JP4347651B2 - Multi-modality imaging method and apparatus - Google Patents

Description

  The present invention relates generally to an imaging system capable of scanning an object with multiple modalities, and more particularly to a multi-modality system in which each modality has a different field of view.

  The present invention relates to a multimodal imaging system capable of scanning using different modalities such as, but not limited to, positron emission tomography (PET) and computed tomography (CT). The difference between multi-mode and multi-modality is that multi-mode systems are used to scan in different modes (eg, fluoro mode and tomosynthesis mode), while multi-modal systems (Multi-modality system) is the point used to scan with different modalities (eg CT and PET). The advantages of the present invention are believed to be useful in all multi-modality imaging systems such as, but not limited to, CT / PET imaging systems.

  At least some multi-modality systems have different fields of view (FOV) for different modalities. For example, a CT / PET system may have a CT FOV that is smaller than a PET FOV, and under some scanning conditions, a portion of the patient extends beyond the area measured by the CT detector. It may lead to image artifacts and incomplete representation of the imaged object. Several known methods have been published that deal with artifact reduction, but do not address imaging of the part of the patient located outside the CT FOV.

  For example, in a multi-modality system such as an integrated PET-CT system, registration is essentially aligned between a PET image acquired by the system and a CT image. Since the patient remains lying on the same table during both the PET acquisition and CT acquisition, the patient is in a consistent position and orientation during these two acquisitions and correlates the CT and PET images. The process of adding and synthesizing is greatly simplified. This makes it possible to provide attenuation correction information for reconstruction of the PET image using the CT image, and the image reader is presented to the anatomical information presented in the CT image and the PET image. It is possible to easily correlate the function information. However, it would be desirable to be able to provide attenuation information for reconstruction of a PET image of a portion of a patient that extends beyond the CT FOV. It is also desirable to be able to provide accurate attenuation information for PET images inside the FOV (note that artifacts caused by truncation produce biased attenuation information).

  In one aspect, the method of the present invention scans an object with a first modality having a first field of view and includes a fully sampled field of view data and a partially sampled field of view data. Obtaining the modality data of The method also includes scanning the object with a second modality having a second field of view that is larger than the first field of view to obtain second modality data; the second modality data and the first modality. Reconstructing the image of the object using the partially sampled field-of-view data.

  In another aspect, an imaging device is provided. The apparatus includes a computed tomography (CT) system including an x-ray source and a detector responsive to x-rays arranged to receive x-rays emitted from the source, and gamma rays. A positron emission tomography (PET) system including a detector responsive to the CT system and a computer coupled in operation to the CT system and the PET system. The computer receives data including fully sampled field data and partially sampled field data from the subject's CT scan and uses the fully sampled field data to receive the received partially Configured to augment the sampled field-of-view data, receive data from the subject's PET scan, and reconstruct the image of the subject using the received PET data and the augmented partially sampled field-of-view data ing.

  In another aspect, a computer readable medium encoded with a program is provided. The program augments the partially sampled field of view data from the first modality with the fully sampled field of view data from the first modality to generate the augmented first modality data. Used to instruct the computer to reconstruct the image with the second modality.

  The present invention also includes a method for use with first and second image data sets corresponding to first and second fields of view (FOV), respectively. The first data set includes a plurality of projection views each including first to last attenuation measurements, each corresponding to a first to last parallel trajectory through the first FOV, wherein the first FOV is Smaller than the second FOV and contained within the second FOV, only the common area of the first and second FOV is traversed by each of the projection views and is also within the second FOV And the area outside the first FOV is traversed only by a subset of the projection views. The method uses attenuation measurements from one or more projection views to enhance attenuation measurements from one or more other projection views and a second FOV for the one or more other projection views. Combining the attenuation measurements corresponding to trajectories traversing at least a portion of the second, compensating the second data set for attenuation using the augmented projection view, and combining the compensated second data set. And constructing an image.

