FR3004654A1 - OPTIMIZATION OF ADAPTIVE RADIOTHERAPY PLANNING - Google Patents

Abstract

The invention relates to a method of optimizing a treatment planning in adaptive radiotherapy of at least one tumor structure to be treated, a method implemented by an optimization device, said method comprising, during a current treatment session, at least one acquisition of a tracking image comprising a representation of said at least one minus a tumor structure to be treated.

Description

Optimization of adaptive radiotherapy planning. Field of the Invention The invention relates to the processing of medical imaging data. The invention relates more particularly to a medical data processing method for optimizing planning in adaptive radiotherapy using cumulative dose drift and prediction of the final dose delivered. It is recalled that radiotherapy is a treatment method that uses radiation to destroy cancer cells. The purpose of irradiation is to destroy tumor cells while sparing healthy peripheral tissue. To allow irradiation, a linear electron accelerator is used which produces an irradiation beam. 2. Prior art The history of radiotherapy is relatively recent. Radiation therapy becomes more efficient as the evolution of technical progress evolves. Thus, thanks to the implementation of a scanner, the planning of three-dimensional radiotherapy treatments became possible, which represented a major step forward compared to two-dimensional treatments. CT-based treatments allow radiation oncologists to more accurately determine the radiation dose distribution using CT images of the patient's anatomy. Imaging techniques have led to the development of image-guided radiotherapy (IGRT), which allows better control of the position of the area to be treated as treatment progresses. Innovations in treatment devices such as the appearance of multi-blade collimators have led to the introduction of intensity-modulated conformal radiotherapy (IMRT), which makes it possible to adapt irradiation more precisely to the shape of the organs. treat. It has been possible to visualize and treat tumors more effectively, resulting in a better prognosis for patients, better preservation of healthy organs and fewer side effects. The optimizations developed over the past few years are, however, unfortunately limited in their clinical effectiveness by the fact that, while all planning is based on a single planning image (also called TDM image), anatomical displacements and deformations, which are not classically detectable by treatment devices, may occur during irradiation sessions delivered during treatment weeks. The dose administered may no longer correspond to the planned dose, exposing the patient to a risk of tumor underdosage (and therefore recurrence) and overdose in healthy tissue (and therefore toxicity). To take into account these geometrical uncertainties, image-guided radiotherapy (IGRT) makes it possible to locate the tumor target under the treatment device at the time of the irradiation session. Due mainly to its complexity, the use of the IGRT is still limited. In addition, changes in the position, volume and interrelation of targets and organs at risk are not taken into account, in particular because the dosimetric quality indexes (or optimization criteria for a new planning) in progress treatment are non-existent, poorly defined or without interaction with subsequent irradiation sessions. 3. Summary of the invention The proposed technique does not have these disadvantages of the prior art. More particularly, the proposed technique relates to a method for optimizing an adaptive radiotherapy treatment planning of at least one tumor structure to be treated, a method implemented by an optimization device, said method comprising, during of a current treatment session, at least one acquisition of a tracking image comprising a representation of said at least one less a tumor structure to be treated. Such a method comprises: an estimation phase, for said current session, of a cumulative dose during the treatment, for said at least one less a tumor structure to be treated; a phase of predicting a dose received at the end of treatment as a function of at least one piece of data collected during treatment; an evaluation phase of cumulative doses between a first treatment session and said current treatment session, delivering a data representative of a cumulated dose drift; a generation phase of at least one new treatment constraint when said data representative of a cumulated dose drift exceeds a determined ceiling. In at least one other embodiment, the invention relates to a device for optimizing a treatment planning in adaptive radiotherapy of at least one tumor structure to be treated, said device comprising, acquisition means of a tracking image comprising a representation of the at least one minus one tumor structure to be treated. Such a device characterized in that it comprises: estimation means, for said current session, a cumulative dose during treatment, for said at least one less a tumor structure to be treated; means for predicting a dose received at the end of treatment as a function of at least one piece of data collected during treatment; cumulative dose evaluation means between a first treatment session and said current treatment session, delivering a data representative of a cumulated dose drift; means for generating at least one new processing constraint when said data representative of a cumulative dose drift exceeds a determined ceiling. According to a preferred implementation, the various steps of the methods according to the invention are implemented by one or more software or computer programs, comprising software instructions intended to be executed by a data processor of a relay module according to the invention. invention and being designed to control the execution of the various process steps. Accordingly, the invention is also directed to a program that can be executed by a computer or a data processor, which program includes instructions for controlling the execution of the steps of a method as mentioned above.

