SD) by E (this is the average inter-patient variation of the set-up).

 The random error of a patient is the standard deviation (SD) of the set-up error from one fraction to another (this is the inter-fraction variation). It can be seen immediately that the censorship interval and all the new opportunities for error do not

<Desc / Clms Page number 174>

do not enter into account in the reasoning. Otherwise An average is still calculated to obtain the average of the error per range of likelihood for a population of patients, denoted by #. A specific type of set-up error (eg, a translation in a certain direction is char- acterized by the set (, #, #).) Also note that if lasers and the light field are considered, in a reference situation (the simulator or the CT scan) and set up the treatment together, as appropriate indicators of the isocenter, would be close to zero, but # can still deviate significantly from zero [due to inter-patient variations (eg, errors in placement and voluntary and involuntary movements of the patient, during the CT scan or simulation, due to the mobility of the skin and changes in weight of the body by weight loss or weight gain, do not enter either) and especially no existence of biases.For measuring the errors of the establishment, a system of coordinates fixed is recommended by the I protocol nternational Electrotechnical
Commission (IEC 61217). In this system, the y axis is perpendicular to the ground, the z axis is parallel to the ground and the x axis is perpendicular to the first two axes.

5.5.2.3. - Intervals of censoring the measured data
JP Sy and JM Taylor produced a stimulating and thoughtful review re-analyzing one. extensive compilation of published clinical data on the late effects of radiation over time. It seems appropriate in this context to carefully consider the clinical and scientific implications of the present invention and to try to see it in a slightly broader context. Given the existence of possible bias of aim which can lead to differentiation, in the population. treated patients, those who are immediately likely, because of a purpose more next to the lesion than in the lesion, to reoffend and those who are not because the impeccable aim was essentially the lesion force to integrate the notion of censorship in practice. The most likely hypothesis is that some data on the failure time of literature come from a population. patients that is made up of certain topics that are susceptible and others that are not susceptible to the event of interest. In addition, the data studied have, at the end of the follow-up period, usually a censoring weight, and a standard survival analysis might not always be appropriate, if interval censorship is not taken into account. This is why it is important to give equal opportunities to patients by avoiding, at all costs, the high quality of ballistic sighting.

<Desc / Clms Page number 175>

 the biases. In such situations where there is good scientific or empirical evidence of an unaffected population, the mixture or cure model extended by Farawell (Biometrics 38, 1982: 1041-1046.) Used. Kuk and Chen (1992, Biometrika 79: 531-541.) May otherwise resolve the censorship interval question, using proportional risk regression of Cox for latency. This is why JP Sy and JM Taylor have developed maximum likelihood techniques for joint estimation, in this model using not only the nonparametric form of the likelihood and the EM algorithm, but also regression parameters. incidence and latency. It is the inverse of the observed information matrix which is then used to calculate the standard errors (JP Sy, and JM Taylor: Estimation in a Cox proportional hazards cure model Biometrics 2000 Mar; 56 [1]: 227 -36.) Inherent in the fractionation of treatment and the time interval between fractions.

An important clinical contribution to this discussion is also the analysis of the time factor made by W. Pan, in his article, A two-sample test with internai censored data via multiple imputation (Stat Med 2000 Jan 15; [1]: 1 -11), in which he retrospectively reviewed the medical records and showed that interval censored data arising naturally in large-scale panel studies where subjects can only be monitored, such as during radiotherapy fractionation, periodically and the event of interest can only be recorded as occurring between two exam times. It would be, in this case, the inter-fractions of the X-ray treatment. We must then take into account the question of the comparison of the two interval-censored samples. It proposes to impute the exact failure times from the interval-censored observations to obtain censored data on the right, then to apply the existing techniques, such as the Harrington and Fleming tests to the imputed data censored on the right. . For proper consideration, a Bayesian bootstrap (ABB) based multiple imputation algorithm is discussed by this author. The advantage of its proposal is the simplicity of its implementation and its adaptability to many incorporate many techniques of comparison of the two existing samples for data censored right but difficult to implement in daily practice in radiation therapy. The method is illustrated by the re-analysis of the data series of the Breast Cosmesis Study and shows that we must be very circumspect, since we are dealing with censored data (W. Pan: A two -sample test with internai censored data via multiple impu-

<Desc / Clms Page number 176>

 tation. Stat Med 2000 Jan 15; 19 (1): 1-11.). This is however the case of interval-censored CT scan data and processing planning contours, that of the virtual simulation data, etc., which do not undergo any pre-comparison treatment. This naturally leads us to define, in the context of the present invention, a good estimator of all the elements involved in a radiotherapy strategy and to minimize the importance of interval censoring.

5. 5.3. - Virtuality and Simulation in Radiotherapy Strategy Since it is a question of doing also with the device of the present invention of pre- and intraoperative clinical simulation and since virtual reality is about to catch up with fiction, it can therefore that it is about the planning of the treatment to make us enter real mirages, like the current art. So there are limits to virtuality. Although the military has been interested in the concept of virtuality since the early days of computer simulation (the first flight simulation programs were provided to NASA by General Electric in 1970), the first to really capture the immense scope of virtuality is the American Jaron Lanier. Among, for example, the various thoracic anatomical structures and the pathologies that affect them, there are now few entities that can elude the combined potential of computed tomography (CT) with multidetector lines and three-dimensional volume rendering (3D ) is very useful. Technological advances in this area have created with 4-D the ability to rapidly acquire unprecedented high-quality multiphase, or cineccanographic, exploration data and the potential for real-time 3D visualization and visualization. Powerful way is huge. In vascular imaging, for example, this technique provides a diagnostic image quality that matches or surpasses conventional angiography. Its use has expanded to aid diagnosis and surgical planning, often bypassing conventional or digitized angiography and reducing the costs of the procedure. It is reliable in describing clot and pulmonary vasculature, and can also be used to map an organ in its tissue environment and thus plan for X-ray therapy.

Multi-detector row CT has increased the conventional roles of computed tomography with 3D volume rendering and challenged the supremacy of other imaging modalities. In many cases, although axial images may be appropriate for diagnosis, 3D images in volume rendering may provide more information about the pathological nature and

<Desc / Clms Page number 177>

better communication with clinicians. It will henceforth play a dominant role with the advent in future research and radiological practice of 4-D polyconic CT. We discuss later some examples of pathologies where this role has already been asserted forcefully. The present invention has applications in almost all the locations treated, in radiotherapy, by external beam:
As an example for thoracic anatomical structures, we will cite the experience of the harmonization gauge used by MJ Samson et al. (1999) to analyze in an article in detail the intra-fractional movement of. anatomical structures visible on the portal images AP Multiple portal images per treatment fraction were obtained, contoured, and harmonized with the first gantry image of the corresponding fraction. Relevant structures that showed a small movement were the chest wall and the trachea. For the chest wall, the SD means intra-fraction movement in. the X and Y directions were respectively 0.8 and 0.6 mm, and for the trachea respectively 0.8 and 1.7 mm. The vertebrae, the sternum and occasionally the trachea had been leveled for the lateral fields, expecting the intra-fractional movement of these structures to be negligible, and a pilot study was carried out by the same authors to verify this expectation. (Hans CJ de Boer, John R. van Sornsen of Koste et al, 2001). This is a field where the present invention advances arguments that are more relevant for this type of estimation, where the only remote virtuality of the clinic can bring nothing more.

 As for the implementation errors in the simulator of the art, they have for example been evaluated by Hans C.J. de Boer et al. (2001) for 39 cancer patients. Small-cell lung (NSCLC) undergoing a planning CT, digitally reconstructed radiographs (DRRs), calculated for anteroposterior and lateral bundles. The frontal and lateral laser lines were CT-marked by ink lines (21 patients) or tattoos (18 patients) as well as by a longitudinal lead tape, which was with as many opportunities. error clearly visible in CT section. The patients were examined and treated without any immobilization device, in the supine position, except for one arm resting at the top of the table, and the other arm was elevated and the feet were at rest. CT cut thickness and spacing were 5 mm. The CT data was further transferred to the Varian-DoseTek CadPlan treatment planning system (TPS) to construct an isocentric 3-D plane (Hans C. J. de Boer, John R. van

<Desc / Clms Page number 178>

 Sômsen of Koste et al, 2001). In retrospect, the visible position of the anatomy with respect to the planned isocentre was compared with the corresponding position on the digitalized radiographs of the simulator, using a contour harmonization software later on.

 The deferred accuracy of the reference image of the establishment to the treatment unit compared to the implementation in the simulator was in this study measured with a new opportunity of error using a electronic portal imaging device (EPID), which is more clinical but not diagnostic enough, in 40 patients for at least 5 fractions of the two non-simultaneous orthogonal beams per patient. The set-up corrections, based on a simple off-line decision protocol, were applied with parameters derived from the knowledge of the studied patients. The standard deviations (SD) of the random CT planning implementation errors in the respective lateral, longitudinal, and anteroposterior directions were estimated at 4.0; 2.8; and 2.5 mm. The standard deviation (SD) of rotations around the anteroposterior axis was 1.6 and 1.3 around the left-right axis. The implementation error at the treatment unit was evaluated with a small component randomly in all three directions (1 SD = 2 mm). The systematic components were, particularly in longitudinal direction (1 SD = 3.6 mm), larger but they were reduced with the decision protocol at 1 SD = 2 mm with an average of 0.6 correction of placement per patient (Hans CJ de Boer, John R. van Sômsen of Koste et al, 2001). Implementation errors, which become systematic errors in the simulator were comparable, when the reference implementation is defined by simulation, in the case where no off-line protocol would have been applied to the systematic errors of setting up. , to the processing unit.

 Knowing that this type of intervention is very uncertain and is also an opportunity for error. From there, Hans C.J. de Boer et al. (2001) still infer that omitting a separate simulation step during processing could reduce systematic errors as efficiently as applying an off-line correction protocol. Random errors were thus small enough to make an off-line protocol feasible. However, we must relativize these results by saying that these errors do not take into account possible cumulative biases or by the various successive tests nor by accumulating opportunities for error, or by censoring these data.

<Desc / Clms Page number 179>

5.5.3.1. - Image surveillance
The resulting entrance gates of the treatment were simulated and the radiation oncologist judged the placement accuracy of the field based on the simulator images, on the beam's-eye-view (BEV) fields of view and standard CT (scout) topograms (which suffer from incorrect beam divergence in the transverse plane and non-divergence in the super-inferior direction). Boer et al did not use any DRRs during the simulation, as the objective of the study was to explore the accuracy of the simulation procedure used, which did not involve the use of DRRs. At the beginning of the simulation, cha-. that patient was positioned by CT markings, using the simulator lasers.

 The next step was to execute, according to the final position of the isocenter of the treatment plan, a (precomputed) translation of the table. Radiographs obtained in this situation often led to additional adjustments indicated by the clinician of the table and subjective visual comparison, based on a new one. opportunity for reading error, with the BEV plots. Once the placement was deemed satisfactory, the final position of the isocenter was marked on the patient using skin tattoos, long laser lines, and baseline placement radiographs were obtained at 0. and 90 or 270 angles of the stand (NB: these angles may differ from the angles of the treatment beam). (Hans C. J. de Boer, John R. van Sornsen of Koste et al, 2001), knowing that the outer contour is also subject to some mobility.

 The DRRs were retrospectively constructed from the 3-D planning data, at the same stand angles, for an intervention procedure.

 They were generated with homemade software from TPS, which benefited. from the diagnostic point of view of a higher spatial resolution than the DRRs. It has been verified that the TPS DRRs coincide geometrically with the DRRs used in this analysis (difference <0.5 mm). This difference in placement between the DRRs and the simulator images was measured, using the anatomical harmonization software and the gauging software described above. These errors. implemented in the simulator were analyzed by calculating the mean error and the standard deviation by translation direction and axis of rotation respectively producing systematic errors of (Rx, Ry, Rz, RAP, RLR) and (#Rx, # Ry, #Rz, #RAP, #RLR), in terms explained above (Hans CJ de Boer, John R. van Sôrnsen of Koste et al, 2001), ignoring neither the measurement error nor the the dependent operator error. The rotation around the z axis does not appear either

<Desc / Clms Page number 180>

 at first glance be a concern in this study. Thus, even in such a rigorous study, there is a certain subjectivity, as for 16 patients, the delineation of the contours in AP and lateral on the digitalized snapshots of the simulator as well as the gauging procedure harmonizing with the corresponding DRRs carried out by two observers. The SD is estimated from the analysis error for a single observer, SDobs, from the SD of the differences between the two observers (Hans C. J. de Boer, John R. van Sornsen of Koste et al, 2001). In the intervention process itself there are many opportunities for error insofar as internal mobility is not evaluated. It is even in the best centers like this one a possible source of ballistic aiming bias.

 5.5.3.2. - Approach to the issue of physiological motility It is therefore not surprising that despite a 65 Gy protocol, the complete local remission rate in patients with non-small-cell lung cancer cells (NSCLC = Non small cell Lung Cancer) is only about 20%, in the study of T. Le Chevalier et al. (1992). Unless irradiated volumes are reduced, radiation dose escalation may lead to unacceptable pulmonary and oesophageal toxicity under current irradiation conditions, particularly when concurrent chemotherapy is used (R.

 Byhardt, C. Scott, W. T. Sause, et al, 1998). This probably lacks proper image-guided dynamic aiming and the lack of knowledge due to ballistic aiming. Volume reduction can sometimes be achieved by omitting elective ganglion irradiation (JM Robertson, RK Ten Haken, MB Hazuba, et al, 1997, Hayman, MK Martel, RK Ten Haken, et al, 1999), but being also able to use in planning small safety margins. Such small margins would also frequently reduce toxicity to (usually) standard dose levels (Hans C. J. de Boer, John R. van Somsen of Koste et al, 2001). In line with the recommendations of ICRU Report 50, treatment planning also requires the exact definition of a Clinical Target Target (CTV), which must continuously encompass macroscopic tumor, subclinical disease, and probably involve lymphoganglionic relays. But as this target volume (VC) is for example estimated in a thorax mu by respiratory movements and transmitted cardiovascular beats remains a subject to caution, which the current art does not always have a satisfactory solution. However, the CTV requires to be further expanded to

<Desc / Clms Page number 181>

 means of certain geometric margins to a forecast target volume or planning target volume (PTV). These margins must ensure adequate CTV coverage during treatment and therefore be based on accurate knowledge of the range of motion, with respect to the treatment fields, the target and the accuracy of the machine (the eg reproducibility of block positions) and not just the relevance of an intervention protocol. ICRU Report 50, however, does not provide clear recommendations on how to choose CTV-to-PTV margins? And each one takes, his liberties in relation to that. For example, Stroom et al (1999) derived a general method of calculation, based on the probability of CTV dose coverage to derive a 3-D margin from known set-up errors and internal organ movements. , which can have any probability of the density function. They found that a clear distinction must be made between systematic and random errors, in which the systematic error is the average error of setting the target volume, in all fractions for a certain given patient. and random errors are inter-fractional variations. Their calculations incorporated the presence of small rotations (1SD - 1). A very similar result was found by van Herk et al (2000), by an analytical calculation of the simplified situation at a spherical target volume in an ideally conformational homogeneous dose, two of the results of Hans C. J. de Boer et al. (2001) confirm the intuitive notion that, in establishing planning margins, systematic errors are more important than random errors, without taking into account the biases inherent in current technology that is not guided by a simultaneous image.

Most of the placement studies in patients with lung cancer, for example, are two-dimensional (2-D) (D, Yan et al, 1997, J. van de Steene et al, 1998, L. Ekberg et al. 1998, MJ Samson et al., 1999) where they do not appropriately separate and quantify, in terms of random and systematic components (R. Halperin 1999, LE Schewe et al., 1996, GC Bentel et al. , 1997), implementation errors. None of them implement simultaneous image-guided procedures. These studies usually belong to small numbers (<20) of patients (J. van de Steene et al., 1998, L. Ekberg et al.
1998; MJ Samson et al, 1999; R. Halperin et al, 1999; JE Schewe et al, 1996).

In addition, none of these studies take into account the implementation errors that are made in the simulator [which have proved to be of some importance for

<Desc / Clms Page number 182>

 other tumor sites (Bel et al, 1994, F. Lohr et al, 1997) or other unassessed opportunities for errors that inexorably add up to approximately estimated errors. The intervention in the form of an off-line 3-D correction protocol for systematic errors during treatment can not have the desired impact and leaves something to be desired. As noted above, knowledge of systematic and random misstatements (and how much they can be reduced) is a prerogative for calculating optimal planning margins. This is how Hans C. J. de Boer et al. tried to measure the extent of systematic errors in simulator placement and treatment with this type of intervention as well as errors in random placement of 3-D treatment in 40 lung cancer patients.

De Boer et al. also report a reduction, through an off-line verification protocol of implementation, of systematic treatment errors.

The opportunities for error are, in this last report, too numerous and diversified to exhaust the question of the precise evaluation of the accumulation of errors. For example, the studies did not evaluate the errors in estimating the extent of the lesion and its inclusion in the target volume, as well as those attributable to prescribing and measuring the effects of radiation, previously time of diagnosis and during treatment, management of dose delivery and instant recognition of the relative severity of the disease and its evolution over time. Bias is caused by blind ballistic targeting without any instantaneous visualization of the target. Geographical misfires or tumor targeting errors are not only caused by external errors of placement but also, according to the literature (L. Ekberg et al, 1998, CS Ross et al, 1990) and at the same time by the unappreciated internal movements of the target volume. The study of H.C.J. Boer et al., Focused on external errors, for a group of lung cancer patients who had CT planning, set-up consists of: (a) errors made using the simulator snapshots for the definition of the implementation of reference (A. Bel et al, 1994;

Lohr et al., 1997) throughout the treatment, and (b) errors in placement at the treatment unit compared to the reference set-up (Hans CJ de Boer, John R. van Sômsen of Koste et al, 2001). In these studies the intervention did not necessarily reduce the frequency of errors and it is not even possible to know if the frequency of errors was increasing or if there had been a reduction in errors compared to proposed intervention. But, the waiting time for the sick

<Desc / Clms Page number 183>

 increased and staff days were lengthened.

 5.5.3.3. Clinical applications of the device of the present invention Conformational radiotherapy already appears, in the therapeutic approach of patients with brain tumors, as undeniable progress. Continuation of the clinical work, notably on the determination of the macroscopic tumor volume, that of the anatomoclinical volume for the gliomas and on the interest of the increase of the total doses, however remain necessary. Data on the toxicity of healthy tissues are better documented.

A comparison of the information provided by the dose-volume histograms with the results of neuropsychological evaluations may be considered. Finally, the development of beam intensity modulation will lead to comparative studies with proton therapy or stereotaxic radiotherapy, for small volume tumors. These are the issues that the present invention is optimistic about.