  The invention further includes a method for use with a structural data set and a functional data set that indicate the structural and functional characteristics of the object being imaged. The structure set and function set correspond to the first and second fields of view (FOV), respectively, and the structure data set corresponds to the first to last parallel trajectories corresponding to the first to last parallel trajectories through the first FOV, respectively. A plurality of projection views each containing attenuation measurements, wherein the first FOV is smaller than the second FOV and contained within the second FOV, and is common to the first and second FOVs Only areas are traversed by each of the projection views, and areas that are inside the second FOV and outside the first FOV are traversed only by a subset of the projection views. The method includes, for each projected view, adding all attenuation measurements to form a view attenuation measurement, identifying a maximum view attenuation measurement, and a view attenuation measurement that is less than the maximum attenuation measurement. For each of one or more subsets of values, the associated projected view is augmented to form an augmented attenuated view, and the sum of all the attenuated measurements in the augmented view is substantially similar to the maximum attenuated measurement Compensating the second data set for attenuation using the augmented projection view and the non-enhanced projection view; combining the compensated second data set to construct an image; Is included.

  In addition, the present invention provides first and second detections configured to collect first and second data sets, respectively, from a plurality of projection angles around a first and second field of view (FOV). A method used with a container. Each projection angle data constitutes one projection view, the second FOV is larger than and includes the first FOV, and each first set of projection views is the second FOV. Only part of the FOV is traversed. The method of forming an image of an object located within a second FOV includes collecting first and second data sets, and when the object extends beyond the first FOV Identifying one or more first set projection views that are likely to contain the entire object as a complete projection view, and a first set projection from which the object extends. Identifying the view as a truncated projection view, forming a first augmented set by augmenting the data of each truncated projection view with complete projection view data, and augmenting Combining the first set and the second set to form a compensated second set, and combining the compensated second set to form an image.

  In addition, the present invention includes an imaging device for use with a structural data set and functional data set that indicate the structural and functional characteristics of the object being imaged. The structure set and function set correspond to the first and second fields of view (FOV), respectively, and the structure data set corresponds to the first to last parallel trajectories corresponding to the first to last parallel trajectories through the first FOV, respectively. A plurality of projection views each containing attenuation measurements, wherein the first FOV is smaller than the second FOV and contained within the second FOV, and is common to the first and second FOVs Only areas are traversed by each of the projection views, and areas that are inside the second FOV and outside the first FOV are traversed only by a subset of the projection views. The device adds, for each projection view, all attenuation measurements to form a view attenuation measurement, identifies the maximum view attenuation measurement, and one or more of the view attenuation measurements that are less than the maximum attenuation measurement. For each subset of, enhance the associated projected view to form an enhanced attenuated view, such that the sum of all attenuated measurements in the augmented view is substantially similar to the maximum attenuated measurement Including a computer configured to compensate the second data set for attenuation using the projected and unenhanced projected views and to combine the compensated second data set to form an image. .

  This document provides a truncation compensation method and apparatus for an extended field of view of a rotational acquisition system. As will be described in detail below, in one aspect, the method of the present invention, in the case of a parallel sampling geometry, can be used to calculate the total attenuation amount over all channels of the parallel sampling geometry as the projection angle. Based at least in part on the property of being independent. The apparatus and method of the present invention will be described with reference to the drawings, wherein like reference numerals designate the same elements throughout. These drawings are presented for purposes of illustration and not limitation, and are included herein to facilitate the description of exemplary embodiments of the apparatus and method of the present invention.

  In some known CT imaging system configurations, the x-ray source projects a fan-shaped beam, which is the XY plane of a Cartesian coordinate system, generally the “imaging plane”. It is collimated so that it lies in a plane called. The X-ray beam passes through an imaging target such as a patient. The beam enters the array of radiation detectors after being attenuated by the object. The intensity of the attenuated radiation beam received by the detector array depends on the amount of attenuation of the X-ray beam by the object. Each detector element in the array generates a separate electrical signal that is a measurement of the beam intensity at the detector location. Attenuation measurements from all detectors are acquired separately to form a transmission profile (cross section).

  In the third generation CT system, the X-ray source and detector array rotate with the gantry around the imaging object in the imaging plane so that the angle at which the X-ray beam intersects the imaging object changes constantly. A group of x-ray attenuation measurements or projection data from a detector array at one gantry angle is referred to as a “view”. A “scan” of an object includes a set of views that are formed at various gantry or view angles during one revolution of the x-ray source and detector.

  In an axial scan, the projection data is processed to construct an image corresponding to a two-dimensional slice obtained through the object. One method for reconstructing an image from a set of projection data is referred to in the art as the filtered backprojection method. This method converts attenuation measurements from the scan into integers called “CT numbers” or “Hounsfield units” and uses these integers to control the brightness of the corresponding pixel on the cathode ray tube display. To do.