This program can use any programming language, and be in the form of source code, object code, or intermediate code between source code and object code, such as in a partially compiled form, or in any other form desirable shape.

The invention also provides a data carrier readable by a data processor, and including instructions of a program as mentioned above. The information carrier may be any entity or device capable of storing the program. For example, the medium may comprise storage means, such as a ROM, for example a CD ROM or a microelectronic circuit ROM, or a magnetic recording medium, for example a diskette (floppy disc) or a disk hard. On the other hand, the information medium may be a transmissible medium such as an electrical or optical signal, which may be conveyed via an electrical or optical cable, by radio or by other means. The program according to the invention can be downloaded in particular on an Internet type network. Alternatively, the information carrier may be an integrated circuit in which the program is incorporated, the circuit being adapted to execute or to be used in the execution of the method in question. According to one embodiment, the invention is implemented by means of software and / or hardware components. In this context, the term "module" may correspond in this document as well to a software component, a hardware component or a set of hardware and software components. A software component corresponds to one or more computer programs, one or more subroutines of a program, or more generally to any element of a program or software capable of implementing a function or a program. set of functions, as described below for the module concerned. Such a software component is executed by a data processor of a physical entity (terminal, server, gateway, router, etc.) and is capable of accessing the hardware resources of this physical entity (memories, recording media, bus communication cards, input / output electronic cards, user interfaces, etc.).

In the same way, a hardware component corresponds to any element of a hardware set (or hardware) able to implement a function or a set of functions, as described below for the module concerned. It may be a hardware component that is programmable or has an integrated processor for executing software, for example an integrated circuit, a smart card, a memory card, an electronic card for executing a firmware ( firmware), etc. Each component of the previously described system naturally implements its own software modules. The various embodiments mentioned above are combinable with each other for the implementation of the invention. 4. Figures Other features and advantages of the invention will appear more clearly on reading the following description of a preferred embodiment given as a simple illustrative and nonlimiting example, and the appended drawings, among which: - Figure 1 shows a synoptic of the proposed technique; FIG. 2 illustrates an example of calculation of the cumulative dose; FIG. 3 illustrates the evaluation of a dose distribution on an information retrieval device; Figure 4 illustrates a side-by-side 3D evaluation of two dose distributions on an information retrieval device; FIG. 5 illustrates a superimposed 2D evaluation of two dose distributions (solid line for the planned dose and dashed lines for the predicted dose) on an information retrieval device; FIG. 6 illustrates a superposition of the HDVs calculated during each treatment session; FIG. 7 illustrates a 3D evaluation of two subtracted dose distributions (subtraction of the predicted dose matrix from the planned dose matrix) on an information retrieval device; Figure 8 describes a device for implementing the proposed technique. 5. Description of an embodiment 5.1. Recall of the principle of the technique The proposed technique is to optimize the IGRT in routine practice to improve cancer healing. In the embodiment presented later, cancer of the prostate was chosen as a model because of its frequency, the major role of radiotherapy for these cancers and the mobility of the prostatic target posing a therapeutic problem. Of course, the proposed technique is applicable to any other type of cancer for which radiotherapy treatment is possible.

The proposed technique can be implemented in the form of a reliable tool for decision support that takes into account the deformable nature of the organs. The proposed technique is based on an original and operational approach for "adaptive radiotherapy" defining decision support indices for intraoperative re-planning of treatment.