 # 5.5.3.3.1. -Applications at the level of the Skull The good clinical results of the radiosurgical treatments of the Brain Tumors could make think that the present invention would have no place in this localization. Radiation is the most common treatment modality for brain and skull cancer. Although commonly used, radiation delivery processing techniques have changed substantially. One of the most important modifications in the implementation of the rays is the widespread application of stereotaxic techniques and their acceptance in the course of the dominant radiotherapeutic delivery. The distinctive features of stereotactic radiosurgery and its current and future applications are important for all physicians to understand. The article by JM Buatti, SL Meeks, WA Friedman, and FJ Bova (Stereotectic Radiosurgery: Techniques and Clinical Applications, Surg On Clin Clin N Am 2000 Jul; 9 (3): 469-87, viii) discusses these treatment techniques. and applications from the perspectives of a surgical Oncologist.

More specifically, in comparison with a radiation therapy gamma knife apparatus, the present invention provides for radiation purposes a more flexible geometry, and thus an irradiated target volume size and more flexible shape. Moreover, since the present invention is constructed in combination with 4-D computer-aided tomography radioscanners, it necessarily has a cost and a weight.

<Desc / Clms Page number 184>

 below the gamma knife, whose irradiation sources are all stationary.

In contrast to the gamma knife, the present invention offers in addition to mixed cycling an ability to perform diagnostic imaging at lower X-ray energies than those used in the present invention for orthovoltage with its complete scanner unit. 4-D that is capable not only of contrast enhancement, after injection i. v. in iodine bolus, but also tissue perfusion images and vascular mapping of the target region. Also, patient imaging can take place contemporaneously with the radiotherapy procedure, thus allowing accurate 3-D verification in real time of the perfect coincidence of the target region chosen for therapeutic irradiation with the lesion itself. . Thus the present invention provides, as a radiotherapy unit, the more accurate locating means of the region to be irradiated such as in the use of the prior art gamma knife. The present invention also requires no stereotaxic apparatus for positioning more accurate than gamma knife but a thermoformed compression for the reproducibility of the position of the head and neck. Finally, the present invention utilizes a lower X-ray energy level than the 60Co y-energy which thereby allows higher RBE, thus minimizing the requirement for a higher required total dose. for the success of the radiation treatment and improving the dose profile due to the continuous rotation and the increased number of fields of radiation provided simultaneously by the use of a scanner unit with its rotating gantry which does not exist in Currently used in the gamma knife for radiation therapy procedures, or even with high dose escalation potential without major collateral damage, including single fraction.

In another study on skull irradiation, three different techniques for supernatant delivery of posterior fossa (posterior fossa boost) in patients with medulloblastoma were compared. Five patients underwent CT simulation for the planning of the supertreated treatment of the posterior fossa. For each slice, the posterior fossa was, in addition to the cochlea, the brain of the non-posterior fossa, the pituitary gland, the mandible, the parotid glands, the thyroid gland, the pharynx, and the cervical spinal cord. , contoured. The plans for the three techniques for delivery of the supercharged posterior fossa were compared, for each patient. Technique A used opposite parallel lateral fields with bone markers (2-dimensional radiotherapy); the other two techniques were planned using radiotherapy

<Desc / Clms Page number 185>

 three-dimensional. Technique B used a pair of coplanar filtered posterior oblique bundles, while Technique C finally used a pair of oblique posterior fields and a vertex field. The dose-volume histograms (DVHs) were obtained for each of the contoured organs as well as for each technique and each patient. The maximum dose, the minimum dose, and the average dose of each organ were determined using, in this treatment planning system proposed by AC Paulino, A. Narayana, MN Mohideen, and S. Jeswani (Posterior fossa boost in medulloblastoma Radiotherapy Oncol Biol Phys 2000 Jan 15; 46 (2): 281-6.) The DVH program, with fixed bone landmarks needed for reproducibility ballistics.

In 3 of the 5 patients in the study, the target planning volume (PTV) was not with technique A included in the treatment field. The cochlea received 100%, 50%, and 42% of the prescribed dose at the posterior fossa respectively, using techniques A, B, and C. The mean non-posterior fossil cerebral dose was higher with Technique C, intermediate with technique A, and weaker for technique B. The average brain dose of the non-posterior fossa with technique B was comparable to the average brain dose of the non-posterior fossa, delivered using parallel lateral fields opposites based on the CT definition of PTV. The mean dose to the pituitary gland was similarly lower in technique B. Both techniques B and C were associated with a higher average dose to the thyroid gland, mandible, parotid gland, and pharynx.

5.5.3.3.2. - Irradiation of lymph node areas, such as in the
HODGKIN 1. - Upper thoracic and mediastinal level This is about protecting the lungs; Until now, a lead coat was needed to do this. For female patients, radiotherapy treatment for Hodgkin's disease invariably results in undue irradiation of breast tissue that can lead to radiation-induced secondary cancers. The risk of developing secondary breast cancer is dose-related. The present invention allows an irradiation technique to increase during the irradiation by interaction of simultaneous bundles of fields in mantlet breast backup for female patients, thanks to its printer function for the manufacture of molded parts which

<Desc / Clms Page number 186>

 reproduce faithfully the technique in procubitus proposed by Vuong for example.

 To minimize the volume of irradiated breast, T. Vuong et al. (2000) propose a virtual simulation technique using a styrofoam breast immobilization board that was developed on which the patient is lying procubitus with the breasts positioned in gloves inside the board. Breast position is adjusted using Styrofoam filters, and breast placement is verified using an A-P view of CT-pilot. A series of scanographic sections of the neck and thorax are taken, and the lymph nodes, the. Breast volume and critical structures are silhouetted. The virtual simulation of the mantle fields (typically APIPA isocentric beams) is performed, and the beam blocks are drawn on digitally reconstructed radiographs (DRRs) generated by the virtual simulation package. The protection is designed to allow adequate margins. around the lymph nodes, while lowering the level of protection of lung and breast tissue. Para-aortic fields are easily determined by virtual simulation, where multiplicative reconstructions (MPR) and 3-D volume renditions of the patient's CT data are used to determine the field boundaries and holes of the patient. beam. To additionally allow the geometrical optimization of the placement of breasts under the pulmonary fields, the virtual simulation technique provides the information necessary for a 3-D dosimetry analysis, including, according to T. Vuong et al, (2000). ) the dose-volume histograms (DVHs) of the irradiated breast volume and in the context of the present invention the interaction diffusion resultant of the multiple simultaneous beams.

. The 3-D breast rescue technique of T. Vuong et al. which seemed relevant to us and was qualitatively and quantitatively compared to non-CT-based techniques and other 3-D techniques commonly available to assess breast protection. In a preliminary analysis, the images of the virtual simulation (DRRs, rendering 3-Det multiplanar reconstruction) demonstrated the advantage of using the technique of breast backup. Further analysis of DVHs, to which we do not give much credit, showed a reduction of at least 50% in the volume of irradiated breast tissue when using the breast placement board and virtual simulation in comparison with conventional simulation techniques where a breast immobilization board was not used. The use of an immobilization board

<Desc / Clms Page number 187>

 Intra and virtual simulation technique is recommended for the planning and treatment of patients with Hodgkin's disease. DVH analysis has shown that this approach will reduce the risk of secondary breast malignancies in patients with Hodgkin's disease (Vuong et al, 2000); but that of irradiation of the lungs and reduce the dose to the heart at the same time, and all the more so with dynamic X-ray imaging.

 2. - At the lympho-gaglionary level of the abdominopelvic In the prophylactic irradiation of the infradiaphragmatic lymph nodes (LN), the width of the para-aortic and pelvic fields is given in the rules of art by a line joining the points of the apophyses transverse of the 1st 1st thoracic vertebra to the fourth lumbar vertebra. Thus the edge of the field follows a straight line at the most lateral point of the acetabulum observed on the simulation plate. It is therefore necessary to have in this way constant anatomical landmarks. Another method is to set up the most peripheral point of the field and project the lateral contours of the large abdominal vessels from T (1) - weighted by coronal MR images of the abdomen on the simulator radiographs and add a margin of 2 cm along the vessels thus delimited.

Nevinny-Stickel et al. have shown, however, that lymph node measurements are both 1) lateral to the transverse processes of the thoracic and lumbar vertebrae and 2) 2 cm outside the margin of safety from the lateral contour of the abdominal vessels. . These data clearly show that the traditional fields of irradiation of infra-diaphragmatic lymph nodes do not appear to have been broad enough to encompass with certainty almost all retroperitoneal and pelvic lymph nodes. They recommend an expansion of the fields. (M. Nevinny-Stickel, S. Ennemo-ser, R. Sweeney, I. Bangerl, D. Zur Nedden, and P. Lukas: Comparison of standard and individual planned infradiaphragmatic fields in irradiation of retroperitoneal lymph nodes Int J Radiat Oncol Biol Phys 2000 Aug 1; 48 (1): 147-51.). This therefore requires case-by-case exploration and increased collateral radiation toxicity. The printer function is responsible for making the mantle according to the exact position of the ganglia detected by CT with contrast injection. Better still, the irradiation, in the context of the new invention, is carried out by electronic mantle at the level of the thorax. La'table performing, as in a tomotherapy

<Desc / Clms Page number 188>

 back and forth on all lympho-ganglionic areas (abdominopelvic field).

5.5.3.3.3. - Irradiation of the breast
The technique of thoracic wall tangent radii discussed below, taking into account the homilateral axillary, subcalvicular and mammary internal lymphoganglionic regions, is not in question by the present invention. The rotating tubes are permanently extinguished, during the irradiation of the breast and the axillary lympho-ganglionic areas, but not during the irradiation of the internal mammary areas the stationary sources are oriented to irradiate by tangent fields and not. necessarily parallel and opposite. The interaction region and the isocenter are returned to the anterior mediastinum and the sternum in internal mammary lymphoganglion irradiation.

 Find an irradiation technique for loco-regional irradiation of breast cancer patients who, compared to a standard technique, improves distribution. dose adjustment to internal mammary lymph nodes - medial supraclavicular (1M-MS = Mammary-Medial supraclavicular interna), such as, for example, the improved technique by C. W. Hurkmans et al. (2000) is intended to minimize the lung dose and to reduce the dose to the heart can not be conceived outside the control of the parameters of thoracic motility to which has become particu- larly important. attached the present invention. The standard technique consists of an anterior field IM-MS electron / photon mixed. In the improved technique cited above, an oblique electron field and an asymmetric oblique photon field are combined to irradiate the lymph nodes. To irradiate the MS lymph nodes, a combination of anterior electron field and a field. anterior asymmetric photon is used. For both the standard and the improved technique, tangential photon fields are used to irradiate the breast. Three-dimensional (3-D) treatment planning was performed for 8 patients with varying breast sizes for both techniques. Dose-volume histograms (DVHs) and normal tissue complication probabilities (NTCPs = Normal Tissue Complication Probabilities) were compared for both techniques. The field dimensions and the energy of the standard technique were determined by the simulation, while for the improved technique the fields were designed by the dynamic CT image processing planning.

<Desc / Clms Page number 189>

The dose in the breast planning target volume was essentially the same for both techniques. For the improved technique, combined with the 3-D location information, an improvement in coverage of the planning target would have been observed. The volume within the 95% isodose surface was estimated at 25% (0 to 64%) and 74% (43-90%), respectively, on average for the standard technique and the improved technique. The heart would have generally received less dose with the improved technique. However, a small but acceptable increase in pulmonary dose is sometimes found, so much so that these estimates are not unbiased. Otherwise, the improved technique, combined with IM-MS lymph node localization information, greatly improves the dose distribution in the planning target volume for a large group of patients without significantly increasing the dose to the organs. risk.

(CW Hurkmans, AE Saarnak, BR Pieters, Borger JH, and LA Bruinvis: An improved technique for breast cancer irradiation including locoregional lymph nodes.) Int J Radiat Biol Onol Phys 2000 Jul 15; 47 (5): 1421-9. and it is in this sense that the present invention resolutely works to further improve the therapeutic ratio, by controlling the bias factors, including physiological motility and blind practice.

G. Baroni et al. report a technical and clinical validation of a real-time analytical control system of the patient's position on the radiotherapy simulation and treatment units, which to some extent applies the concept of image guidance on which relies on the present invention, in the irradiation of the breast and allows, if ever it could generalize the use of the 4D system of a high-performance polyconic multi-layer CT embedded on a radiotherapy apparatus, as in the present invention , to grasp the exact scope of this example of literature. The positioning control system uses motion analysis technology consisting of an optical component (a pair of TV cameras) and a unit for real-time image processing. The system can provide three-dimensional coordinates (3-D) in real time (up to 100 times in one second) of a series of passive markers (small plastic hemispheres, 5 mm 0) previously positioned on the patient and localized to inside the field of view of the TV camera system. The method for positioning quality control is based on the analytical comparison between the current positions of the series of markers placed on the patient's skin and a corresponding reference model, the last acquired at the end of the simulation procedure.

<Desc / Clms Page number 190>

or at the first therapeutic irradiation session. The system was used at the
Radiotherapy Division of the European Institute of Oncology for the analytical control of repositioning in three patients undergoing, after conservative surgery (siblingectomy) breast cancer, irradiation. This showed the clinical feasibility of the technology as well as that of the. method, and quantified the improvement in patient positioning compared to the current repositioning procedures. In particular, the possibility of evaluating the respiratory phases of. Patient could better distinguish the different factors (repositioning inaccuracies and cyclical and random movements of the patient) contributing to global location errors. The multi-phase 4-D exploration of the present invention makes it possible to overcome all these disadvantages even more precisely by means of a very precise statistical estimation.

 The method described above differs from the 4-D system and is independent of the geometric irradiation parameters and nevertheless would allow one in real time. efficient quantitative control of the specific repositioning quality and real immobility of the patient during the irradiation. It should be remembered that the improvement in the accuracy of conventional repositioning methods based on superficial laser alignment only shows a plausible performance of the optical centering procedure (with 3-D displacement <5 mm) only for markers placed near skin markers used for laser alignment. Otherwise, the markers inside the irradiation field but not directly controlled, during the repositioning procedure, displayed location errors> 5 mm; this often occurred in favor of inaccuracies in relation to the respiratory movements of the patient excluded from the analysis. Furthermore, '. the quantitative evaluation of global location errors confirmed the high influence of respiratory movements on repetition and position retention, which can only be accounted for by physiological exploration, such as cinecanography. In the opposite case, even markers that were averaged well repositioned would come out being significantly displaced during delivery of the radiation dose. It is in this context of aiming bias that the present invention has been devised which is a physiological exploration and an image-guided therapeutic irradiation to closely follow all these variabilities as a function of time.

 These results confirm that movement analysis based on optoelectronic techniques can play a role as a means of improving

<Desc / Clms Page number 191>

 patient positioning and assess immobility, crucial. Thus, the system makes it possible to quantify the target volume localization errors, which makes it possible to adopt suitable countermeasures to reduce the uncertainties, during the actual irradiation. This is a crucial requirement, especially when the complexity of the irradiation geometry (conformal radiotherapy) or the radiation radiation type (hadrontherapy) calls for optimal application of the plan at each actual irradiation session. simulated treatment.

# 1. - Importance of respiratory motion in breast irradiation
A study of healthy volunteer women by H. Lu et al. (2000). suggested that deep breath-free apnea could reduce heart rate at the entry level of left breast cancer treatment. The reduction in irradiated cardiac volume may be important in considering late morbidity and late cardiac mortality as well as the frequent coexisting use of potentially cardiotoxic chemotherapy in can patients. breast. In this study (HM Lu, Cash E., Chen MH, Chin L., WJ

Manning, J. Harris, and B. Bomstein: Reduction of Cardiac Volume in Left-Breast Treatment Fields by Respiratory Maneuvers: CT study. Int J Radiat Oncol Biol Phys
2000 Jul; 47 [4]: 895-904.) Was evaluated the volume of the heart in the fields and, as a result, the true benefit of this respiratory maneuver in patients with breast cancer who had undergone CT simulation. Fifteen patients (mean age 53 years) were studied. For each patient, CT sections were taken, when the patient was breathing normally (calm breathing) and when the patient held the breath after a deep breath. Tangential fields have been planned for normal breathing and apnea configurations using. the same medial, lateral, superior, and inferior edges on the skin. The left heart and lung volumes within the tangential fields were calculated for both respiratory configurations. A series of multiple sections were made in apnea configuration to provide more accurate delineation of cardiac tissue and to study the reproducibility of the patient's position between different patients. deep cycles of inspiration. This is perfectly achieved with less gymnastics for the patient multiphasic physiological exploration or 4-D cinescanography, for radioscanometry.

None of the patients reported difficulty in holding their breath for 20 seconds. Heart volume in the field was reduced (-86 ~ 24%;
0.001), when the patients held their breath, after a deep breath

<Desc / Clms Page number 192>

compared to when he was breathing normally. For 7 patients (47%), deep inspiration completely shifted the heart out of the radiation fields. Expansion of pulmonary tissue due to deep inspiration also increased absolute lung volume in tangential fields (183 cm3vs 97 cm3, p <
0.001). The resolution of this question is related in the present invention to image-guided respiratory synchronism. Overall, however, the fractional volume of the left lung in the field was essentially unchanged. For all but 1 patient, the maximum difference between the external body contours from different apnea cycles was 5 mm and took place in lateral view of the breast. In median view, as indicated by the position of the midline marker, the variations were well within the commonly accepted tolerance for placement, during the tangential treatment, of the patient. All this today would be worth doing in cinecanography. Apnea of deep inspiration substantially reduces the heart volume in the fields. tangential for the treatment of left lateral breast cancer. The variation between the patient's positions at different cycles of apnea has been shown to be reasonably small. Therefore, it seems feasible to reduce cardiac radiation by treating patients with intra-treatment minifractions for 10-15 sec while patients hold their breath. This is entirely. possible in respiratory synchronism, or even cardiorespiratory provided in the present invention.

2. - Importance of Accuracy of Beam Alignment in Breast Irradiation
With the increase of conservative breast therapy for cancer first. Stage, tangential breast irradiation is one of the most important treatment techniques in modern radiotherapy. The goal of tangential irradiation is to deliver a homogeneous dose of radiation to the entire mammary gland, with optimal safeguarding of normal tissues, particularly the lung. A factor that influences the homogeneity of the dose and the volume of treatment is the accuracy of the implementation of the treatment. Small deviations in patient positioning with respect to beam placement could have a relatively large impact on the volume of treatment, especially the amount of irradiated lung tissue (unnecessarily or even dose heterogeneity introduced into the breast tissue , to the extent that the aim is located more or less elsewhere, This dose will inevitably miss the tumor tissue at the end of the race). Therefore, it is

<Desc / Clms Page number 193>

It is important to control the placement and to control all parameters during radiotherapy (RW Byhardt el al, 1978, Kirby and Williams,
1991; MJ Richards and DA Buchler, 1977). And a certain role has been given to evaluation by the gantry cliché, which plays an important role in detecting the inadequate posi- tioning of the. patient or other geometric errors that may occur during treatment (B. Kihlen and B. Ruden, 1989, I. Rabinowitz et al,
1985), let alone the concept of image-guided therapy. In vivo dosimetry studies were conducted to test the validity of the dose calculation procedures in the actual treatment situation (R. Boellaard et al, 1998). Variations). Many studies have shown that movement during a fraction of radiotherapy does not play a major role (A. Fein et al., 1996, C. Westbrook et al, 1991 cited by 0. Pradier et al. al, 1999). This failure to have never evaluated as shown in Figure la for lack of adequate tool.