  To shorten the total scan time, a “helical” scan (spiral scan) can also be performed. To perform a “helical” scan, the patient is moved while data for a predetermined number of slices is being acquired. Such a system generates a single helix from a single fan beam helical scan. Projection data is obtained from the helix mapped by the fan beam, and an image at each predetermined slice can be reconstructed from the projection data.

  A reconstruction algorithm for helical scanning typically utilizes a helical weighting algorithm that weights the collected data as a function of view angle and detector channel number. Specifically, before performing the filtered backprojection, the data is weighted according to a helical weighting factor that is a function of both the gantry angle and the detector angle. The weighted data is then processed to generate a CT number and an image corresponding to the two-dimensional slice obtained through the object is constructed.

  In order to further improve the performance of CT systems, multi-slice CT systems have been constructed. In such a system, multiple projections are acquired simultaneously with multiple detector rows. As with the helical scan, a weighting function is applied to the projection data before performing the filtered backprojection method.

  At least some CT systems are also configured to perform positron emission tomography (PET) and are referred to as CT / PET systems (and PET / CT systems). Positrons are positively charged electrons (anti-electrons) that are emitted by radionuclides produced using a cyclotron or other device. The most frequently used radionuclides for diagnostic imaging are fluorine-18 (18F), carbon-11 (11C), nitrogen-13 (13N) and oxygen-15 (15O). A radionuclide is used as a radiotracer called a “radiopharmaceutical” by including it in a substance such as glucose or carbon dioxide. One common use of radiopharmaceuticals is in the medical imaging field.

  In order to use a radiopharmaceutical for imaging, the radiopharmaceutical is injected into a patient and accumulated in an organ or blood vessel or the like to be imaged. It is known that certain radiopharmaceuticals are concentrated in some organs, or in the case of blood vessels, certain radiopharmaceuticals are not absorbed by the vessel wall. The concentration process often involves processes such as glucose metabolism, fatty acid metabolism and protein synthesis. In the following, for the purpose of simplifying the description, the organ to be imaged including blood vessels is generally referred to as “organ of interest”, and the present invention will be described for a temporary organ of interest.

  After the radiopharmaceutical is concentrated in the organ of interest, the radionuclide decays and emits positrons. The positron collides with the electron after moving a very short distance, and when it collides with the electron, it disappears and is converted into two photons, that is, γ rays. This collision event is characterized by the following two features, which are related to medical imaging, specifically medical imaging using photon emission tomography (PET). That is, first, each γ-ray has an energy of approximately 511 keV when extinguished. Second, these two gamma rays are substantially in opposite directions.

  In PET imaging, if the overall position of these disappearances can be identified in three dimensions, a three-dimensional image of the organ of interest can be reconstructed and observed. A PET camera is used to detect the disappearance position. An example of a PET camera includes a plurality of detectors and in particular a processor including coincidence detection circuitry.

  The coincidence circuitry identifies substantially coincident pulse pairs corresponding to detectors located substantially opposite the imaging area. Thus, a coincident pulse pair indicates that annihilation has occurred on a straight line between a pair of related detectors. Millions of annihilation are recorded over several minutes of acquisition time, with each annihilation associated with a unique detector pair. After the acquisition time, the recorded annihilation data is used in any of several different known backprojection procedures to construct a three-dimensional image of the organ of interest.

  As used in this document, the term element or step described in the singular and followed by a singular indefinite article should be understood as not excluding the presence of a plurality of such elements or steps unless explicitly stated otherwise. . Furthermore, references to “one embodiment” of the present invention should not be construed as excluding the existence of other embodiments that also incorporate the recited features.

  Also, the phrase “reconstruct an image” as used herein does not exclude embodiments of the invention in which data representing the image is generated but no visible image is formed. Accordingly, the term “image” as used herein broadly refers to both the visible image and the data representing the visible image. However, many embodiments form (or are configured to form) one or more visible images.