A system, for example a system consisting of hardware devices and software modules, which implements the proposed technique aims to optimize dose planning in adaptive radiotherapy during treatment. Its initial input parameters are the patient's own data and the parameters of the dosimetric planning: target volumes, organs at risk, margins, dose per fraction ... At each session, such a system calculates the cumulative dose during previous sessions and, by geometric and statistical algorithms, evaluates the predicted dose at the end of treatment and drift compared to the planned dose. The system checks the risks of non-compliance with the radiotherapy quality criteria at the end of the treatment and deduces new optimization constraints for the adaptive planning of the following treatment sessions. The general principle implemented is as follows: At each session, the system calculates the cumulated dose during the previous sessions and, by geometric and statistical algorithms, evaluates the predicted dose at the end of the treatment and the drift with respect to the planned dose. The system checks the risks of non-compliance with the radiotherapy quality criteria at the end of the treatment and deduces new optimization constraints for the adaptive planning of the following treatment sessions. Considering the second treatment session, the solutions are presented globally in Figure 1. The planning image (usually a computed tomography image (CT or CT image)), acquired prior to treatment to define the radiation ballistics that will be used during the different treatment sessions (for example 5 treatment sessions per week for 8 weeks). In this image, the expert has delineated the structures of interest (the different target volumes and the organs at risk). From these contours, the treatment planning system (TPS) optimizes the ballistics according to dosimetric criteria. A treatment ballistics is thus obtained, as well as the corresponding three-dimensional dose matrix. a follow-up image (usually CT or cone beam CT (or Cone Beam Computed Tomography image) corresponding to the 2nd treatment session, which is acquired on a regular basis (daily, weekly, ...) during treatment to visualize the patient's anatomy and / or reposition the tumor before an irradiation session.The proposed technique is divided into three main phases: the estimation of the cumulative dose during treatment: for each session Depending on the acquired follow-up image, the dose received by the tissues is estimated and reported on the planning image, and can be accumulated locally to track cumulative dose during treatment. information collected during treatment (anatomical and dosimetric), as well as possibly data from other patients, the dose received at the end of treatment is predicted. During the course of the treatment, an evaluation of the cumulative doses (cumulative dose of the 1st to the first session and the dose predicted at the end of the treatment) is carried out. This evaluation makes it possible in particular to evaluate a possible drift of the cumulative dose compared to the planned dose. In case of significant drift, new ballistic optimization constraints are proposed to allow a rescheduling of the ballistics of treatment. 5.1.1. Cumulative Dose Calculation Cumulative Dose Calculation involves a step of resetting images between the tracking image and the planning image. An example of a cumulative dose process is illustrated in Figure 2. The first step is the calculation of the daily dose matrix, that is to say the dose received by the tissues during the ii th calculated treatment session. on the corresponding tracking image. This calculation takes into account the repositioning of the patient used during the treatment sessions. Two different methods are used, depending on the location considered (pelvis, ENT ...) and the modality used for the follow-up images: recalculation of the dose following the treatment ballistics: if the external contour of the patient is significantly modified ( for example, in the case of tumor malignancy in ENT) and if the follow-up image allows it (CT image, even CT with cone beam (the use of the latter being more complex)), the dose matrix can be calculated by the treatment planning system according to the treatment ballistics used. direct transposition of the planned dose matrix: if the external contour of the patient is little modified (case of the pelvis), the planned dose matrix can be directly transposed on the follow-up image (that is to say displaced according to the repositioning of the patient and possibly interpolated). The estimation of the deformations of the organs between the planning image and the tracking image is then carried out. In the approach developed, this estimation is carried out in 3 main stages: propagation of the contours of the planning image to the tracking image. To do this, an image registration process between the planning image and the tracking image is used. It provides a field of deformations that reflects the deformations of the organs between the two images. The deformation field obtained is then used to deform the contours associated with the planning image so that they correspond to the anatomical configuration of the tracking image. contour correction / validation. The contours obtained are not always satisfactory, a step of manual correction by an expert may be necessary. Contour registration: Deformations between the tracking image and the planning image are estimated from the contour information. First, three-dimensional distance maps are generated for both sets of contours. These maps represent, for each voxel of the 3D image, the distance to the nearest organ. These distance maps are then recalibrated, providing a new field of movement. The strain field obtained is then used to propagate the daily dose, corresponding to the tracking image, to the planning image. This step is performed by deforming the daily dose matrix according to the deformation field obtained and by applying a trilinear interpolation. Thus, if this process is applied for all the tracking images, so many doses propagated on the planning image are obtained. These different dose matrices are all associated with the same anatomical reference, they can be added together (they are cumulative). Finally, a cumulative dose, corresponding to the dose received from sessions 1 to i, is therefore obtained. 5.1.2. Projection of the dose at the end of treatment During a treatment in n sessions, from the different cumulative doses obtained during the first i sessions, the system allows to predict the dose "really" delivered at the end of treatment. The system therefore models the cumulative doses from i to n, in order to sum them to cumulative doses of 1 to i (obtained by the cumulative dose calculation step). In order to extrapolate the last (n-i) sessions to be performed, several highly variable complexity scenarios can be used: - use of the planned dose (before the start of treatment) to complete the missing sessions. postponement of the last session performed (ie last cumulative dose obtained) on the remaining sessions. average of the sessions already done. use of a probabilistic model based on previous sessions, or based on data from the study of a patient population (eg frequency and magnitude of organ movements). use of a mechanical model established using the kinetics of the volumes is observed on the previous sessions (starting from the images of follow-up), or based on the study of a population of patients (ex: kinetics of the volumes during a radiotherapy treatment ...) use of more complex models from the fields of biology, finance (eg predictive analysis method) ... integrating data specific to patients (age, weight, pathology, tumor volume localization, and any other diagnostic data ...). use of hybrid models (probabilistic, mechanical, etc ...) The system also offers the possibility of other scenarios such as the most optimistic (taking the most favorable dose distribution) or the most pessimistic (taking again the most unfavorable dose distribution ). The system then summarizes these n dose distributions (3D matrices) defined in the same coordinate system (reference of the planning image, or of the last acquired tracking image) to obtain a single distribution. During this summation, depending on the dose at each fraction, the system can correct the dose received by each voxel at each fraction by a corrective factor calculated using known biological model (LU: Linear Quadratic, LQL: Linear Quadratic Linear, LUC: linear quadratic cubic ...) or user-defined. 5.1.3. Evaluation of cumulative doses (at the fraction and at the end of treatment) Various tools are used to evaluate the different dose distributions obtained in the previous steps (planned dose, distorted dose, that is to say fraction, dose cumulated and predicted dose). These tools are of different natures, they can be in: three dimensions, the clinician can thus observe in space the dose distribution studied and thus identify by a color system (LUT) the hot spots (over-dosage) and the cold spots (underdosing). two dimensions, such as one-dimensional dose-volume histograms as geometric, dosimetric and biological indicators. For example statistics on inter-organ distances (eg rectum-PTV), organ volume, maximum / minimum / average doses received by each organ or received by so many percent (or cubic centimeter) of the volume of the organ. organ ... are useful to the user in order to quantify the quality of the treatment. Tolerances registered by users or from the bibliography allow the system to trigger different levels of alerts. These tolerances are either absolute (eg "the maximum dose per fraction for the marrow must be less than 1Gy", or "the maximum dose for the entire treatment for the marrow must be less than 45 Gy" ...) relative (eg "the maximum dose for the entire treatment for the marrow should be less than 110% of the maximum tolerated dose when planning"). 5.1.4. Evaluation of the drift The dosimetric drift is estimated in two ways: Observed drift: subtraction between the cumulated dose at the session i and the corresponding planned dose (that is to say related to the number of sessions performed). Predicted drift: subtraction between the predicted dose at the end of treatment and the planned dose.