3. Patients and treatment technique
30 consecutive patients with breast cancer were included in this study.

 All patients with stage tumors between T1N0 and T2N2 were irradiated after lumpectomy and axillary dissection. All these tumors were adenocarcinomas on histology. Patients were treated using the fields). medial and lateral tangential, with an isocentric technique. Conventional planning was used but CT CT was used to assess lung volume. The plan could not be generated on a single CT cut because of the difference in arm position between the simulation and the CT scan. A rotation of the collimator was introduced to reduce the pulmonary volume. Compensating filters were used to avoid inhomogeneities in the dose distribution inherent in the irregular morphology of the breast. Patients were treated with the linear accelerator, Saturn 43 (6 MV X-rays). They were positioned with an arm support, in abduction of 90-110. A total dose of 45 Gy, in fractions of 1.8 Gy was delivered. An overprint of 15 Gy a). applied with electron beams or brachytherapy. The sagittal and lateral laser beams were used to position the fields. Simulation snapshots were taken and localization snapshots were daily exposed in order to verify the accuracy of the treatment. The measurement method used by O. Pradier et al. (1999) was derived from the work of van Tienhoven et al (1991). Five linear measurements were taken for each simulation and snapshot

<Desc / Clms Page number 194>

 of verification.

 In the latter, the central lung distance (CLD) represents the distance from the edge of the dorsomedian beam to the inner chest wall in the central plane of the beam. Cranial lung distance (CrLD = Cranial lung distance) represents the distance from the edge of the dorsomedian beam to the inner chest wall in the plane of the beam at 4 cm above the central plane. This value of 4 cm is fixed and measured for each entrance door. The edge of the central beam at the distance of the skin (CBESD = central beam edge to skin distance) is in the central plane of the beam. the distance from the skin to the edge of the ventrola-beam. Teral. The central irradiated width (CIW = Central irradiated width) is defined as the distance from the dorsomedian edge of the beam to the skin. The cranio-caudal distance (CCD = short distance) is defined as the distance from a particular landmark to the caudal periphery of the field. All this seems very tedious to start.

. Determining the accuracy of the patient's position in radiotherapy of breast cancer is not, in these conditions, a cure. To do this, portal images were obtained by using a rapid electronic imaging system of megavoltage in 30 cases. of breast cancer. The quantitative analysis of 530 gantry megavoltage images and the comparison with 30 digitized simulation snapshots was. performed. Five linear measurements proposed by O. Pradier et al (1999) were taken for each simulation and verification snapshot. The central distance from the lung (CLD = Central Lung distance) represents, in the central plane of the beam, the distance from the dorsomedial edge of the beam to the inner chest wall. The cranial lung distance (CrLD = Cranial lung distance) is, in the plane of the beam at 4 cm from the central plane, the distance from the dorsomedial edge of the beam to the inner chest wall. The distance from the central beam edge to the skin (CBESD) is, in the central plane of the beam, the distance from the skin to the ventrolateral edge of the beam. The irradiated central width (CIW =
Central irradiated width) is defined as the distance from the dorsomedial edge of the. beam to the skin. The cranio-caudal distance (CCD = craniocaudal distance) is finally defined as the distance from a particular reference point to the caudal edge of the field. With regard to the patient's position in the field, the average standard deviations of the difference between the simulation and treatment images were estimated at 3.9 mm for the CLD; at 3.2 mm to + 4 cm; 3.6 mm for the CIW; 3.3 mm for the CBESD; at 3.8 mm for the CCD. In 90% of errors

<Desc / Clms Page number 195>

 of all set-ups for treatment were found to be less than 1 cm. Variation in CLD was considered the greatest error of implementation.

This last parameter seems to be clinically the most significant. In total the delivery of future treatment should be improved by introducing patient positioning devices such as thermoplastic shells. The Electronic Portal Imaging Device (EPID) seems to be an adequate tool to study the accuracy of the placement of treatment like this. There is a great deal of arbitrariness in all these considerations. Breast irradiation plays an important role in the treatment of breast cancer after surgery, so it is urgent to adapt technique and technology to this expectation. The dose inhomogeneities that result in overdose to certain areas that may occur in the breast and underdosing in other areas since the implementation of the treatment differs from that prescribed in the treatment plan (J.

Das et al, 1993; J.R. Gray et al., 1991). The cosmetic effect has been shown to be more likely to be poor and pulmonary fibrosis is more frequently observed in older patients (J.R. Gray et al, 1991). This may be caused by greater inhomogeneity of the dose (L. M. Chin et al, 1989) in the breast tissue and a larger irradiated lung volume (O. Pradier et al, 1999). It is therefore necessary more delicacy than what is admitted today. If the irradiated site is well immobilized, we can observe a very good accuracy of the treatment, the best example being the immobilization of the head and the neck with the devices of compression, Orfit. Hess et al (1995) reported an error of less than 5 mm for more than 80% of fields. Comparable results were observed for irradiation of the brain and cranial meninges, when the head was immobilized with an Orfit compression device (RD Kortmann et al, 1995, cited by O. Pradier et al, 1999), but the repeats snapshots are missing at these different measures.

In routine work, field placement errors were quantitatively evaluated for a large homogeneous group of patients receiving standardized treatment. The accuracy of the field alignment was analyzed separately, day after day, for the transmission of the simulator to the treatment machine and the variations during the evolution of the treatment. The distribution of the cumulative frequency of our data showed that in 90% of the cases the errors of setting up were lower than 1 cm. Mean differences, mean standard deviation, and maximum deviation were analyzed and compared to results from other studies (Table 3). The average value of the difference between simulation and

<Desc / Clms Page number 196>

the daily port images are 1.88 mm for the CLD with a standard deviation of 3.95. The average difference is - 1.94 and - 0.63 for the parameters
CBESD and CCD (Table 2). The standard deviations are 3.31 and 3.8, respectively. For the CIW parameter, the deviation of our results is less than that of
CL Creutzberg et al (1993) or A. Lirette et al., (1995).

 Transfer errors inherent in implementation based on CT scan data and simulation can significantly affect the overall accuracy of the treatment, as shown in Figure la. The accuracy of the field alignments is influenced by various factors such as immobilization of the patient or. the treated organ, which must be evaluated exactly with the help of an appropriate tool. This is because the breast is a movable organ, and even when the isocenter is marked on the breast skin, the breast is regularly lifted by the respiratory amplitude of the chest wall. Breast mobility could influence the accuracy of the implementation of radiotherapy. In this connection a prospective study a. was undertaken by O. Pradier et al. (there is no need for a study to understand it) to measure the accuracy of field positioning in their establishment. Disagreements in setting up are considered slightly higher than those described by G. van Thienhoven et al. (1991) or A. Lirette et al (1995). Subjectivity is here of rigor. Inter-fractional disagreements were there. less than the difference between simulation and daily port pictures.

Nevertheless, the second important discovery in the study of O. Pradier et al. (1999) was that CLD, representing the volume of the irradiated lung, showed a greater degree of variability. This displacement is, since the volume of the irradiated pulmonary tissue is correlated with radicular pneumonia done in a particular way. according to SL Kwa et al. (1998). The reasons for moving the field in conservative radiotherapy could be either the unstable positioning of the patient or the difficulty of marking the isocenter on the breast. Both problems could be addressed by creating a breast fixation device, as proposed by
CL Creutzberg et al (1993). Patients were not immobilized with one. modified hemicorporeal cache (cradle hemibody customized) variation, a CLD variation of 3-5 mm in 26%, 5-10 mm in 25%, and> 10 mm in 6%.

 The very poor results, obtained compared for the same treatment location under study with the variation of CLD being during the spread of radiotherapy treatment of 5 mm in 35%, 5-10 mm in 25%, and> 10 mm in 10%. The data of O. Pradier et al. (1999) agree with the reported observations

<Desc / Clms Page number 197>

 previously. The highest variation in treatment implementation occurred with transmission from the simulator to the accelerator. Subsequently, variations during treatment delivery were lower. The change in central lung distance was the biggest mistake of placement. This parameter is considered clinically the most significant. The future delivery of treatment will need to be improved by introducing patient positioning devices such as thermoplastic restraints to maintain the reproducibility of the patient. positioning. (Behind such a device there may still be a decline in volume by muscle wasting or expansion by weight gain.

. The question can not be definitively closed in this way).

# 5.5. 3.3. 4. - Irradiation of the Prostate
The object of the present invention is to irradiate irrespective of the localization under conditions close to the interstitial irradiation of local implants without the disadvantages of the latter, ie nonuniform dose rates are inevitable. Prostate treatments involving, as is done in the art, internal sources, flow rates from, according to MA, Ebert and SF Zavgorodni (Med Phys 2000 Feb;
27 [2]: 393-400. ) effects of electronic imbalance as well as nonuniformity of activity distribution, L For the purpose of evaluating, using in patients undergoing radiotherapy for prostate carcinoma electronic portal imaging, l accuracy of daily setups. JA Stryker et al. (1999) used a liquid ion chamber to evaluate the accuracy of placement in 25 consecutive patients undergoing a 6-week course of prostate radiotherapy. Electronic images (EPIs) were collected during the 33 treatments at each of the four gateways. The positions of the anatomical structures on the EPIs were compared with the same structures observed on the digitally reconstructed radiographs (DRRs) made after the CT simulation before starting the radiotherapy. Displacements of the EPIs compared to the DRRs were i. calculated by computer in millimeters in the lateral, longitudinal and rotational directions for each entrance door. 11 patients had displaced entrance doors because of disagreements between PPEs and DRRs; required displacements in the first five treatments to correct systematic errors (simulator). In the right-left and anteroposterior directions, approximately 95% of the EPIs were within 5 mm of the position of the

<Desc / Clms Page number 198>

 simulated port. In the super-inferior direction, 98% of the entrance doors were within 5 mm of the position of the simulated door. Two patients had errors in the rotational plane on the lateral entrance doors (respectively 8 degrees and 10 degrees). It was concluded that daily electronic imaging would be an effective technique for evaluating the accuracy of the set-up, in prostate radiotherapy (JA Stryker, J. Shafer, and RE Beatty: Assessment of the accuracy of daily facilities in prostate radiotherapy using electronic imaging, Br J Radiol 1999 Jun; 72 (858): 579-83.).

In addition, a transabdominal mobile targeting system (BAT) based on ultrasonography has been developed that can stereotactically locate the position of the prostate every day of treatment and directly integrate this information into the treatment planning system. Daily target verification facilitates a marked reduction in treatment planning margins by correcting for potential organ movement and placement errors. Preliminary studies have been performed to establish the accuracy of ultrasound localization.

This report quantifies the magnitude of the patient's isocenter displacement parameters taken into account during the clinical implementation of this system (J.

Lattanzi, S. McNeeley, S. Donnelly, E. Palacio, A. Hanlon, T. E. Schultheiss, G.E.

Hanks: Ultrasound-based stereotactic guidance in prostate cancer- quantification of organ motion and facility errors in external beam radiation therapy.

Retrograde urethrogram is commonly used to define the apex of the prostate during simulation. The study by S. Malone (2000) evaluated the hypothesis that urethrography causes displacement of the prostate, resulting in an error in treatment planning. Forty-five patients with carcinoma of the prostate evaluated. Gold grains were placed in the apex, posterior wall, and base of the gland. In the first 20 patients, the position of the grain-defined apex was compared to the simulation (with urethrogram) and Day 1 of treatment (without uregrogram). In the second cohort of 25 patients, the effects of urethrogram on the prostate position were evaluated directly at the simulation by comparing the position of the pre- and post-urethrogram apex. An analysis was performed to estimate the possible impact of urethrogram-induced prostate movement on target coverage. The mean superior displacement (displacement) in the first and second cohorts was 5.2 mm and 6.8 mm, respectively (combined with the average shift of 6.1 mm) with a field margin of 10 mm below the tip. the cone of the urethrogram, 56% of patients in

<Desc / Clms Page number 199>

 this study may have inadequate coverage of Target Planning Volume (PTV). Retrograde urethrogram causes a significant superior rise of the prostate. Strict confidence in urethrography by determining the lower margin of the field may result in inadequate treatment (S. Malone, R. Donker, M. Broader, S. Dahrouge, J. Szanto, L. Gerig, G. Bociek, and J. Prostate Radiation Oncol Biol Phys 2000 Jan 1; 46 (1): 89-93.) Conformational radiotherapy should be used instead to identify this soft reality. - the targeting of the target including the prostate enclosed in the pelvis not only from the point of view of the contours but also from the point of view of the histobiological aspects of the tumor. Recent studies have demonstrated that magnetic resonance spectroscopic imaging (MRSI) of the prostate can effectively distinguish between prostate cancer epithelium and normal regions. This diagnostic imaging tool takes the advantage over increased choline plus creatine versus the ratio of citrates found in malignant prostate tissue compared to normal prostate tissue. The same result can be achieved with the tissue perfusion kinecanography of the present invention. To describe a new brachytherapy treatment planning optimization module, using an integer programming technique that uses biology-based optimization, a method that records the MRSI to US images obtained intraoperatively. incorporates this information into a treatment planning system, is described to achieve dose escalation to intra-prostatic deposition of the tumor. The MRSI was obtained for a patient with clinically localized Gleason 7 prostate cancer. The ratios of choline plus creatine to citrate were analyzed for prostate and for malignant cells from high-risk areas were identified. The peaks of the representative ratios on the NMR spectra were calculated on a spatial grid covering the prostatic tissue. A procedure for mapping the points of the images is described. A programming integer technique is described as an optimization module for determining the optimal grain distribution for permanent interstitial implantation.

The MRSI data is incorporated into the treatment planning system to test the feasibility of scaling and dosing positive voxels with a relative backup of the surrounding normal tissues. The resulting tumor control probability (TCP) is estimated and compared to the TCP for the standard implantation planned for

<Desc / Clms Page number 200>

 brachytherapy. The proposed brachytherapy treatment planning system is capable of achieving a minimum dose of 120% of the prescription of 144 Gy to 100% -50% positive MRS voxels to the prostate and 100% to 120% to the urethra. maintained. When compared to a standard plan without MRS-guided optimization, the TCP estimated for the MRS-optimized plan is higher. Enhanced TCP was more pronounced for smaller volumes of intra-prostate tumor deposition compared to estimated TCP values for larger lesions. Using this brachytherapy optimization system, we can demonstrate the feasibility of MRS-optimized dose distributions for 125I permanent prostatic implants. Based on estimates of the likelihood of improved TCP anticipated, this approach may impact the ability to safely escalate the dose and potentially improve outcome for patients with confined organ but aggressive prostate cancers. The magnitude of the enhancement of TCP and therefore the risks of ignoring the NMR data appear to be more substantial, when the tumor is well localized; however, the achievable gain in TCP may be quite dependent on the efficiency of tumor detection by MRS.

The study by K. Kitamura et al. (2000) on the relationship between the technical parameters of external radiotherapy and prostate treatment complications was performed to retrospectively review, as a complication of external beam radiotherapy for localized prostate cancer, and to analyze the relationship between the technical parameters of radiotherapy and complications, the clinical course of chronic rectal bleeding. 71 patients with stage A2, B and C were treated, between 1989 and 1998, with local field radiotherapy (total dose 52.2 - 66 Gy, daily dose of 2.0-3.28 Gy, area of the field
30-81 cm2, number of fields 3-15 ports, planning simulations based on X-ray or CT), in three institutions. During the same period, the protocols were consistent with these institutions. The multivariate analysis revealed that the sum of Psa and Gleason of pre-treatment, in a mean follow-up period of 42 months (ranging from 12-119 months), were statistically significant predictors of 5-year antigen-free recurrence rates. prostatic specific (Psa). The significant risk factors for the higher gradmg of acute morbidity was an equivalent biological dose, alpha / beta = 10 (BED10)> or = 65 Gy, the dose per fraction> or = 3.0 Gy, the field> or = 42 cm2, fewer gates (rarer) and radiographic planning simulation. However, no parameter was associated with the high grading of late morbidity. IIpatient CI 5.4%)

<Desc / Clms Page number 201>

 experienced a late complication GI: grade 1 (4.2%), grade 2 (9.8%), grade 3 (1.4%). The median time to onset of rectal bleeding was 12 months after radiotherapy and the mean duration of morbidity was 11 months. The highest dose and dose per fraction, the widest field area, the least wear and the radiographic simulation increased the grades of acute morbidity. A majority of chronic rectal bleeds were transient and responded to conservative treatment. K. Kitamura, H. Shirato, K. Suzuki, N.

Shinohara, T. Demura, T. Harabayashi, T. Nishioka, K. Kagei, N. Takayama, Y,
Shinno, K. Kawakura, T. Koyanagi, and K. Miyasaka: The relationship between external radiation therapy and localized prostate cancer. Jpn J Clin Oncol 2000 May; 30 (5): 225-9. ). But all this could be avoided by an irradiation whose aim is irreproachable and therefore unbiased, this is also true for the calculation of the dose delivered.

5. 5.4. - Modeling for REAL TIME DOSIMETRY and EN
LINE of the device
In the foregoing description, the particular reference is made to traveling wave linear accelerators, since each of the linear accelerators is in common use, the simulation and dosimetry of which can be made from the data of the stereoradioradiotherapy apparatus and the present invention. invention. However, it is obvious that the advantages of our invention also apply well to the dosimetry of the sources of telechiotherapy as well as to the multiple sources of orthovoltage X-ray photons, by means of an integrated modeling in the computer of the device.

 High resolution dosimetry is except for use in stereotactic radiotherapy. in the treatment of small, well-defined lesions, a concern in all areas of radiotherapy. Dynamic StereoRadiotherapy is a technique of bicephalic external irradiation with mirror symmetry. Unlike conventional fractional radiotherapy, a high dose of radiation in stereotactic radiation therapy is often administered in a single session.

. To avoid side effects, the preservation of healthy tissue by fractionation of the dose should be compensated for by a corresponding lower dose to the normal tissue adjacent to the Radiotherapy target. This basic requirement can be completely fulfilled by a carefully elaborated irradiation technique, allowing selective delivery including multiple simultaneous beams and exactly reproducible dose to target volume. Planning

<Desc / Clms Page number 202>

 Exact treatment requires detailed knowledge in the individual beam of the dose distribution to determine the dose distribution inside. of the patient. The physical data required by the treatment planning system must often be experimentally determined. Several studies on the specific characteristics of photon and electron beams as well as methods for performing narrow photon beam dosimetry measurements have been conducted by different teams (Rice et al, 1987, Heyderian et al, 1996, Rustzi and Frye). 1995, McLaughlin et al., 1994, et al., 1998, Francescon et al., 1998, McKerracher and Twaites 1999). They may be in addition to the energy kernel method adapted to multiple sources of simultaneous photon irradiation and the present invention may serve as a tool for calculating such procedures for other more conventional therapeutic apparatuses.

5.5.4.1. - Mon-Carlo dosimetric procedures that can be applied online and in real time
In the early 1990s, computer console technology improved to such an extent that the size and cost of the computers required for Monte Carlo transport applications made them feasible for more widespread use. This technology, coupled with advances in LLNL's MonteCarlo transport methods, has provided a unique opportunity to apply advanced computer modeling technology to Radio-Oncology. This is how Chandler et al. have proposed the calculation of the dose in the body, using the three-dimensional Monte-Carlo transport. Neutrons, protons, deuterons, newts, helium-3, alpha particles, photons, electrons, and positrons are transported in a fully coupled manner, using the Monte-Carlo All-Particle
Method (MCAPM) developed at LLNL. The major elements of Chandler et al. discussed below include a computer, a digitized data file containing the patient description (a CT scan), a description of the radiation source, a Monte Carlo transport software, and digitized output files containing the distributions. of the dose. The user can define, either manually or from a single computed tomography section (CT) of the patient, a Cartesian mesh of transport. If the transport mesh is obtained from a scan section, the x and y dimensions of the area are defined as the dimensions of each pixel on the scan section, while the z ~ dimension represented by the space between the sections CT. Each zone within said mesh is, according to Chandler et al. (WO 97/32630) Assigned

<Desc / Clms Page number 203>

 of a material number associated with a specific atomic composition and density. Materials are assigned using a software-based thresholding algorithm based on CT scans or can be assigned manually.

The source of radiation can be defined as a set of external beams (or internal radioactive sources), each element of the source of radiation (beam or internal source) is characterized by its shape and position in space, as well as by the distribution of energy and the direction of the particles that it emits. For external beams, the radiation source is defined as a complex beam or set of beams in an isocentric gantry and standard table geometry. The user specifies the gantry of the accelerator, the collimator and the angles of the table, as well as the location (x, y, z) of the center (isocentre), in the transport mesh, rotation table / accelerator . The energy and direction of each particle is described in space (the plane of definition of the beam) with respect to a plane which is defined as being perpendicular to the radius connecting the source of radiation to the gantry / table isocenter. The particles are transported from the point source to the beam definition plane. The beam definition plane is divided into a series of concentric rings. Each concentric ring may have an energy and a different angular distribution associated with the latter. Depending on the location territorially located on the definition plane of the beam where the particle lands, a new energy and a new angle are given to it by sampling from the chosen energy and angular distributions of the digitized libraries which include monoenergetic options, monodirectional, computer as well as energies and angular distributions derived from Monte Carlo simulations of actual beam delivery systems. The angular distributions chosen by the user modify, in terms of the definition of the beam, the trajectory of the particle. The shape of the field is defined when the particles pass through the jaws of the collimators and through an optional compensator and / or a block modeling the field. Particle transport through the jaws of the collimator is done by a tracer ray, whereas Monte-Carlo methods are used instead to transport through the compensator and the block. Once each emerging particle leaves the beam collimation system (jaws, compensators, and block) it is, according to Chandler et al. (WO 97/32630) transported to the patient's mesh.

<Desc / Clms Page number 204>

Chandler et al. thus realize the calculations of Monte-Carlo transport, in a coordinate system of the patient without the means in the current art to establish it instantly, knowing that a particle is first selected from the source of radiation. Depending on the type, it is passed to the neutron, the photon, the electron / positron, appropriate or to the transport conditioning of the particle, where it is tracked until it undergoes a collision . For charged particles (electrons, positrons, and heavy charged particles), the energy is deposited, by combining the distance traveled with the information of the falling (slowing down) of the predetermined stopping energy. Each nuclear reaction and each photon interaction is manipulated using the Monte Carlo AllParticle Method (MCAPM). This passage of a given incident particle returns the energies and angles of all the secondary particles resulting from the collision. These latter particles are stored in a secondary bank. The control is then returned to the Monte-Carlo tracker which selects from the source of radiation a particle from the secondary bank or when the bank is exhausted. This process continues until all particles from the source and the secondary bank are monitored (Chandler et al: WO 97/32630).

PCAPM software conditioning is a system of data and algorithms that performs particle collision physics for the simulation of many types of particles. The nuclear and atomic data libraries mentioned above describe with the matter the interaction of neutrons, photons, electrons, and heavy charged particles. Each interaction can result in the production of any of these particles. Additional data describing Coulomb interactions of electrons are included in Chandler et al. (WO 97/32630).

Although the cross sections of the neutron and charged heavy particle beams are, on average, in the Chandler et al. System above the energy groups or chests, each particle has a discrete energy value allowing collision kinetics to run on a pure continuous energy method basis (storing as much information as necessary to fully describe the dependence of each collision parameter on energy) and the pure multi-group method (storing all collision parameters on a group of energy or on the vault base), allowing a database of smaller section, while retaining the exact kinematics. The photon sections are interpolated between discrete energy points. Neutrons and photons are

<Desc / Clms Page number 205>

 transported, using the standard Monte Carlo methods. The energy transferred to the particles that are not being tracked is deposited with the locally expected energy deposition providing the reduction of the variance. For heavy charged particles, nuclear reactions and wide-angle Coulomb scattering are processed on an event-by-event basis, while all other Coulomb interactions are accounted for by the condensed history method. The Condensed History method uses the ideas of Transport Physics to combine the effects of multiple interactions as it traverses a specified distance. To account for multiple Coulomb collisions that occur within a condensed history step of the charged particle, Chandler et al. (WO 97/32630) proposes to modify the energy of the particle and its trajectory by using pseudo-Gaussian distributions with variances determined from the Rutherford diffusion section (energy straggling) of the Highland approximation to a distribution (multiple diffusion) of Molière.

* Given that electrons and positrons are transported, using a Class II condensed history method, Coulomb wide-angle scattering and bremsstrahlung production are modeled on an event-by-event basis, while all others Coulomb interactions are taken into account with the condensed history method. After each stage of the condensed history, the trajectory of the electron (or positron) is modified by sampling from a Molière distribution. This system, Macro Response Monte-Carlo Method, can, as an alternative to condensed history be used in the future. The Monte Carlo transport conditioning calculates the energy deposited from the radiation source, creating a three-dimensional Cartesian mapping of the energy deposited inside the patient. When the Monte Carlo calculations are complete, the mapping of the energy deposit is converted into a mapping of the absorbed dose, and it is, according to Chandler et al. (WO 97/32630), recalled back to the user (client treatmentplanning code). The input (patient and radiation source description and user options of the control code) as well as the output (three-dimensional dose mapping) are read and written in a set of files, which conform to the specifications. of the TapelNetwork Format for Exchange of Treatment Planning Information, Version 3.00 [01/10/94], which is routinely maintained by NCI Prostate Collaborative Working Group RTOG 3D QA Center,

<Desc / Clms Page number 206>

 Washington University, St. Louis (USA). These files are referred to as standard AAPM files. The form of the input and output data can, according to Chandler et al. (WO 97/32630), change in the future to accommodate the standards of the new industry or allow for a more direct link with the commercial patient treatment planning system. So we spend his time calculating! Calculate the target volume, the safety margins around the target volume, calculate the various uncertainties, the computer generated images to check the ballistic alignment, etc. and there is very little time left for radiotherapy. This is how we move ever farther and farther from the goal of doing better, cheaper and faster, not to mention the number of staff involved in all these different processes.

5.5.4.2. - Monte Carlo dosimetric procedures for electron beams
A hybrid approach was made by Jiang et al. and developed to delegate electron-beam systems to Monte Carlo processing planning.

The approach is based on the assertion that accelerators of the same type have very similar electron beam characteristics and a major difference from a fine tuning at the exit window of the incident electron energy on the electron beam. single site. For an accelerator type, a reference machine can be selected and simulated with the Monte Carlo method. A multiple source model can be constructed for the complete Monte Carlo simulation of the reference beam. By proposing the electron beams from other accelerators of the same type, the energy spectrum was, in the source model, set to harmonize the measured dose distributions. A Variait Clinac accelerator
For example, 2100C was chosen as a reference machine and a model, based on Monte Carlo simulations of a four-source beam. This simplified beam model can be used to accurately generate (in the range of 2% / 2 mm compared to those calculated with space data of the various phases), the Monte Carlo distributions of the dose electron beams from a reference machine having different nominal energies, applicator dimensions (different) and different SSDs.

 Three electron beams have been proposed, adjusting the energy spectrum in the source model. The calculated dose distributions with the adjusted source model were compared to the calculated dose distributions, using the spatial phase data of these beams. The agreement is in most cases

<Desc / Clms Page number 207>

 realized in the interval of 1% and that of 2% in all other situations.

 This preliminary study showed an approaching ability to manipulate wide variation of incident electron energy. The possibility of making this approach more flexible is also discussed by Jiang et al. (2000).

 The simulation of Monte-Carlo transport of radiation is considered by these authors as one of the most accurate methods of calculating radiotherapy dose. With the rapid development of computer technology, Monte Carlo based on radiotherapy treatment planning has become practical. A basic requirement of Monte-Carlo treatment planning is a detailed knowledge of the radiation beams of medical accelerators.

 A practical approach to achieve this is to perform the simulation of Monte Carlo radiation transport in the medical accelerator, and what is more, the Monte-Carlo modeling of the head of the processing machine can also improve the understanding of the clinical features of the beam, providing more than 5, realistic beam data. Ma et al, (1999) summarize in their article the work of the past two decades on the Monte Carlo simulation of the electron beams of medical accelerators. Otherwise, one must be aware that factors exist to systematically contradict this dosimetric forecast.

 The introduction in radiotherapy of dynamic intensity modulation, using the '. conventional photon beams in which particle beam scanning requires additional and efficient methods of dose verification. Dynamically generated dose measurements in dose distributions using a single ionization chamber require full application of the treatment field for each single measurement. The measurements are therefore 5. carried out by the simultaneous use of multiple ionization chambers. The measurement is carried out by a computer-controlled system and consists of the following steps: (a) automated positioning of the ionisation chambers, (b) measurement at these points, (c) comparison with the dose calculated from the treatment planning system, and (d) documentation of the measure. The ionization chambers are read aloud by a multi-channel electrometer and are densely packaged in a polymethyl methacrylate assembly, which is attached to the arm of an aqueous three-dimensional motor-driven phantom. The measured and planned values of the dose are both numerically and graphically displayed.

 The average deviation between measured and planned doses as well as their standard deviations are calculated and displayed. Thanks to the complete documentation

<Desc / Clms Page number 208>

 The measurement is obtained and a quick decision can be made when the dose distribution is acceptable to the patient. The system is now routinely used by Karger et al. (1999) for dose verification of the heavy ion therapy project at Gesellschafifur Schwerionenforschung, Darmstadt.

 Until the date of publication of the article by these authors, 242 measurements were made for the heavy ion treatment of 30 patients. The latter system allows for efficient verification and documentation of carbon ion fields and is in principle also applicable to modulated intensity photon beams.

In addition, small fields where electronic equilibrium is not achieved are becoming increasingly important in clinical practice. These complex situations give rise to problems and inaccuracies in dosimetry and analytical / empirical dose calculation together, and therefore require other methods than conventional ones. This is why a detector in. natural diamond and a parallel flat ionization chamber were selected by
H. de Vlamynek et al. for clinical dosimetry in 6 MV beams. The results and simulations using the BEAM / EGS4 Monte Carlo system, to model the beam geometry, were thus compared with the dose measurements. A modification of the module of the existing component of the collimators. Multiblades (MLCs) would allow the modeling of an SL 25 linear accelerator (Elektra Oncology Systems) equipped with an MLC having curved end blades. A mechanical method of measuring with spacer plates and a technique for detecting the bright edge of the field are described as methods for obtaining data for application in the Monte Carlo system. geometric openings of the collimator. Generally a good agreement is found between measurements and calculations of the depth dose distributions and the deviations typically remain below 1%. The profiles calculated for a 10 x 2 cm2 field of the lateral dose slightly exceed the distributions measured near the highest level of penumbra, but they agree well. in all other cases with the measurements of the profiles. The simulations are also able to predict the variations of the output factors as well as the ratios, as a function of the field width and the offset field, of the output factors. The Monte Carlo results published by H. de Vlamynck et al. (1999) demonstrate that qualitative changes in energy spectra are too small to explain. these variations and that the geometric factors particularly affect the

<Desc / Clms Page number 209>

 outlets as well as curves and depth dose profiles.

The technique of extracting energy spectra from electrons from measured dose distributions along the central axis of clinical electron beams is explored in detail. Clinical spectra measured with this simple spectroscopy tool are shown to be sufficient in accuracy and resolution for use in Monte Carlo processing planning. A series of mono-energetic curves of the appropriate depth-spacing energy dose, precalculated for a single beam model, with the Monte-Carlo method are deployed from the depth curve of the measured dose.

The beam pattern that included a dot electron and a photon source placed in an empty chamber with a distance from the source to the surface of 100 cm.

The systematic error introduced by this model affects the depth dose curve calculated at about 2% / 2 mm. The component, first subtracted from deployment, of the dose due to bremsstrahlung of the treatment head is estimated from the Schiff spectrum of the thin target in 0.3% of the maximum total dose (from the electrons and photons) on the beam axis. Optimal deployment settings are chosen based on physical principles. The deployment is done with the FERDO code of the public domain. Comparisons were made with the published spectra, previously measured with magnetic spectroscopy and with the spectra calculated with the simulation of the treatment head.

The approach gives smooth spectra with an average resolution of the 27 studied beams of 16 3% peak average energy. The average peak energy of the magnetic spectrometer spectra was calculated within 2% for the AECL T20 scanning beam accelerators, 3% for the Philips SL25 machine, based on the scattering sheet. The number of low energy electrons in the Monte Carlo spectra is estimated by deployment with an accuracy of 2%, relative to the total number of electrons in the beam. The central axis depth dose curves calculated from the deployment spectra are within 0.5% / 0.5 mm of the measured and simulated depth dose curves, except near the range. in practice, where 1% / 11mm errors are evident (BA Faddegon and I. Blevis: Electron spectra derived from depth dose distributions, Med Phys 2000 Mar; 27 (3): 514-26).

<Desc / Clms Page number 210>

REFERENCES - AAPM: A protocol for the determination of absorbed dose from high-energy photons and electrons. Med Phys 10; 1983: 741.

- AAPM: Clinical electron beam dosimetry: reportof AAPM Radiation Therapy
Committee Task Group N 25. Med Phys 18; 1991: 73-109.

A. Ahnesjo: Analytic modeling of photon from flattening filters in photon therapy beams, Med Phys 21; 1994: 1227-1235.

A. A. al-Beteri and D.E. Raeside: Optimal electron-beam treatment planning for retinoblastoma using a new three-dimensional Monte Carlo-based treatment planning system. Med Phys19; 1992: 125-135.

- Ali M. and C. Storey: Topographical multilevel single linkage; J Global
Optimization 5; 1994: 349-358.

- Ali M. and C. Storey: A new topographical algorithm for global optimization. J
Faculty Sci 7, 1995: 132-138.

- MM Ali, C. Storey and A. Torn: Application of some global stochastic optimization algorithms to practical problems. J Optimization Theory Appl 95;
1997: 545-563.

- P. Andreo: Monte Carlo simulation and electron transport in water. Absorbed dose and fluence distributions. Report FANZ / 80/3 (University of Zagora, Spain:
Department of Nuclear Physics); 1980.

- P. Andreo: Monte Carlo techniques in medical radiation physics. Phys Med Biol
36; 1991: 861-920.

- P. Andreo and A. Brahme: Restricted energy loss straggling and multiple scattering of electrons in mixed Monte Carlo procedures. Radiat Res. 100; 1984: 16-29.

- P. Andreo, A. Brahme, A. Nahum, and O. Mattsson: Influence of energy and angular spread on stopping power ratios for electron beams. Phys Med Biol 34;
1989: 751-768.

 - P. Andreo and A. Fransson: ratios and their uncertainties for clinical electron beam dosimetry. Phys Med Biol 34; 1989: 1847-1861.

- KM Ayyangar and SB Jiang: Do we need the Monte Carlo treatment planning for
Linac based stereotactic radiosurgery? - A case study. Med Dosim 23; 1998:

<Desc / Clms Page number 211>

161-167.

- JP Balog, TR Mackie, DL Wenman, Glass M., G. Fang and D. Pearson:
Multileaf collimator interleaf transmission. Med Phys 26; 1999: 176-186.

- JM Balter, RK Ten Haken, Lawrence SL, et al: Uncertainties in CT-based radiation therapy treatment planning associated with patient breathing. Int J
Radial Oncol Biol Phys 36; 1996: 167-174.

 J. Barkhof, G. Schut, J. B. Flanz, M. Goitein, and J. M. Schippers: Verification of the alignment of a therapeutic radiation beam relative to its patient positioning.

 MedPhys 26 [11]; 1999: 2429-2437.

. - G. Baroni, G. Ferrigno, R. Orecchia, and A. Pedotti: Real-time three-dimensional motion analysis for patient positioning verification. Radiother Oncol 54 [There
2000: 21-7.

 - G. Baroni, R. Orecchia, F. Torretta, G. Ferrigno, and A. Pedotti: Automatic opto-electronic control of a patient in radiotherapy for breast carcinoma [abstract]. Radiol Med (Torino) 98 [1-2]; 1999: 78-84.

- A. Bel, M. van Herk, and J. Lebesque: Target margins for random geometrical treatment in conformai radiotherapy. Med Phys 23; 1996: 1537-
1545.

A. Bel, H. Bartelink, RE Vijlbrief, et al: Transfer errors planning CT to simulator: A possible source of setup inaccurancies? Radiother Oncol 31; 1994:
176-180.

 - C.C. Bentel, L.B. Marks, R. Krishnamurthy: Impact of cradle immobilization on the reproducibility of external radiation therapy for lung cancer. Int J Radiat Oncol Biol Phys 38; 1997: 527-531.

SM Bentzen and S. Dische: Late Morbidity: The Damocles Sword of Radiotherapy? Radiotherapy and Oncology, 2001, Vol 61, No 3: pp. 223-232 - MJ Berger and SM Sletzer: The influence of scattering foils on absorbed dose distributions from electron beams. Report NBSIR 78-1552 (Gaitherburg: NBS);
1978.

. - AF Bielajew: Monte Carlo Modeling in external radiotherapy electron beam - why leave it to chance? Proc 11th Int Conf on the Use of Computers in
Radiation Therapy ed. AR Hounsell, JM Wilkinson and PC Williams (Manchester: Christie Hospital, North Western Medical Physics
NHS Trust); 1994: pp 2-5.