The matrices of doses obtained are studied according to the modes described above (average, maximum values ... of differences in dose). 5.1.5. Adaptation of treatment plan If the drift between the planned dose distribution and the predicted dose distribution becomes too great, a new treatment plan must be established. For this new dosimetric constraints are established, based on the initial constraints, but taking into account the dose already received by the patient. The basic principle is to subtract the dose already received from the initial constraints. The definition of new volumes can then be required. For example, if at planning the maximal dose at the marrow should be less than 45 Gy, and that during the sessions preceding the current session, part of the marrow has already received 30 Gy, then the system will produce this new volume ( It is a fictitious volume resulting from the isodose 30 Gy, useful for the new optimization.) And will recommend there a stress of 15 Gy. Similar mathematical operations can be applied to other indexes (ex: average dose ... ) to generate new dosimetric constraints to correct the dose drift observed on the sessions performed on the remaining sessions. 5.2. Implementing device The method described is preferably implemented by means of a dedicated computing device described in relation with FIG. 5. Such equipment comprises a memory 41, a processing unit 42 equipped for example with a microprocessor, and driven by the computer program 43, implementing at least a part of the method as described. In at least one embodiment, the invention is implemented in the form of an application installed on a planning device. Such a device comprises: means for estimating, for said current session, a cumulative dose during the treatment, for said at least one less a tumor structure to be treated; means for predicting a dose received at the end of treatment as a function of at least one piece of data collected during treatment; cumulative dose evaluation means between a first treatment session and said current treatment session, delivering a data representative of a cumulated dose drift; means for generating at least one new processing constraint when said data representative of a cumulative dose drift exceeds a determined ceiling. These means are implemented via dedicated software and / or hardware modules. These modules are able to implement the reception and data transmission interfaces present on said device (input interface I and transmission interface T). In at least one embodiment, the described technique can be implemented via a communication network to which a device is connected. In at least one embodiment, the technique described is implemented within a radiotherapy treatment planning console. The invention may also be implemented via a decision support system for optimizing dose planning in adaptive radiotherapy comprising: a patient specific data input module; a data entry module of the radiotherapy treatment; a data entry module for the dose delivered during previous treatment sessions; a cumulative dose assessment module at the current session; a module for evaluating the dose predicted at the end of the treatment; a module for assessing the risks of non-compliance with the radiotherapy quality criteria at the end of treatment; a module for defining new optimization constraints for the adaptive planning of the following treatment sessions.