<Desc / Clms Page number 212>

 A.F. Bielajew, H. Hirayama, W. R. Nelson and D.W.O. Rogers: History, overview and recent improvements of EGS4 Report PIRS-0436 (Ottawa: NRCC); 1994.

AF Bielajew and DWO Rogers: PRESTA: the parameter reduced electron-step transport algorithm for electron Monte Carlo transport. Nucl Instrum
Methods B; 1987: 165-181.

AF Bielajew and Rogers DWO: Variance-reduction techniques Monte Carlo
Transport of Electrons and Photons TM Jenkins et al (New York: Plenum);
1988 pp 407-419.

- AF Bielajew and Rogers DWO: A stardard timing benchmark for EGS4
Monte Carlo calculations. Med Phys 19; 1992: 303-304.

- AF Bielajew and PE Weibe: EGS-Windows - a graphics interface to EGS
Report PIRS-0274 (Ottawa: NRCC); 1991.

- R. Boellaard, M. van Herk, H. Uiterwaal, B. Mijnheer: First clinical tests using a liquid-filled electronic device and a convolution model for the verification of the midplane dose. Radiother Oncol 47; 1998: 303-312.

- R. Boellaard, M. van Herk, H. Uiterwaal, and BJ Mijnheer: A convolution model to convert dose transmission images to exit dose distributions. Med Phys 24;
1997: 189-199.

- C. G. E. Boender, A.H.G. Rinnooy and G. T. Timmer: A stochastic method for global optimization. Math Program 22; 1982: 125-140.

- Hans J. de Boer, John R. van Sômsen of Koste at al: Analysis and Reduction of
3-D Systematic and Random Setup Errors During the Simulation and Treatment of Lung Cancer Patients with CT-based External Beam Radiotherapy Dose
Planning. Int J Radiat Oncol Biol Phys 49 [3]; 2001: 857-868.

- R. Bogaerts, D. Huyskens, C. Weltens, and A. Dutreix: Variation of relative transit dose profiles with patient-detector distance. Radiother Oncol 54; 2000:
2937.

- R. Bogaerts, A. van Esch, Reyman R., and D. Huyskens: A method to estimate the transit dose on the beam axis for portal images. Radiother Oncol 54; 2000: 39-46.

- S. N. Boon, H. Meertens, P. van Luijk, J. M. Schippers, S. Vinckier, J. M. Denis, J.

 Medin, and E. Grusell: A fast 2D phantom dosimetry system for scanning proton beams. Med Phys 25; 1998: 464-475.

S. Boon N., P. Luijk, T. Bohringer, A. Coray, A. Lomax, E. Pedroni, B. Schaffner,

<Desc / Clms Page number 213>

and JM Schippers: Performance of a fluorescent screen and CCD camera as a two-dimensional dosimetry system for dynamic treatment techniques. Med Phys
27 (10); 2000: 2198-2208.

Borrell-Carbonell, JP Patau, M. Terrissol and D. Trunk: Comparison between experimental measurements and calculated transport simulations for electron dose distributions inside homogeneous phantoms. Strahlentherapie 156; 1980:
186-191.

T. Bortfeld, AL Boyer, W. Schlegel, DL Kahler, and TJ Waldron:
Realization and verification ofthree-dimensional conformal radiotherapy with modulated fields. Int J Radiat Oncol Biol Phys 30 (4); 1994: 899-908.

A.L. Boyer, T. G. Ochran, C. E. Nyerick, and T. J. Waldrom: Clinical dosimetry for implementation of a multileaf collimator. Med Phys 19; 1992: 1255-1261.

A. Brahme, B. Lind, and Nafstadius: Radiotherapeutic computed tomography with scanned photon beams. Int J Radiat Oncol Biol Phys 13; 1987: 95-101.

- JF Briesmeister: MCNP - a general Monte Carlo N-particle transport code, version 4A Report la-12625 (Los Alamos, NM: Los Alamos National
Laboratory); 1993.

- C. Brusasco, A. Cattai, R. Cirio, G. Dellascasa, M. Donetti, P. Isoards, F.

Marchetto, C. Peroni, V. Rolando, M. Ruspa, A. Solano, and C. Zambernardi:
Strip ionization chambers as 3-D detector for hadron therapy. Nucl Instrum
Methods Phys Res A 389; 1997: 499-512.

- 25. JM Buatti, SL Meeks, WA Friedman, and FJ Bova: Stereotectic radiosurgery: techniques and clinical applications. Surg Oncol Clin N Am 2000
Jul; 9 [3]: 469-87, viii.

 T.E. Burlin, J.M. Sidwell and B.M. Wheatley: Applications of Monte Carlo methods in medical radiology. Br J Radiol 46; 1973: 398-399.

D. Burns, S. Duane, and McEwen: A new method to determine the ratios of electron stopping powers to an improved accuracy. Phys Med Biol 40; 1995:
733-739.

RW Byhardt, JD Cox, A. Hornburg, G. Liermann: Weekly localization films and detection of field placement errors. Int J Radiat Oncol Biol Phys 4; 1978:
881-887.

 R. W. Byhardt, C. Scott, W. T. Sause, et al: Response, Toxicity, Failure Patterns, and Survival in Five Radiation Therapy Oncology Group [RTOG] trials of sequential and / or concurrent chemotherapy and radiotherapy for locally

<Desc / Clms Page number 214>

advanced non-small cell carcinoma of the lung. Int J Radial Oncol Biol Phys 42;
1998: 469-478.

 M. F. Chan, K. M. Ayyangar: Confirmation of target localization and dosimetry for 3D radiotherapy treatment by MR imaging of a ferrous sulphate gel head phantom. Med Phys 22 (7); 1995: 1171-5.

- R. Corvo, F. Foppiano, M. Orsatti, F. Bisi, G. Ghiso, R. Martinelli, L. Valanzola,
V. Vitale: The Application of the Multileaf Collimator in Radiotherapy: The Dosimetric Problems and the Technical Implications for its Clinical Use (Abstract) Radiol Med (Torino) 94 (5): 512-9.

-1. Daftari, C. Castenadas, Petti PL, Singh RP, and LJ Verhey: An Application of GafChromic MD-55 Film for 67.5 MeV clinical proton beam dosimetry. Phys
Med Biol 44; 1999: 2735-2745.

M. Essers, J.W. Lanson, G. Leunens, T. Schnabel, and B. J. Mijnheer: The accuracy of CT-based inhomogeneity corrections and in vivo dosimetry for the treatment of lung cancer. Radiother Oncol 37; 1995: 199-208.

- Mr Essers, Mr BR Hoogervorst, Mr van Herk, Mr H. Lanson, and Mr BJ Mijnheer:
Dosimetric characteristics of a liquid-filled electronic portal imaging device.

 Int J Radiat Oncol Biol Phys 33; 1995: 1265-72.

- M. Essers, and B. J. Mijnheer: In vivo dosimetry during external photon beam radiotherapy. Int J. Radiat Oncol Biol Phys 43; 1999: 245-59.

C. Fiorino, C. Uleri, G. M. Cattaneo, and R. Calandrino: On-line exit dose profile measurements by a linear array diode. Phys Med Biol 41; 1996: 1291-304.

- JB Flanz, S. Durlacher, M. Goitein, A. Levine, P. Reardon, and A. Smith:
Overview of the MGH-Northeast Proton Therapy Center plans and progress,
Nucl Instrum Methods Phys Res B 99; 1995: 830-834.

- J. B. Flanz et al: The approach for a highly accurate positioning of patients for NPTC, in Advances in Hadrontherapy, Excerpta Medica, edited by U.

 Amaldi, B. Larssen, and Y. Lemoine; Elsevier, Amsterdam, 1997: pp 260-268.

- R.M. Cardinale, S.H. Benedict, Q. Wu, R.D.Zwicker, H.E. Gaballa, and R.

Mohan: A comparison of three stereotactic techniques; ARC vs. noncoplanar fixed fields vs. intensity modulation. Int J Radiat Oncol Biol Phys 42; 1998:
431-6.

- William P. Chandler, Christine L. Hartamann-Siantar, and James A. Rathkopf:
Calculation of Radiation Therapy with All Particle Monte Carlo Transport.

 PCT / US 97/03328. Calculation of dosimetry in Radiotherapy using the

<Desc / Clms Page number 215>

 transport all particles according to the Monte Carlo process. EP 0 956 099 Al (International Publication Number WO 97/32630 - 12.09.1997, Gazette 1997/39).

- E.L. Chaney, T.J. Cullip and T.A. Gabriel: A Monte Carlo study of accelerator head scatter. Med Phys 21; 1994: 1383-1390.

- LM Chin, CW Cheng, RL Siddon, Rice RK, BJ Mijnheer, Harris JR: Three dimensional photon dose. Int J Radiat Oncol Biol Phys 17; 1989:
1327-1335.

- J.C. Chu, B. Ni, R. Kriz, and V. Amod Saxena: Applicationsof simulator computed tomography for photon dose calculations during radiotherapy treatment planning. Radiother Oncol 55 [1]; 2000: 65-73.

C. Chui, G. Kutcher, and T. LoSasso: A convolution method for incorporating uncertainties in dose calculations. Med Phys 19; 1992: 814.

T. Craig, J. Battista, V. Moisenko, J. van Dyk: Considerations for the implementation of target volume protocols in radiation therapy. Int J Radiat Oncol Biol
Phys 49 [He; 2001: 241-250.

 C. L. Creutzberg, V.G. Althof, H. Huizenga, A. G. Visser, P.C. Levendag: Quality assurance using portal imaging: the accuracy of patient positioning in irradiation of breast cancer. Int J Radiat Oncol Biol Phys 25; 1993: 529-239.

- T. Csendes: Interval methods for bounding level sets: revisited and tested with global optimization problems. ILO 30; 1990: 650-657.

- I.J. Das, G.E. Desorby, S.W. McNeeley, E.C. Cheng, and T.E. Schultheiss: Beam characteristics of a double-focused multileaf collimator, Med Phys 25; 1998: 1676-1884.

- IJ Das, CW Cheng, Fein DA, Coia LR, WJ Curran, Fowble B. Dose estimation to a significant effect. Int J
Radiat Oncol Biol Phys 37; 1997: 1023-9.

- I.J. Das, C. W. Cheng, H. Fosmire, K. Kase, T. Fitzgerald: Tolerances in set-up and dosimetric errors in radiation treatment of breast cancer. Int J Radiat Oncol Biol Phys 26; 883-890.

J. Debus, EB Hug, Liebsch, D. O'Farrel, D. Finkelstein, J. Efird, et al:
Brainstern tolerance to conformal therapy of skull base tumors Int J Radiat
Oncol Biol Phys 39; 1997: 967-975.

<Desc / Clms Page number 216>

- S. D'haese, V. Vinh-Hung, P. Bijderke, M. Spinnoy, M. De Beukeleer, N. Lochie,
P. De Roover and G. Storme: The effect of timing of the supply of information on cancer patients receiving radiotherapy.

 J. Cancer Educ 2000 Winter; 15 (4): 223-7.

J. O. Deasy, P. R. Alamond and M. T. McEllistrem: Measured electron energy and angular distributions from clinical accelerators. Med Phys 23; 1996: 675-684.

- WR de Gersem, S. Derycke, Colle CO, Wagter C., and WJ De Neve:
Inhomogeneous target-dose distributions: a dimension more for optimization?
Int J Radiat Oncol Biol Phys 44 [2]; 1999 461-468.

- WR de Gersem, S. Derycke, Wagter C., and WC de Neve: Optimization of beam weights in conformal radiotherapy planning of stage III non-small cell lung cancer: effects on therapeutic ratio. Int J Radiat Oncol Biol Phys 47 [1];
2000: 255-60.

J.J. De Marco, T. D. Solberg, and J. B. Smathers: A CT-based Monte Carlo simulation tool for dosimetry planning and analysis. Med Phys 25; 1998: 1-11.

- J. E. Dennis Jr and J.J. More: Quasi-Newton methods, motivation and theory.

 SIAMRev 19; 1977: 46-89.

C. DesRosiers, V. Moskvin, A. F. Bielajew, and L. Papiez: 150-250 MeV electron beams in radiation therapy. Phys Med Biol 45 [7]; 2000: 1781-805.

- K. de Vlamynck, H. Palmans, F. Verhaegen, C. de Wagter, W. de Neve, and H.

 Thierens: Dose measurements compared with Monte Carlo simulations of narrow 6 MV multileaf collimator shaped photon beams. Med Phys 26 [9]; 1999: 1874-1882.

- G. E. Desorby and A. L. Boyer: An analytic calculation of the energy fluence spectrum of a linear accelerator. Med Phys 21; 1994: 1943-1952.

- G.X. Ding and D.W.O. Rogers: Energy spectra, angular spread, and dose distributions of electron beam from various accelerators used in radiotherapy.

 Report PIRS-0439 (Ottawa: NRCC); 1995.

- GX Ding, Rogers DWO and Mackie TR: calculation of backscatter in a beam monitor chamber for electron beams [abstract]. Med Phys 21; 1994: 923-
924.

- G.X. Ding, D.W.O. Rogers and T.R. Mackie: Calculation of energy using realistic clinical electron beams. Med Phys 22; 1995: 489-501.

- G.X. Ding, D.W.O. Rogers R. Mackie Mean energy, energy-range

<Desc / Clms Page number 217>

relationship and depth-scalling factors for clinical electron beams. MedPhys 23;
1996: 361-376.

- GX Ding, Rogers DWO and Mackie TR: Electron Fluence Correction Factors used in the conversion of measured dose in plastic to dose in water. Med Phys 24;
1997: 161-176.

C. Duzenli, B. McClean, and C. Field: Backscatter in the beam monitor chamber: Implications for dosimetry of asymmetric collimators, Med Phys 20;
1993: 363-367.

- MA Ebert and PW Hoban: A model for electron-beam applicator scatter. Med
Phys 22; 1995a: 1419-1429.

- M. A. Ebert and P.W. Hoban: A Monte Carlo investigation of the electron-beam applicator scatter. Med Phys 22; 1995b: 1431-5.

- M. A. Ebert, P.W. Hoban, and P. J. Keall: Modeling clinical accelerator beams: a review. Australas Phys Eng Sci Med 19; 1996: 131-150.

- L. Ekberg, O. Holmberg, L. Wittgren, et al: margins should be added to the clinical target for radiotherapy treatment planning for lung cancer?
Radiother Oncol 48 1998: 71-77.

 O. Esik, T. Bortfeld, R. Bendl, G. Nemeth, and W. Schlegel: Inverse radiotherapy planning for a concave-convex PTV in cervical and upper medial regions. Simulation of radiotherapy using an Alderson-Rando phantom.

 Planning target volume. Strahlenther Onkol 173 [4]; 1997: 193-200.

 # L. Eyges: Multiple scattering with energy loss. Phys Rev 74; 1948: 1534-5.

- B.A. Faddegon: Spectroscopy for advanced treatment planning in electron therapy [abstract]. Med Phys 22; 1995: 666.

- BA Faddegon, Balogh J., Mackenzie R., and Scora D. Clinical considerations of
Monte Carlo for electron radiotherapy treatment planning. Radiat Phys Chem
53; 1998: 217-227.

- B.A. Faddegon, P. O'Brien and D.L.D. Mason: The flatness of Siemens linear accelerator x-ray field. Med Phys 26; 1999: 220-8.

 - B.A. Faddegon and I. Blevis: Electron spectra derived from dose dose distributions; Med Phys 27 [3]; 2000: 514-26.

A. Fein, K. McGee, T. Schultheiss, B. Fowble, G. Hanks: Intra- and interfractional reproducibility of tangential breast fields: a prospective on-line portal imaging study. Int J Radiat Oncol Biol Phys 34; 1996: 733-740.

<Desc / Clms Page number 218>

E. Fermi: B. Rossi and K. Greisen: Cosmic-ray theory. Rev Mod Phys 13;
1941: 265-268.

EM Fernandez, GS Shentall, WPM Mayles, and DP Dearnaley: The acceptability of a multileaf collimator
Radiother Oncol 36; 1995: 65-74.

C. Fiorino, S. Broggi, D. Corletto, GM Caettaneo, and R. Calandrino: Conformal Irradiation of Concave-shaped PTVs in the Treatment of Prostate Cancer by Simple
1-D intensity-modulated beams. Radiother Oncol 55 [1]; 2000: 49-58.

JB Flanz, S. Durlacher, M. Goitein, A. Levine, P. Reardon, and A. Smith:
Overview of the MGH-Northeast Proton Therapy Center plans and progress.

 Nucl Instrum Methods Phys Res B99; 1995: 830-834.

A. Floudas and Pardalos: Recent Advances in Global Optimization (Princeton)
University Press); 1992.RRL Ford and WR Nelson: The ECG code system: computer programs for the
Monte Carlo simulation of electromagnetic waterfall showers (version 3) Report
SLAC-210 (Stanford, CA: SLAC); 1978.

 A. Frazier, D. Yan, M. Du, J. Wong, F. Vicini, R. Matter, M. Joyce, and A.

Martinez: Effects of treatment with the multileaf collimator on the effects of radiation therapy using the multileaf collimator vs the cerrobend block. Int J
Radiat Oncol Biol Phys 33; 1995: 1247-1256.

N. Fukunaga-Johnson, M. Sandler M., R. Marsh, and M.K. Martel: The use of 3-D conformal radiotherapy (3D CRT) to separate the cochlea in patients with medulloblastoma. Int J Radiat Oncol Biol Phys 41; 77-82.

 N. Fukunaga-Johson, J. H. Lee, H.M. Sandler, P. Robertson, E. McNeil, and J. W.

 Goldwein: Patterns of failure following treatment for medulloblastoma: Is it necessary to treat the entire posterior fossa? Int J Radiat Oncol Biol Phys 42; 1998b: 143-6.

AS Garden, Maior MH, WK Yung: Outcome and patterns of failure following a limited-volume irradiation for malignant astrocytomas. Radiother
Oncol 20; 1991: 99-110.

Gaspar, BJ Fisher, MD Mac Donald: Supratentorial malignant glioma: patterns of recurrence and implications for external beam local treatment. Int J
Radiat Oncol Biol Phys 24; 1992: 55-57.

D. Georg and A. Dutreix: A formalism to calculate the output ratio in a mini phantoms for a GE multileaf collimator. Phys Med Biol 42; 1997: 521-536.

<Desc / Clms Page number 219>

- From Goldberg: Genetic Algorithms in Search, Optimization and Machine
Learning (Reading, MA: Addison-Wesley); 1989.

- R. Gordon: A tutorial on ART. IEEE Trans Nucl Sci 29; 1974: 78-93.

J.C. Gore and S. Leeman: The reconstruction of objects from incomplete projections. Phys Med Biol 25; 1980: 129-136.

 # J.R. Gray, B. McCormick, L. Cox, J. Yahalom: Primary breast irradiation in large breasted or heavy women: analysis of cosmetic outcome. Int J Radiat Oncol Biol Phys 21; 347-354.

S. Gripp, R. Doeker, M. Glag, P. Vogelsang, and H. Pape: Conventional and virtual simulation in retrobulbar irradiation. [Abstract] Strahlenther Onkol 2000
Tue; 176 [3]: 131-4.

- Guan Huaiqun, Fang-Fang Yin, Zhu Yunping, and Jae Ho Kim: Adaptative Gantry CT Reconstruction: A simulation study. Med Phys 27 [10], 2000: 2209-
2214.

- Guan Huaiqun, Fang-Fang Yin, Zhu Yunping, and Jae Ho Kim: Adaptative Gantry CT Reconstruction: A simulation study. Med Phys 27 [10], 2000:
2209-2214.

 H. Guan and Y. Zhu: Feasibility of CM Portal Gmgavoltage using an electronic gantry imaging device (EPID) and a multilevel algebraic technical reconstruction scheme (MLS-ART). Phys MedBiol 43 1998: 2925-2937.

- H. Guan and R. Gordon: A projection of order for convergence speed of algebraic reconstruction techniques (ART): A multilevel scheme (MLS) for computed tomography. Phys Med Biol 39; 1994: 2005-2022.

- OCL Haas: Radiotherapy Treatment Planning: New System Approaches
London, Springer); 1999: pp. 135-176.

- Th Habarer, W. Becher, D. Schardt, and G. Kraft: Magnetic scanning system for heavy ion therapy, Nucl Instrum Physics Methods Res A.330; 1993: 296.

- C. Haie-Meder, A. Beadre, C. Breton, B. Biron et al: The conformational radiotherapy of brain tumors. Cancer Radiother 3; 1999: 407-413.

- John W. Haller, PhD, Timothy C. Ryken, MD, Michael W. Vannier, MD: Image-guided Therapy: for Practical Applications. Acta Radiol 2001; 8:
888-897.

R. Halperin, W. Roa, M. Field, et al: Reproducibility in radiation therapy for lung cancer: A comparison between T-bar and expanded foam immobilization

<Desc / Clms Page number 220>

 devices. Int J Radial Oncol Biol Phys 43; 1999: 211-216.

- K. Han, Ballon D., CS Chui and Mohan R.: Monte Carlo simulation of a cobalt-
60 beam. MedPhys 14; 1987: 414-419.

CL Hartmann-Siantar et al: LawrenceLivermore National Laboratory's
PEREGRINE project. Proc 12th Int. Use of Computers in Radiation
Therapy ed DD Leavitt and G. Starkschall (Madison: Medical Physics
Publishing); 1997: pp 19-22.

JA Hayman, MK Martel, RK Ten Haken, et al: Dose escalation in nonsmall cell lung cancer using conformai 3-dimensional radiation therapy: Update of a
Phase 1 trial [abstract]. J Clin Oncol 18; 1999: 459a.

- J.A. Heilbeib and T.A. Mehlhorn: ITS: the Integrated Tiger Series of coupled electron photon Monte Carlo transport codes. Report SAND84-0573 (Albuquerque, NM: SandiaNational Laboratories); 1984.

JS Hendricks and JF Briesmeister: Recent MCNP developments. Report LA-UR-91-3456 (Los Alamos, NM: Alamos National Laboratory); 1992.- Herman WG: Image Reconstruction from Projections: The Fundamentals of
Computerized Tomograph. Academic, New York; 1980.

 Hess FF, Kortmann R.D., Jany R, Hamberger A, Bamberg M .: A retrospective analysis of clinical practice.

 Radiother Oncol 34; 1995: 69-72.

 B.M. Hesse, L. Spies, and B.A. Groh: Tomotherapeutic portal imaging for radiation treatment verification. Phys Med Biol 43; 1998: 3607-3616.

 K.R. Hogstrom, M. D. Mills and P. R. Almond: Electron Beam Dose Calculations.

 Phys Med 26; 1981: 445-459.

 A.A. Holmes, T.R. Mackie, W. Sohn, P.J. Reckwerdt, T.J. Kinsella, A.F.

 Bielajew, and D.W.O. Rogers: The application of correlated sampling to the computation of electron beam dose distributions in heterogeneous phantoms using the Monte Carlo method by Phys Med Biol 38; 1993: 675-688.

 # T. Holmes and T.R. Mackie: A comparison of three inverse treatment planning algorithms. Phys Med Biol 39; 1994: 91-106.

- A. R. Hounsell and J. M. Wilkinson: Head scatter modeling for irregular field shaping and beam intensity modulation, Phys Med Biol 42; 1997: 1737-1749.

 # A.R. Hounsell: Monitor chamber backscatter for intensity modulated radiation

<Desc / Clms Page number 221>

 therapy using multileaf collimators, Phys Med Biol 43; 1998: 445-454.

M.A. Hunt, T. E. Schulttheiss, G. Desorby, and G. E. Hanks: Convolving setup uncertainties with dose distributions. Med Phys 20; 1993: 929.

 C. W. Hurkmans, A. E. Saarnak, B. R. Pieters, J. H. Borger, and LA. Bruinvis: An improved technique for breast cancer irradiation including locoregional lymph nodes. Int J Radiat Oncol Biol Phys 47 [5]; 2000: 1421-9.

DP Huyskens: Does Gated Media Dosimetric Control Have a Future?
Cancer Radiother 4 Suppl; 2000: 36s-43s
IAEA: An International Code of Practice IAEA Technical Report Series 277 (Vienna: IAEA); 1987.

 - D.A. Jaffray and J. H. Siewerdsen: Cone-beam computed tomography with flet- panel imager: Initial performance characterization. American Association of Physicists in Medicine 2000; S0094-2405 (00) 0 1306-7. In Bibliographic Review of /, Radiol 881; 2000: 1774.

- TM Jenkins, Nelson WR, A. Rindi, AE Nahum, and Rogers DWO: Monte
Carlo Transport of Electrons and Photons (New York: Plenum); 1988.

- S.B. Jiang and K.M. Ayyangar: On compensator design for photon beam intensity modulated radiation therapy. Med Phys 25; 1998: 668-675.

 S.B. Jiang, A. Kapur and C-M. Ma: Electron beam modeling and commissio ning for Monte Carlo treatment planning. Med Phys; 1999.

- SB Jiang, ZM Luo and KM Ayyangar: Incorporation of the electron energy - loss straggling into the Fermi-Eyges equation. Radiat Phys Chem 53; 1998:
477-482.

 S.B. Jiang, A. Kapur, and C.M. Ma: Electron beam modeling and commissio- ning for Monte Carlo treatment planning. Med Phys 27 [1]; 2000: 180-91.

B. Jones and RG Dale: Radiobiological modeling and clinical trials. Int J Radial
Oncol Biol Phys 48 [1]; 2000: 259-65.

 - T.J. Jordan and P.C. Williams: The design and performance characteristics of a multileaf collimator. Phys Med Biol39; 1994: 231-251.

- P.A. Jursinic: Clinical Implementation of a two-component x-ray source model for the calculation of head-scatter factors, Med Phys 24; 1997: 2001-2006.

K. Kagei, H. Shirato, T. Nishioka, T. Arimoto, S. Hashimoto, M. Kaneko, K.

 Ohmori, A. Honma, Y. Inuyama, and K. Miyasaka: Ipsilateral irradiation for

<Desc / Clms Page number 222>

 carcinomas of tonsillar region and soft palate based on computed tomographic simulation. Radiother Oncol 2000 Feb; 54 (2): 117-21.

 - A.H.G. Kan Rinnooy and G. T. Timmer: Stochastic Methods for Global Optimization. Am JMath Management Sci 4; 1984: 7-40.

AHG Kan Rinnooy and Timmer GT: Stochastic global optimization methods;
Part 1: Clustering methods Math Program 39; 1987a: 27-56.

- AHG Kan Rinnooy and Timmer GT: Stochastic global optimization methods;
Part II: Clustering methods Math Program 39; 1987b: 57-78.

A. Kapur and C-M. Ma: Stopping power ratios for clinical electron beams from a scatter-foil linear accelerator. Phys Med Biol 44; 1999: 2321-2341.

- A. Kapur, CM. Ma, EC Mok, DO Findley: Characterization of small field electron beams for radiotherapy using Monte Carlo simulations Proc 12th Int
Leavitt and G.DDD in Radiation Therapy

 Starkschall (Madison: Medical Physics Publishing); 1997: 157-8 pp.

A. Kapur, C-M. Ma, E.C. Mok, D.O. Findley and A. L. Boyer: Monte Carlo calculations of electron beam output factors for a medical linear accelerator.

 Phys Med Biol 43; 1998: 3479-3494.

 C. P. Karger, O. Jakel, and G.H. Hartmann: A system for three - dimensional dosimetric verification of treatment plans in intensity - modulated radiotherapy with heavy ions. Med Phys 26 [10]; 1999: 2125-32.

AB Karim, B. Maat, R. Haltevoll, J. Menten, EH Rutten, DG Thomas, et al. :
Randomized trial on dose-response in radiation therapy of low-grade glioma:
European Organization for Research and Treatment of Cancer (EORTC) Study
22844. Int J Radiat Oncol Biol Phys 36; 1996: 549-556.

 Mr. G. Karlsson, Mr. K. Karlsson, and C-M. Ma: Treatment head design for multileaf collimated high-energy electrons. Med Phys 26; 2161-2167.

Kassae MD Altschuler, S. Ayyalsomayajula and P. Bloch: Influence of cone design on the electron beam characteristics on clinical accelerators. Med Phys
21; 1994: 1671-1676.

- Kawrakow, M. Fippel and K. Friedrich: 3-D electron dose calculations using a Voxel-based Monte Carlo algorithm (VMC). Med Phys 23; 1996: 445-457.

- P.J. Keall and P.W. Hoban: The angular and energy distribution of the primary electron beam. Australas Phys Eng Sci Med 17; 1994: 116-123.

- P.J. Keall and P.W Hoban: Calculating the standard angular deviation of electron

<Desc / Clms Page number 223>

 beams using Fermi-Eyges theory. Phys Med Biol 41; 1996: 1511-1515.

 P. J. Kelly, C. Daumas-Duport, D. B. Kispert, B. A. Kall, B. W. Scheithauer, J. J.

 Illig: Imaged-based stereotaxic serial biopsies in untreated intracranial neoplasms. J Neurosurg 66; 1987: 865-74.

H. Kestelman: Lebesgue measure, in Modern Theories of Integration 2ndrevisited edn (New York: Dover); 1960: pp 67-91.

FM Khan: The Physics of Radiation Therapy (Baltimore, MD: Williams
Wilkins); 1994.

 B. Kihlen, B.I. Ruden: Reproductibility of field alignment in radiation therapy: a large scale clinical experience. Acta Oncol 28; 1989: 689-692.

 Y. Kinoshita, K. Kohshi, N. Kunungita, T. Tosaki, and A. Yokota: Preservation of tumor oxygen after hyperbaric oxygenation by magnetic resonance imaging. British Journal ofCancer 82 [1]; 2000: 88-92.

 M. Kirby, P. Williams: Portai imaging for the verification of breast treatments.

 Radiother Oncol 22; 1991: 314-316.

K. Kitamura, H. Shirato, K. Suzuki, N. Shinohara, T. Demura, T. Harabayashi,
T. Nishioka, K. Kagei, N. Takayama, Y. Shinno, K. Kawakura, T. Koyanagi, and
K. Miyasaka: The relationship between technical parameters and external radiation. Jpn J Clin
Oncol 30 [5]; 2000: 225-9.

EE Klein, WB Harms, DA Low, V. Willcut, and JA Purdy: Clinical Implementation of Multileaf Collimator: Dosimetry, Networking, Simulation, and Quality Assurance. Int J Radiat Oncol Biol Phys 33; 1995: 1195-
1208.

- M. Kocher, H. Treuer, J. Voges, M. Hoevels, V. Sturm, and RP Muller:
Computer simulation of cytotoxic and vascular effects of radiosurgery in solid and necrotic brain metastases. Radiother Oncol 54 [2]; 2000 149-56.

 L. Kocis and W. Whiten: Computational investigations of low-discrepancy sequences ACMTrans Math Software 23; 1997: 266-294.

- R.D. Kortmann, C. F. Hess, W. Hoffmann, R. Jany, M. Bamberg: Is the standardized technical helmet adequate for irradiation of the brain and the cranial meninges? Int J Radiat Oncol Biol Phys 32; 1995: 241-244

<Desc / Clms Page number 224>

HD Kubo, Len PM, S. Minohara, and H. Mostafavi: Breathing-synchronized radiotherapy program at the University of California Davis Cancer Center. Med
Phys 27 [2]; 2000: 346-53.

J. Kurhanewicz, DB Vigneron, SJ Nelson: Three-dimensional magnetic resonance spectroscopic imaging of brain and prostate cancer. Neoplasm 2 [1-2];
2000: 166-189.

GJ Kutcher, CS Chui, and T. LoSasso: Incorporation of installation uncertainties in treatment planning calculations. Int J Radiat Oncol Biol Phys
21; 1991: 123.

- S. L. Kwa, J.V. Lebesque, C. M. Theuws, L. B. Marks, M. T. Munley, G. Bentel, et al: Radiation pneumonitis as a function of average lung dose: an analysis of pooled data of 540 patients. Int J Radiat Oncol Biol Phys 42; 1998: 1-9.

DA Jaffray, DG Drake, M. Moreau, AA Martinez, and JW Wong: A radiography and tomographic imaging system integrated into a medical linear accelerator for localization of bone and soft-tissue targets. Int J Radiat Oncol
Biol Phys 45 (3); Oct 1999: 773-89.

C.P. Karger, O. Jâkel, G. Hartmann, and Heeg: A system for three-dimensional dosimeters in radiotherapy with heavy ions. Med Phys 26; 1999: 2125-2132.

- AI Katsevitch: Novel techniques for image computation and a concomitant reduction of the x-ray dose in transmission tomography. Proc SPIE 2074;
1966: 405-416.

 H. D. Kubo, B.C. Hill: Respiratory gated radiotherapy treatment: a technical study. Phys MedBiol 41 (1); 1996: 83-91.

 K.J. Kuchala, G. H. Olivera, E. A. Schloesser, and T. R. Mackie: Megavoltage CT on a tomotherapy system. Phys Med Biol 44; 1999: 2597-2621.

 J.H. Kung, G. T.Y. Chen, and F.K. Kuchnir: A monitor unit auditing in modulated radiation as a dosimetry quality assurance.

 Med Phys 27 (10); 2000: 2226-2230.

- KL Lam, MS Muthuswamy, and Ten Haken RK: Applatants-filter-based empirical methods to parametrize the head scatter factor. Med Phys 23; 1996:
343-352.

 M. Langer, R. Brown, S. Morrill, R. Lane, and O. Lee: A generic genetic algorithm for generating the beam. MedPhys 23; 1996: 965-971.

<Desc / Clms Page number 225>

J. Lattanzi, S. McNeeley, A. Hanlon, TE Schutheiss, and GE Hanks:
Ultrasound-based stereotactic guidance of precision conformal external beam radiation therapy in clinically localized prostate cancer. Urology 55 [1]; 2000:
73-8. See also J. Lattanzi, S. McNeeley, S. Donnelly, E. Palacio, A. Hanlon,
TE Schultheiss, GE Hanks: Ultrasound-based stereotactic guidance in prostate cancer-quantification of radiation therapy.

 T. S. Lawrence, R.J. Tesser, and R.K. Ten Haken: An Application of Dose Volume Histograms to the Treatment of Intrahepatic Malignancies with Radiation Therapy.

 Int J Radial Oncol Biol Phys 19; 1990: 1041-1047.

T. Lawrence, R.K. Ten Haken, M. L. Kessler, J. M. Robertson, J. T. Lyman, M.L.

 Lavigne, M. B. Brown, D. J. DuRoss, J. C. Andrews, W. D. Ensminger, and A. S.

 Lichter: The use of 3-D dose volume analysis to predict radiation hepatitis. Int J Radiat Oncol Biol Phys 23, 1992: 781-788.

# T. Knight, R. Arriagada, Tarayre, et al: Significant effect of adjuvant chemotherapy on locally advanced survival nonsmall cell lung carcinoma. J
Natl Cancer Inst 184; 1992: 158.

 M.C. Lee, S.B. Jiang and C.M. Ma: Monte Carlo and experimental investigations ofmultileaf collimated electron beams for modulated electron radiation therapy. Med Phys 27 [12]; 2000: 2708-18.

- P.C. Lee: Monte Carlo simulations of the differential beam hardening effect of a flattening filter on a therapeutic x-ray beam. Med Phys 24; 1997: 1485-1489.

- SW Lee, Fraass BA, LH Marsh, K. Herbort, Gebarski SS, MK Martel, et al: Patterns of high-dose following high-dose conformai radiotherapy for high-grade astrocytomas: quantitative dosimetric study. Int J Radiat Oncol Biol
Phys 43; 1999: 79-88.

- J. Leong: Implementation of random positioning error in computational radiation treatment planning. Phys Med Biol 32;
1987: 327-334.

R. Lewis, S.J. Ryde, D.A. Hancock, and C.J. Evans: An MCNP-based model of a linear accelerator x-ray beam. Phys Med Biol 44; 1999: 1219-1230.

- DG Lewis, W. Swindell, EJ Morton, PM Evans, and Z R. Xiao: A
Megavoltage CT scanner for radiotherapy verification. Phys Med Biol 37;
1992: 1985-1999.

 - R.M. Lewitt: Processing of incomplete measurement data in computed

<Desc / Clms Page number 226>

 tomography. Med Phys 6; 1979: 412-417.

- JG Li, AL Boyer, and L. Xing: Clinical implementation of radiotherapy treatment planning. Radiother
Oncol 53 [3]; 1999: 257-64.

- JG Li, L. Xing, AL Boyer, RJ Hamilton, DR Spelbring, and JV Turian:
Matching photon and electron fields with dynamic intensity modulation. Med
Phys 26 [11]; 1999: 2379-84.

- A. Lirette, J. Pouliot, M. Aubin, M. Larochelle: The role of electronic portal imaging in tangential breast irradiation: a prospective study. Radiother Oncol
37; 1995: 241-245.

H.H. Liu, T.R. Mackie and E.C. McCullough: A dual source photon beam model used in convolution / superposition dose calculations for clinical megavoltage x-ray beams. Med Phys 24; 1997: 1960-1974.

- HH Liu, EP Lief, and EC McCullough: Measuring dose distribution for enhanced dynamic wedges using a multichamber detector array. Med Phys 24;
1997: 1515-1524.

- S. Liu, P. Andreo, A. Satherberg, I. Gudowska and A. Brahme: Monte Carlo simulation of accelerators: application to a MM50 Racetrack microtron [abstract]. Radiother Oncol 37 [suppl]; 1995: 119.

* - F. Lohr, 0. Schramm, P. Schraube, et al: Simulation of 3D-treatment plans in head and neck tumors aided by matching of digitally reconstructed radiographs (DRR) and on-line distortion corrected simulator images. Radiother Oncol 45; 1997: 199-207.