Within this system, patient-specific data includes any relevant data (eg, age, weight, sex) that allows the patient to be associated with the meta-data of a population. The different input modules receive data automatically or manually from the diagnostic and therapeutic tools during the various treatment sessions or from data from the scientific and medical literature. Diagnostic tools include: medical imaging devices external to treatment (eg, x-ray, CT, ultrasound, scintigraphy, positron emission tomography); imaging devices internal to the treatment (for example, CBCT, ultrasound tracking, video tracking); biological dosages; manual diagnostic tools (eg, palpation, interrogations, clinical examinations). Therapeutic tools include: the physical characteristics of treatment devices (eg cobalt therapy, linear accelerators, tomotherapy, radiotherapy); the geometric characteristics of said treatment apparatus; Moreover, the processing data input module receives: the bi, tri or four dimensional contours (3D + time) of the target volumes to be processed; the bi or tri or four dimensional contours of organs at risk; the dosimetric planning parameters of the current treatment. The input module of the treatment data initially receives the prescription (dose at the session, etc.). The data input module of the treatment receives at each session the individual and inter-individual geometric modifications of the target volumes to be processed and the organs at risk of the patient. The data input module of the treatment receives the data concerning the overall movement of the patient in the treatment apparatus during the various previous treatment sessions.

The data input module of the treatment receives the data relating to the repositioning of the patient in the treatment apparatus during the various previous treatment sessions. The input module of the dose delivered during the previous treatment sessions receives the result of a calculation of the cumulative dose obtained from the data previously described. The calculation of the cumulative dose is obtained from algorithms of rigid or non-rigid registration of organ contours and targets defined on the different imaging modalities.