DMJ Lovelock, CS Chui, Kutcher J. and Mohan R.: Analysis of the photon beam. Med Phys 21;
1994: 1969-1977.

- D. M. J. Lovelock, C. S. Chui, and R. Mohan: A Monte Carlo model of photon beams used in radiation therapy. Med Phys 22; 1995: 1387-1394.

- HM Lu, E. Cash, MH Chen, L. Chin, WJ Manning, J. Harris, and B. Bornstein: Reduction of Cardiac Volume in Left-Breast Treatment Fields by Respiratory Maneuvers: a CT study. Int J Radial Oncol Biol Phys 2000 Jul; [4]: 895-
904.

AE Lujan, RK Ten Haken, EW Larsen, and JM Balter Quantification of setup uncertainties in 3-D dose calculations. Med Phys 26 [11]; 1999: 2397-
2405.

<Desc / Clms Page number 227>

AE Lujan, EW Larsen, JM Balter, and RK Ten Haken: A Method for Incorporating Organ Motion due to breathing in 3-D dose calculations. Med
Phys 26; 1999: 715-720.

 W. Lutz, K. R. Winston, and N. Maleki: A system for stereotactic radiosurgery with a linear accelerator. Int J Radiat Oncol Biol Phys 14; 1988: 373-381.

- C-M. Ma, S.B. Jiang: Monte Carlo modeling of electron beams from medical accelerators. Phys Med Biol 44 [12]; 1999: R157-89.

- CM. Ma: Monte Carlo simulation of dosimeter response using transputers. PhD
Thesis University of London (Report ICR-PHY-1/92 (Sutton, UK: ICR)); 1992.

- CM. Ma: Characterization of computer simulated radiation therapy for Monte
Carlo treatment planning. Radiat Phys Chem 53; 1998: 329-344.

- CM. Ma, Faddegon BA, Rogers DWO and Mackie TR: Accurate characterization ofMonte Carlo calculated electron beams for radiotherapy. Med
Phys 24, 1997a: 401-416.

- CM. Ma, A. Kapur, EC Mok, and D. Findley: Monte Carlo Treatment Planning: Procedures for Controlling Computer Simulated Clinical Chemistry (abstract). J
Int Fed Med Biol Eng 35; 1997c: 995.

- C-M. Ma, E.C. Mok, A. Kapur, T. Pawlicki, D.O. Findley, S.Brain, K.Forster, and A.L.Boyer: Clinical Implementation of a Monte Carlo Treatment Planning System. Med Phys 26; 1999: 2133-2143.

- CM. Ma and AE Nahum: Calculation of absorbed dose ratios using correlated
Monte Carlo sampling. MedPhys 20; 1993: 1189-1199.

- CM. Ma, P. Reckwerdt, M. Holmes, Geer Rogers DWO and B. Walters:
DOSXYZ users manual. Report PIRS-0509 (b) (Ottawa: NRCC); 1995.

- C-M. Ma and D.W.O. Rogers: BEAMDP users manual. Report PIRS-0509 (C) (Ottawa: NRCC); 1995a.

- CM. Ma and DWO Rogers: Beam Characterization: A Multiple Source Model
Report PIRS-0509 (C) (Ottawa: NRCC); 1995b.

- CM. Ma and DWO Rogers: BEAMDP - as a general-purpose utility. postponement
PIRS-0509 (C) (Ottawa: NRCC); 1995c.

- C-M. Ma, D. W.O. Rogers, G.X. Ding and T.R Mackie: Electron Beam Characterization: Reconstruction Models and Dose Distributions In a homogeneous phantom [Abstract] Med Phys 21; 895.

- C-M. Ma, D.W.O. Rogers, B.A. Faddegon, G.X. Ding, J. S. Wei, F. Bielajew

<Desc / Clms Page number 228>

 and T.R. Mackie: Simplified models of electron beams from a 2100C accelerator [abstract]. Med Phys 20; 1993a: 1295.

CM. Ma, DWO Rogers, K. Shortt, Ross K., AE Nahum, and AF Bielajew:
Wall correction and absorption rate factors for Fricke dosimetry:
Monte Carlo calculations and measurements. Med Phys 20; 1993b: 1507-1513.

- C-M. Ma and J. P. Seuntjens: Correction factors for waterproofing sleeves in kilovoltage x-ray beams. Med Phys 24; 1997: 1507-1513.

- Chang-Ming Ma and Steve B. Jiang: Monte Carlo modeling of electron beams from medical accelerators. Phys Med Biol 44, 1999: R157-R189.

- L. Ma, P. B. Geis and A. L. Boyer: Quality Assurance for Dynamic Multileaf Collimator Modulated Fields Using a Fast Beam Imaging System. Med Phys 24 (8); 1997: 1213-20.

R. I. MacKay, PA Graham, C. J. Moore, J. P. Loque, and P. J. Sharrock: Animation and Radiobiological Analysis of 3-D motion in conforma! radiotherapy.

 Radiother Oncol 52 [1]; 1999: 43-9.

Tr Mackie: Applications of the Monte Carlo Method in Radiation Radiation Dosimetry of Ionizing Volume 3, ed K. Kase, B. Bjamgard and FH Attix (San Diego, USA).
CA: Academic); 1990: pp 541-620.

- TR Mackie, Holmes TW, PJ Reckwerdt and J. Yang: Tomotherapy: Optimized planning and delivery of radiation therapy. Int J Imaging Syst Technol 6;
1995: 43-55.

T. T. Mackie, T. W. Holmes, S. Swerdloff, P. J. Reckwerdt, J. Deasy, J. Yang, B.

 Paliwal and T. Kinsella: Tomotherapy: a new concept for the delivery of conformal therapy MedPhys 20; 1993: 1709-1719.

- TR Mackie, SS Kubsad, Rogers DWO and AF Bielajew: The OMEGA project; electron dose planning using Monte Carlo simulation [abstract]. Med
Phys 17; 1990: 730.

- TR Mackie et al: The OMEGA project: comparison among EGS4 electron beam simulations, 3-D Fermi-Eyges calculations, and dose measurements. proc
XIth ICCR (Manchester, UK) (Christie Hospital NHS Trust); 1994: pp 152-3.

- T. R. Mackie, J. W. Scrimger and J. J. Battista: A convolution method of calculating dose for 15-MV x-rays. Med Phys 12; 1985: 188-196.

S. Malone, R. Donker, M. Broader, S. Dahrouge, J. Szanto, L. Gerig, G. Bociek, and J. Crook: Effects of urethrography on prostate position: considerations for

<Desc / Clms Page number 229>

radiotherapy treatment planning of prostate carcinoma. Int J Radiat Oncol Biol
Phys 46 [1]; 2000: 89-93.

 Manfredotti, U. Nastasi, R. Marchisio, C. Ongaro, G. Gervino, R. Ragona, S.

Anglesio and G. Sannazzari: Carlo simulation of dose distribution in electron beam radiotherapy treatment planning. Nucl Instrum Methods; 1990 :
A 291 646-54.

- Manfredotti, U. Nastasi, R. Ragona and S. Anglesio: Comparison of three- dimensional Monte Carlo simulation and the pencil beam algorithm for an electron beam from a linear accelerator. Nucl Instrum Methods 987: A 255,355.

- LB Marks, GW Sherouse, S. Das, Bentel GC, DP Spencer, and D. Turner:
Conforma radiation therapy with fixed-shaped coplanar gold non-coplanar radiation beam bouquets: a possible alternative to radiosurgery. Int J Radiat
Oncol Biol Phys 33; 1995: 1209-1219.

- J. Mazurier, F. Salvat, B. Chauvenet and J. Barthe: Simulation of photon beams from a Saturn 43 accelerator using the PENELOPE code. Phys Med 15; 1999:
101-110.

- R.C. McCall, R.D. McIntyre, and W.G. Turnbull: Improvement of linear accelerator depth-dose curves. Med Phys 5; 1978: 518-524.

- C.J. McGinn, R.K. Ten Haken, W. Ensminger, S. Walker, S. Wang, and T.S.

Lawrence: The treatment of intrahepatic cancer with radiation doses based on a normal tissue complication probability model. J Clin Oncol 16; 1998: 2246-
2252.

- J. Maryansky, et al. : relaxation enhancement in gels polymerized and cross-linked by ionizing radiation: A new approach to 3 D dosimetry by MRI.

 Magn Reson Imaging 11, 1993: 253.

McLaughlin, W., J.C. Humphrey, B.B. Radak, A. Miller, and T.A. Olejnik, The response of plastic dosimeters to gamma rays and electrons at high absorbed dose rates. Radial Phys Chem 14; 1979: 535-550.

S.L. Meeks, J.M. Buatti, F.J. Bova, W.A. Frieman, W.M. Mendenhall, R.A.

 Zlotecki: Potential clinical efficacy of intensity-modulated conforma! therapy.

 Int J Radiat Oncol Biol Phys 40; 1998: 483-495.

- V. K. Mehta and D. R. Goffinet: Unsuspected abnormalities noted on CT treatment - schedule scans obtained for breast and chest wall irradiation. Int J.

 Radiat Oncol Biol Phys 49 [3]; 2001: 723-725.

<Desc / Clms Page number 230>

- Thomas Mertelmeier and Bernhard Scholz: Vorrichtung zur Untersuchung von Lebendem Gewebe. (Apparatus for examining living tissue). EP 0 945 103 A1.
Publ 29.9.1999, Bulletin 1999/39 - filing number 99104920. 6, filing date 11.03.1999 (SIEMENS AKTIENGESELLSCHAFT, 80333 München [DE]) - JM Michalski, JW Wong, RL Gerber, D. Yan, A. Cheng , MV Graham, MA

Renna, PJ Sawyer, and CA Perez: The use of on-line image verification to estimate the variation in radiation therapy dose delivery. Int J Radial Oncol Biol
Phys 27; 1993: 701-716.

S. Midgley, R. M. Millar, and J. Dudson: A feasability study for megavoltage cone beam CT using a commercial EPID. Phys Med Biol 43; 1998: 155-169.

- C. Mitine, M.T. Hoornaert, A. Dutreix, and M. Beauduin: Radiotherapy of pelvic malignancies: the impact of two types of rigid immobilization devices on localization errors. Radiother Oncol 52 [1], 1999: 19-27.

- U. Mock, K. Diekmann, U. Wolff, T. H. Knocke, and R. Potter: Gantry based on definition of the planning target volume during pelvic irradiation for gynecological malignancies. IntJ Radiat Oncol Biol 45 [1]; 1999: 227-32.

- R. Mohan: Monte Carlo dose calculations for radiation treatment machine heads
Monte Carlo Transport of Electrons and Photons. Ed TM Jenkins et al (New
York: Plenum); 1988: pp. 453-468.

- R. Mohan: Why Monte Carlo? Proc 12th Int.
Radiation Therapy; DD Leavitt and G. Starkschall (Madison: Medical
Physical Publishing); 1997a: pp 16-18.

- R. Mohan: Monte Carlo dose calculations for radiation therapy planning [abstract]. Med Phys 24; 1997b: 987.

 R. Mohan, C. S. Chui and L. Lidofsky: Energy and angular distributions of photons from medical linear accelerators. Med Phys 12; 1985: 592-597.

- EC Mok, CM. Ma, A. Kapur, D. Findley and AL Boyer: Predicting Air Gap Factors for Electron Fields at Extended Treatment Distances [abstract]. J Int Fed
Med Biol Eng 35; 1997: 1026.

V. Mongioj, A. Brusa, G. Loi, E. Pignoli, A. Gramaglia, M. Scorsetti, et al:
Accuracy evaluation of fusion of CT, MR, and SPECT images using commercially available software packages (SRS PLATO and IFS). Int J
Radiat Oncol Biol Phys 43; 1998: 227-234.

V. Mongioj, A. Brusa, G. Loi, E. Pignoli, A. Gramaglia, M. Scorsetti, et al. :
Accuracy evaluation of fusion of CT, MR, and SPECT images using

<Desc / Clms Page number 231>

commercially available software packaging (SRS PLATO and IFS). Int J
Radiat Oncol Biol Phys 43; 1998: 227-34.

A. Mosleh-Shirazi, P. M. Evans, W. Swindell, S. Webb, and M. Pertridge: A cone-beam megavoltage CT scanner for treatment verification in conformai radiotherapy. Radiather Oncol 48; 1998: 319-328.

- D. Murray: Using Monte Carlo EGS4 in medical radiation physics. Australas
Phys Eng Sci Med 13; 1990: 132-147.

- C. Mwanza Chabunda: Iconographic Clinical Approach and Physical Therapy in Radiological Medicine: Conceptual Foundations of the Representation Through the Medical Image of the Internal Environment, Through the History of Radiological Sciences. PhD Thesis, ULP, 2000.

- A. E. Nahum: Monte Carlo Electron Transport Simulation II: Application to dose planning The Computation of Dose Distributions in Electron Beam Radio- therapy; ed A. E. Nahum (Umea, Sweden: Umea University); 1985: p 319.

- AE Nahum: Overview of Monte Electron and Photon, in Carlo Monte Carlo
Transport of Electrons and Photons, TM Jenkins et al (New York:
Plenum); 1988: pp 3-20.

- A.E. Nahum: Condensed-history Monte Carlo simulations for particles: what can it do for us? Radiate Environ Biophys 38 [3]; 1999: 163-73.

K. Nakagawa, Y. Aoki, T. Fujimaki, A. Terahara, K. Karasawa, K. Sakata, et al:
High-dose conformai radiotherapy influenced the pattern of failure but did not improve survival in glioblastoma multiforme. Int J Radiat Oncol Biol Phys 40;
1998: 1141-1149.

- WR Nelson, Hirayama H. and Rogers DWO: The EGS4 code system, in
Report SLAC-265 (Stanford, CA: SLAC). ; 1985.

- WR Nelson and TM Jenkins: Computer Techniques in Radiation Transport and
Dosimetry (New York: Plenum); 1980.

- H. Neuenscwander, T.R. Mackie and P.J. Reckwerdt: MMC - a high-performance Monte Carlo code for electron beam treatment planning. Phys Med Biol 40 - 1995: 543-574.

- Mr Nevinny-Stickel, S. Ennemoser, R. Sweeney, I. Bangerl, D. Zur Nedden, and
Lukas: Comparison of standard and individually planned infradiaphragmatic fields in irradiation of retroperitoneal lymph nodes. Int J Radiat Oncol Biol Phys
48 [1]; 2000: 147-51.

<Desc / Clms Page number 232>

H. Niederreiter: Random Number Generation and Quasi-Monte Carlo Methods (CBNS-NSF Regional Conf Applied Mathematics 63), Philadelphia, PA:
SIAM; 1992.

- Nilsson and A. Brahme: Contamination of high-energy photon beams by scattered photon. Strahlentherapie 157; 1981: 181-186.

J. Nocedal and J. S. Wright: Numencal Optimization, New York, Springer; 1999.

MC Nutting, V. S. Khoo, V. Walker, H. McNair, C. Beardmore, A. Norman, and D. P. Dearnaley: A Randomized Study of the Use of a Customized Immobilization System in the Treatment of Prostate Cancer with Conformal Radiotherapy. Radiother Oncol 54 [1]; 2000: 1-9.

- K. Ogawa, N. Nakajima, and S. Yuta: A reconstruction algorithm from truncated projections. IEEE Trans Med Imaging 3; 1984: 34-40.

L.E. Olsson et al. : MR imaging of absorbed doses for radiotherapy using ferrous sulphate gels. Phys Med Biol 35, 1990: 1623.

- I. Orion, AB Rosenfeld, FA Dilmanian, F. Telang, B. Ren, and Y. Namito:
Monte Carlo simulation of dose distributions from a synchrotron-produced microplanar beam array using the EGS4 code system. Phys Med Biol; 45 [9];
2000: 2497-508.

- JB Owen and LR Coia: The Changing Structure of Radiation Oncology:
Implications for the Era of Managed Care. Semin Radiat Oncol 7 [2]; 1997 :
108-113.

J.R. Palta, D. K. Yeung, and V. Frouhar: Dosimetric considerations for a multileaf collimator system. Med Phys 23; 1996: 1219-1224.

- W. Pan: A two-sample test with internai censored data via multiple imputation.

 Stat Med 19 [1]; 2000: 1-11.

J. Patau, C. E. Vernes, M. Terrisol and M. Malbert: Calculation of the qualitative characteristics (TEL, FQ, dose equivalent) of a beam of photons of braking for medical use, by simulation of its creation and its transport. Proc 6th Symp Microdosimetry (1978), Brussels, Belgium (EUR 6064), ed J. Booz and H.G.

 Ebert (London: Hardwood Academic); 1978: pp 579-588.

- AC Paulino, A. Narayana, MN Mohideen, and S. Jeswani: Posteriorfossa boost in medulloblastoma: an analysis of dose to surrounding structures using 3-dimensional (conformal) radiotherapy. Int J Radiat Biol Oncol Phys 2000 Jan
15; 46 [2]: 281-6 Comment in: Int J Radial Oncol Biol Phys 2000 Jan 15; 46 [2]: 265-6

<Desc / Clms Page number 233>

- T. Pawlicki: Application of a new mixed-simulation Monte Carlo code system to electron beam radiotherapy treatment planning with computed data, PhD Thesis,
Medical College of Ohio; 1998.

E. Pedroni et al: The 200-MeV proton therapy project at the Paul Scherrer Institute: conceptual design and practical realization. Med Phys 22, 1995: 37-53.

E. Pedroni, R. Bacher, H. Blattmann, A. Coray, A. Lomax, S. Lin, G. Munkel, S.

Schreib, U. Schneider, and A. Tourovsky: The 200-MeV proton therapy project at the Paul Scherrer Institute: design and practical realization. Med
Phys 22; 1995: 37-53.

 - E. Penitsa, J. C. Rosenwald, and C. Kappas: Quality control of dose volume histogram computation characteristics of 3-D treatment planning systems. Phys Med Biol 43; 1998: 2807-2816.

 - C.A. Pelizzari and G. T. Chen: Volume visualization in radiation treatment planning. Crit Rev Diagn Imaging 2000 Dec; 41 [6]: 376-401.

PL Petti, MS Goodman, Gabriel T. and Mohan R.: Investigation of buildup dose from electron contamination of clinical photon beams. Med Phys 10;
1983a: 18-24.

 P. Petti, S. Goodman, J.M. Sisterson, P. Biggs, T.A. Gabriel and R.

 Mohan: Sources of Electron Contamination for the Clinac-3525-MV Photon Beam. Med Phys 10; 1983b: 856-861.

- MH Phillips, K. Singer, E. Miller, K. Stelzer: Commissioning an image-guided localization system for radiotherapy. Int J Radiat Oncol Biol Phys 48 [II; 2000:
267-76.

O. Pradier, H. Schmidberger, E. Weiss, H. Bouscayrol, A. Daban, and CF. Hess:
Accuracy of alignment in breast irradiation: a retrospective analysis of clinical practice. BJR 72; 1999: 685-690.

 W. Price: Global optimization by controlled random search. J Optimization Theory Appl 40; 1983: 333-348.

 - I. Rabinowitz, J. Broomberg, M. Goitein, K. McCarthy, J. Leong: Accuracy of radiation field alignment in clinical practice. Int J Radiat Oncol Biol Phys 11; 1985: 1857-67.

 - D.E. Raeside: Monte Carlo principles and applications. Phys Med Biol 21; 1976.

 181-197.

<Desc / Clms Page number 234>

CR Ramsey, D. Arwood, D. Scaperouth, and Oliver AL: Clinical application of digitally-reconstructed radiographs generated from magnetic resonance imaging for intracranial lesions. Int J Radiat Oncol Biol Phys 1999 Oct 1; [3]: 797-
802.

MJ Richards, DA Buchler: errors in reproducing pelvic radiation. Int J Radiat Oncol Biol Phys 2; 1977: 1017-9 A. Rizzoti, R. Compri, and Garusi GF: Doseevaluation to patients irradiated by
60Co beams, by means of direct measurements on the incident and on the exit surfaces. Radiother Oncol 3; 1985: 279-83.

- JM Robertson, RK Ten Haken, MB Hazuba, et al: Dose escalation for non-small cell lung cancer using conformal radiation therapy. Int J Radiat Oncol
Biol Phys 37; 1997: 1079-1085.

 D. W.O. Rogers: The role of Monte Carlo simulation of electron transport in radiation dosimetry. Int J Appl Radiat Isot 42; 1991: 965-974.

D.W.O. Rogers and A. F. Bielajew: The use of EGS for Monte Carlo calculations in medical physics. Report PXNR-2692 (Ottawa: NRCC); 1984.

DWO Rogers and AF Bielajew: Difference in electron depthdose curves calculated with EGS and ETRAN and improved energy-range relationships. Med
Phys 13; 1986: 687-694.

- D.W.O. Rogers and A. F. Bielajew: Monte Carlo Techniques of Electron and Photon Transport for Radiation Dosimetry, in The Dosimetry of Ionizing Radiation Vol. 3, ed K. Kase, B. Bjamgard, and F. Attix (San Diego, CA: Academic); 1990: pp 427-539 ..

 D. W.O. Rogers, A.F. Bielajew, T.R. Mackie and S.S. Kubsad: The OMEGA project treatment planning for electron-beam radiotherapy using Monte Carlo techniques [abstract]. Phys Med Biol35; 1990: 285.

DWO Rogers, GM Ewart, AF Bielajew and G. Dyk: Calculations of Electron Contamination in a Co-60 Therapy Beam. Proc IAEA Int Symp on Dosimetry in
Radiotherapy vol I (Vienna: IAEA); 1988: pp 303-312.

DWO Rogers, BA Faddegon, GX Ding, CM. Ma, J. We and R. R. Mackie:
BEAM: Monte Carlo code to simulate radiotherapy treatment units. Med Phys
22; 1995a: 503-524.

 D.W.O. Rogers, C-M. Ma, G.X. Ding and B. Walters: BEAM users manual.

 Report PIRS-0509 (A) (Ottawa: NRCC); 1995b

<Desc / Clms Page number 235>

 - 1.1. Rosen, R. G. Lane, S. Morrill and J. A. Belli: Treatment plan optimization using linear programming. Med Phys 18; 1991: 141-152.

CS Ross, DH Hussey, EC Pennington, et al: of movement of intrathoracic neoplasms using ultrafast computerized tomography. Int J Radiat
Oncol Biol Phys 18; 1990: 671-677.

 # S.N. Rustgi, A. Rustgi K., S.B. Jiang and K.M. Ayyangar: Dose disturbance caused by high density inhomogeneities in small beams in stereotactic radiosurgery. Phys Med Biol 43; 1998: 3509-3518.

F. Salvat, JM Fernandez-Vera, J. Baro and J. Sempau: PENELOPE, year
Algorithm and Computer Code for Monte Carlo Simulation of Electron-Photon
Showers (Madrid: Information Tecnicos Ciemat); 1996.- MJ Samson, JR van Sôrnsen of Koste, JC de Boer, et al: An analysis of anatomical landmark mobility and deviations in radiotherapy for lung cancer. Int J Radiat Oncol Biol Phys 43; 1999: 827-832.

 - P.V. Sankar, O. Nalcioglu, and J. Slansky: Undersampling errors in region-of-interest tomography. IEEE Trans Med Imaging 1; 1982: 168-173.

Sauer, O.A., Shepard, D., and Mackie, T. R.: Application of constraint optimization to radiotherapy planning. Med Phys 1999 Nov; 26 [11]: 2359-66.

- JE Schewe, JM Balter, KL Lam, et al: Measurement of patient setup errors using port films and a computer-aided graphical alignment tool. Med Dosim 21;
1996: 97-104.

- M. Schiebe, and W. Hoffmann: CT-based virtual simulation using the
AdvantageSim system Description of reliability and accuracy Strahlenther Onkol
2000 Aug; 176 (8).

 - S. Shaley: Treatment verification using digital imaging. In: A.R. Smith, ed.

 Radiation therapy physics, Berlin: Springer; 1995: pp 155-173.

 M. B. Sharpe, A. A. Jaffray, J. J. Battista, and P. Munro: Extrafocal radiation: A unified approach to the prediction of beam penumbra and output factors for megavoltage x-ray beams. Med Phys 22; 1995: 2065-2074.

 I H. Shaw: A quasirandom approach to integration in Beyasian statistics. Ann Statist 16; 1988: 895-914.

- Sheikh-Bagheri: Monte Carlo study of photon beams from medical linear accelerators: optimization, benchmarking and spectra, PhD Thesis, Carleton
University, Canada. ; 1998.

<Desc / Clms Page number 236>

H. Shirato, S. Shimizu et al. : Four dimensional treatment planning and fluoroscopy real-time tumor tracking radiotherapy for moving tumor. Int J Radiate Oncol Biol Phys 2000 Sep 1; 48 (2): 435-42.

S. Shimizu, H. Shirato, K. Kagel, T. Nishioka, X. Bo, H. Dosaka-Akita, S.

Hashimoto, H. Aoyama, K. Tsuchiya and K. Miyasaka: Impact of respiratory movement on the computed tomographic images of small lung tumors in three-dimensional (3D) radiotherapy. Int J Radial Oncol Biol Phys 45 [5]; 2000:
1127-1133.

- K.R. Shortt, C.K. Ross, A.F. Bielajew and D.W.O. Rogers: Electron beam dose distributions near standard inhomogeneities. Phys Med Biol 31; 1986: 235-249.

- K. E. Sixel and B. A. Faddegon: Calculation of x-ray spectra for radiosurgical beams. Med Phys 22; 1995: 1657-1661.

J.M. Slater, J.O. Archambeau, D. W. Miller, M.I. Notarus, W. Preston, and J.D.

Slater: The proton treatment center at Loma Linda Medical University Center: Rationale for and description of its development. Int J Radial Oncol Biol Phys
22; 1992: 383-389.

- TD Solberg, JJ DeMarco, Holly FE, JB Smathers and AF DeSalles: Monte
Carlo treatment planning for stereotactic radiosurgery. Radiother Oncol 49;
1998: 73-84.

- WJT Spyra, A. Faridani, KT Smith, and EL Ritman: Computed tomographic imaging of the coronary arterial tree-use of local tomography. IEEE Trans Med
Imaging 1; 1982: 168-173.

- M. Stasi, B. Baiotto, F. Palamara, P. Gabriele, and G. Scielzo: Dosimetric characterization of a multileaf collimator. (abstract) Radiol Med (Torino) 97 (5); 1999: 382-8.

J.C. Stroom, J.C. de Boer, H. Huizenga, et al: Inclusion of Geometrical uncertainties in radiotherapy treatment planning by means of coverage probability. Int J Radiat Oncol Biol Phys 43; 1999: 905-919.

J.A. Stryker, J. Shafer, and R. E. Beatty: Assessing the accuracy of daily prostate irradiation using electronic imaging. Br J Radiol 72 [858]; 1999: 579-83.

- M. Tatcher and B.E. Bjarngard: Head-scatter factors in rectangular photon fields.

 Med Phys 20; 1993: 205-206.

- A. Torn, and S. Vitanen: Topographical global optimization using pre-sampled points. J Global Optimization 5; 1994: 267-276.

<Desc / Clms Page number 237>

- AF Thornton, TJ Hegarty, Ten Haken RK: Three-dimensional treatment planning of astrocytomas: a dosimetric study of cerebral irradiation. Int J Radial
Oncol Biol Phys 20; 1991: 1309-1315.

- A. Torn, M. M. Ali and S. Vitanen: Stochastic global optimization: problem classes and solution techniques. J Global Optimization 14; 1999: 437-447.

- JR Treurniet and Rogers DWO: BEAM and DOSXYZGUI user's manual
Report PIRS-0623 (Ottawa: NRCC) - ST Treves, KD Mitchell, IH Habboush: Three-dimensional image alignment, registration and merger. QJ Nucl Med 42; 1998: 83-92.

- THE. Tsalafoutas, S. Xenofos, A. Papalexopoulos, S. Nikoletopoulos: Dose calculations for asymmetry of fields defined by independent collimators using symmetry of field data. Br J Radiol 73 (868); 2000: 403-9.

 # J.E. Turner, A.A. Wright and R.N. Hamm: Review article: Monte Carlo primer for health physicists. Health Phys 48; 1985: 717-733.

Udale: Monte Carlo investigation of surface doses for broad electron beams.

 Phys Med Biol 33; 1988: 939-954.

 @ Udale-Smith: A Monte Carlo investigation of high-energy electron beams used in radiotherapy, PhD Thesis, Leeds University; 1990.

 M. Udale-Smith: Carlo calculations of electron beam parameters for Philips linear accelerators. Phys Med Biol 37; 1992: 85-105.

 # J. van Dam, G. Marinello: Methods for in vivo dosimetry in external radiotherapy.

 Estro Booklet No. 1 on physics for clinical radiotherapy. Guarantor: Leuven; 1994.

J. van Dam, C. Vaerman, N. Blanckaert, G. Leunens, A. Dutreix, and E. van der
Schueren: Are port films reliable for in vivo exit dose measurements?
Radiother Oncol 25; 1992: 67-72.

- W. van der Zee: Photon beam radiotherapy planning using Monte Carlo systems,
M Sc Thesis, Utrecht University; 1996.

- W. van der Zee, J. Welleweerd: Calculating photon beam characteristics with
Monte Carlo techniques. MedPhys 1999 Sep; 26 [9]: 1883-92.

A. van Esch, R. Bogaerts, G. J. Kutcher, and D. Huyskens: Quality assurance in radiotherapy Radiother Oncol 2000 Jul; 56 [1]: 109-15.

<Desc / Clms Page number 238>

- M. van Herk, P. Remeijer, C. Rasch, et al: The probability of correct target dose-dosing population histograms for deriving treatment margins in radiotherapy. Int J Radiat Oncol Biol Phys 47; 2000: 1121-1135.

J. van de Steene, F. van den Heuvel, A. Bel, et al: Electronic portai imaging avec une-ligne de correction d'ordination in thoracic irradiation: Clinical evaluation. Int J
Radiat Oncol Biol Phys 40; 1998: 967-976.

- G. van Tienhoven, J. Lanson, D. Crabeels, S. Heukelom, B. Mijnheer: Accuracy in tangential treatment with a portal imaging study. Radiother Oncol 22; 1991: 317-322.

A.A. van't Veld, P. van Luijk, F. Praamstra, P.C. van der Hulst: Slit x-ray beam primary dose profiles determined by analytical transport of Compton recoil electrons. Med Phys 27 (5); 2000: 923-34.

- SM Vatnitsky, RWM Schulte, R. Galindo, HJ Meinas, and DW Miller:
Radiochromic film dosimetry for verification of dose distributions delivered with proton-beam radiosurgery. Phys Med Biol 42; 1997: 1887-1898.

- D. Verellen, V. Vinh-Hung, P. Bijdekerke, F. Nijs, N. Linthout, A. Bel, G.

Storme: Characteristics and clinical application of a treatment simulator with
CT option. Radiother Oncol 50 [3]; Mar 1999: 355-66.

F. Verhaegen, AE Nahum, S. Van de Putte and Y. Namito: Carlo Modeling of Radiation Therapy kV x-ray units. Phys Med Biol 44; 1999: 1767-
1789.

T. Vuong, W. Parker, H. J. Patrocinio, P. Fava, D. Parmar, M. D. Evans, and C.R.

Freeman: An alternative mantle irradiation technique using D CT-based treatment planning for female patients with Hodgkin's disease. Int J Radiat
Oncol Biol Phys 2000 Jun 1; 47 [3]: 739-48.

 W. Wagner, K. Ostkamp, and U. Niewohner-Desbordes: A simple method for transposing MRI information to simulation films. Technical note [abstract].

 Strahlenther Onkol 2000 May; 176 [5]: 235-8.

 S. Wallace and B.J. Allen: CT-based 3D Monte Carlo radiation therapy treat- ment planning. Australas Phys Eng Sci Med 21; 1998: 41-50.

- L. Wang, C. S. Chui and M. Lovelock: A patient specifies Monte Carlo dose calculation method for photon beams. Med Phys 25, 1998: 867-878.

- S. Webb: Optimization of conformal radiotherapy dose distributions by simulated annealing. Phys Med Biol 34; 1989: 1349-1370.

<Desc / Clms Page number 239>

- S. Webb: Optimization by simulated annealing of three-dimensional conformal treatment planning for radiation fields defined by a multileaf collimator. Phys MedBiol 36; 1991: 1201-1226.

- S. Webb: The Physics of Three-dimensional Radiation Therapy (Bristol: Institute of Physics Publishing); 1993: 135-176.

- S. Webb: The Physics of Conformal Radiotherapy: Advances Technology (Bristol: Institute of Physics Publishing); 1997: pp 1-95.

C. Weltens, J. Menten, M. Feron, E. Bellon, P. Demaerel, F. Maes, et al:
Interoberver variations in Gross Tumor delineation of brain tumors on CT and impact on MRI. Radiother Oncol 48; 1998: 14.

- C. Westbrook, J. Gildersleve, J. Yarnold: Quality assurance in daily treatment procedure: Patient movement during tangential fields treatment. Radiother Oncol 22; 1991: 299-303.

X. Wu, Y. Zhu, J. Dai and Z. Wang: Selection and determination of beam weights based on genetic algorithms for conformal radiotherapy treatment planning. Phys Med Biol 45; 2000: 2547-2558 - Xingen Wu and Yunping Zhu: A global optimization method for three-dimensional conformai radiotherapy treatment planning. Phys Med Biol46;
2001: 107-119.

D. Yan, J. Wong, Vicini F., et al: Adaptive Modification of Treatment and Treatment. Int J Radiat Oncol
Biol Phys 38; 1997: 197-206.

- M. K. Yu, R. S. Sloboda, and F. Mansour: Measurement of photon beam backscatter from collimators to the beam monitor chamber using target-current-pulse-counting and telescope techniques. Phys Med Biol 41; 1996: 1107-1117.

Y. Yu, M.C. Schell and J.B. Zhang: Decision theoretic steering and genetic algorithm optimization: Application to stereotactic radiosurgery treatment planning. Med Phys 24; 1997: 1742-1750.

- C.X. Yu: Intensity-modulated arc therapy with dynamic multileaf collimation: an alternative to tomotherapy. Phys Med Biol 40 (9); 1995: 1435-49.

K. Yuen, MS Al-Ghazi, CL Swift, and CA White: A Practical Method for the Computation of Multileaf Collimators.
Nov ; 26 [11]: 2385-9.

<Desc / Clms Page number 240>

Zaider, MJ Zelefsky, EK Lee, KL Zakian, HL Amols, J. Dyke, G. Cohen,
YC. Hu, AK Endi, CS. Chui and JA Koutcher: Treatment planning for prostate implants using magnetic-resonance spectroscopy imaging. Int J. Radiat
Oncol Biol Phys 47 [4]; 2000: 1085-96.

 - G. G. Zhang, D. W.O. Rogers, J. E. Cygler and T. R. Mackie: Effects of changes in stopping-power ratios with field size on electron beam relative output factors.

 Med Phys 25; 1998: 1711-1716.

GG Zhang, DWO Rogers, JE Cygler and TR Mackie: Monte Carlo Investigation of Electron Output Factors versus Size of Square Cut-Out. Med Phys
26; 1999: 743-750.

- R. Zielinski: A stochastic estimate of the structure of multi-external problems.

 Math Program 21; 1981: 348-356.

Claims (4)

 1) Integrated radiotherapy device for simultaneously irradiating a single subject characterized in that it comprises in combination a plurality of synchronous sources of orthovoltage photon beams X 1 (1, 2, 3, 5, 5), mounted on rotary gantries for 4-D radioscanography-guided multi-beam isotropic radiotherapy and a plurality of orthovoltage photon beam sources (8,9, 10,11, 12, and 13) mounted on stationary gantries, as well as Imaging means generating 4-D multi-image image data as well as real-time, real-time measurement of the exact irradiation dose for the 3-D clinical planning, treatment and surveillance simulation multibeam delivery of the spokes with great ballistic flexibility to adjust immediately placement errors, voluntary and involuntary movements, rotation on itself by rappelling orto the anatomical landmarks and the poor targeting of the goal of radiotherapy from the data acquisition system (DAS) scan data (14) and the means for controlling the constancy of the irradiation parameters. 2) Device according to claim 1 characterized in that the rotating gantry has an odd number of sources of X-rays, only one of which is intended for CT imaging, secured to the same number of networks of the detectors of the X-rays (15, 16, 17, 18, 19, and 20), diametrically opposed, and mounted to rotate together, so that a given number of the tubes perform the same number of simultaneous helical scan phases, radioscopic examination or radioscanoscopic instantaneous examination, combined with a simultaneous movement of continuous translation in both directions of the axis of a patient support 24, specifically intended for multiphasic clinical monitoring and surveillance. 3) Device according to claim 1 characterized in that the stationary gantries are in the form of half-arches arranged in pairs (15, 16, 17, 18, 19 and 20) with the possibility of lateral tilt of the pair from and other of the central position, and independently of each other, so that a pair can shift <Desc / Clms Page number 242>  at most its two half-arches of 30 and at the end of each half hoop is mounted an X-ray tube (8, 9, 10, 11, 12 and 13). 4) Device according to claim 3 characterized in that it comprises six semicircles (15, 16,17, 18,19 and 20) at the ends of which are arranged six tubes (8,9, 10,11, 12, and 13) stationary with anodes of fixed preferences on both sides of the plane (x, y) of the stand inside which five rotating anode tubes (1, 2, 3, 4 and 5) rotate, the rotating assembly around one and the same isocenter 1 and the patient support (24).