The calculation of the cumulative dose is obtained from dose propagation algorithms on the contours of contouring of the contours of organs and targets. The cumulative dose evaluation module at the current session receives the result of the cumulative dose previously calculated. The cumulative dose evaluation module at the current session receives the patient's specific data. The predicted dose assessment module at the end of treatment provides predictable dosimetric indicators at the end of treatment (eg, dose-volume, mean dose, mean dose, minimum dose histograms) for each of the organs or targets considered.

The module for evaluating the dose predicted at the end of treatment provides predictable radiobiological indicators at the end of the treatment (for example BED, EUD, TCP, NTCP, etc.) for each of the organs or targets considered. . The module for predicting the cumulative dose at the end of treatment receives cumulative dose indicators at the current session.

The module for predicting the cumulative dose at the end of treatment receives cumulative dose indicators individually during the previous treatment sessions. The cumulative dose prediction module at the end of the treatment receives the 3D dose matrix and 3D mapping of the organs and targets considered during the current session.

The cumulative dose prediction module at the end of treatment receives the 3D dose matrix and 3D mapping of organs and targets during previous treatment sessions. The cumulative dose prediction module at the end of the treatment performs probabilistic calculations on the cumulative dose index for the following treatment sessions. The cumulative dose prediction module at the end of the treatment performs probabilistic calculations on the 3D matrices of dose and mapping of organs and targets for the following treatment sessions.

The module for predicting the cumulative dose at the end of the treatment performs the said probabilistic calculations according to different predictive models (for example, trend analysis). The cumulative dose prediction module at the end of treatment receives data from the medical literature regarding national or international recommendations for specification of the prescribed dose (eg RTOG, GETUG, consensus conferences). The module for predicting the cumulative dose at the end of treatment defines, on the basis of the previously defined information, an estimate of the risk of non-compliance with the quality indices.

The module for defining new optimization constraints receives the estimate of the risk of non-compliance with the quality indexes. The module defining new optimization constraints receives the dose accumulation quality index for the following treatment sessions. The module defining new optimization constraints receives the same information as that defined for patient-specific data entry modules and radiotherapy treatment. The module for defining new optimization constraints defines new optimization constraints for the definition of a modified treatment plan.

Claims (3)

REVENDICATIONS1. A method for optimizing a treatment planning in adaptive radiotherapy of at least one tumor structure to be treated, a method implemented by an optimization device, said method comprising, during a current treatment session, at least an acquisition of a follow-up image comprising a representation of the said at least one tumor structure to be treated, characterized in that it comprises: an estimation phase, for the said current session, of a cumulative dose during treatment, for said at least one less tumor structure to be treated; a phase of predicting a dose received at the end of treatment as a function of at least one piece of data collected during treatment; a cumulative dose evaluation phase between a first treatment session and said current treatment session, delivering a data representative of a cumulated dose drift; a generation phase of at least one new treatment constraint when said data representative of a cumulated dose drift exceeds a determined ceiling. 2. Device for optimizing treatment planning in adaptive radiotherapy of at least one tumor structure to be treated, said device comprising tracking image acquisition means comprising a representation of said at least one minus one tumor structure to be treated, device characterized in that it comprises: means for estimating, for said current session, a cumulative dose during the treatment, for said at least one less a tumor structure to be treated; predicting a dose received at the end of treatment based on at least one piece of data collected during treatment; cumulative dose evaluation means between a first treatment session and said current treatment session, delivering a data representative of a cumulated dose drift; means for generating at least one new treatment constraint when said datum representative of a cumulated dose drift exceeds a determined ceiling. 3. Computer program product downloadable from a communication network and / or stored on a computer-readable medium and / or executable by a microprocessor, characterized in that it comprises program code instructions for the execution of a method according to claim 1 when executed on a computer.