  1 and 2 illustrate a multi-modal imaging system 10, which includes a first modality unit and a second modality unit (shown in FIGS. 1 and 2). Not included). These two modality units cause the system 10 to scan the object with the first modality using the first modality unit and the same object with the second modality using the second modality unit. Making it possible. The system 10 allows for multiple scans with different modalities and facilitates improved diagnostic capabilities over a single modality system. In one embodiment, the multi-modal imaging system 10 is a computed tomography / positron emission tomography (CT / PET) imaging system 10, wherein the CT / PET system 10 is a “third generation” CT. It is shown as including a gantry 12 typical of an imaging system in combination with a PET circuit. In alternative embodiments, modalities other than CT and PET are used with the system 10. The gantry 12 includes a first modality unit having an X-ray source 14 that projects an X-ray beam 16 toward a detector array 18 provided on the opposite side of the gantry 12. To do. The detector array 18 is formed by a plurality of detector rows (not shown) including a plurality of detector elements 20, which are collectively projected X through a subject such as a patient 22. Sense a line. Each detector element 20 generates an electrical signal that represents the intensity of the incident x-ray beam and thus allows estimation of the attenuation of the beam as it passes through the subject or patient 22. During a single scan to acquire X-ray projection data, the gantry 12 and components mounted on the gantry 12 rotate about the center of rotation 24. FIG. 2 shows only a single row of detector elements 20 (ie, a detector row). However, the multi-slice detector array 18 includes a plurality of parallel detector rows of detector elements 20 so that projection data corresponding to a plurality of slices can be acquired simultaneously during a single scan.

  The rotation of the gantry 12 and the operation of the X-ray source 14 are controlled by the control mechanism 26 of the CT / PET system 10. The control mechanism 26 includes an X-ray controller 28 and a gantry motor controller 30. The X-ray controller 28 supplies power signals and timing signals to the X-ray source 14, and the gantry motor controller 30 is a gantry motor controller 30. Twelve rotational speeds and positions are controlled. A data acquisition system (DAS) 32 provided within the control mechanism 26 samples the analog data from the detector element 20 and converts this data into a digital signal for subsequent processing. An image reconstructor 34 receives sampled and digitized x-ray data from DAS 32 and performs high speed image reconstruction. The reconstructed image is applied as an input to the computer 36, which causes the mass storage device 38 to store the image.

  The computer 36 also receives commands and scanning parameters from an operator via a console 40 having a keyboard. The attached cathode ray tube display 42 allows the operator to observe the reconstructed image and other data from the computer 36. The commands and parameters supplied by the operator are used by the computer 36 to supply control signals and information to the DAS 32, X-ray controller 28 and gantry motor controller 30. In addition, the computer 36 operates a table motor controller 44 that controls the motorized table 46 to place the patient 22 in the gantry 12. Specifically, the table 46 moves each part of the patient 22 through the gantry opening 48.

  In one embodiment, the computer 36 may receive instructions and / or instructions from a computer readable medium 52 such as, for example, a flexible disk, CD-ROM, DVD, or other digital source such as a network or the Internet and digital means under development. A device 50 for reading data is included. The device 50 is a network such as a flexible disk drive, a CD-ROM drive, a DVD drive, a magneto-optical disk (MOD) device, or an Ethernet device (“Ethernet” is a trademark). Any other digital device including a connecting device. In other embodiments, computer 36 executes instructions stored in firmware (not shown). The computer 36 is programmed to perform the functions described herein, and the term computer used herein is not limited to only integrated circuits referred to in the art as computers. , Processors, microcontrollers, microcomputers, programmable logic controllers, application specific integrated circuits, and other programmable circuits, which are used interchangeably herein. The CT / PET system 10 also includes a plurality of PET detectors (not shown) that include a plurality of detectors. Since both the PET detector and detector array 18 detect radiation, both are referred to herein as radiation detectors. In one embodiment, CT / PET system 10 is a Discovery LS CT / PET system (TM) commercially available from General Electric Medical Systems, Waukesha, Wisconsin, USA and configured as described herein. In another embodiment, CT / PET system 10 is a Hawkey CT / PET system (TM), also commercially available from General Electric Medical Systems, Waukesha, Wisconsin, USA and configured as described herein.

  In addition, although the medical environment is described herein, the advantages of the present invention are not limited, but are not limited to those typically used at transport sites such as airports or railway stations, but are not limited to, for example, a CT system for baggage scanning. It is thought that it will be useful in all CT systems including such industrial CT systems.

  Under some scanning conditions, a portion of the patient 22 may extend beyond the area measured by the detector 18 resulting in image artifacts and an incomplete representation of the object being imaged. Several known methods have been published that deal with artifact reduction, but do not address imaging of the part of the patient located outside the field of view (FOV). However, it is desirable to image a portion of the patient that extends beyond the FOV. This is useful in many areas, including oncology, spin angiography, synthetic imaging systems, and In Economy CT scanners. Current hardware of known multi-slice CT scanners limits the reconstruction field of view (FOV) to about 50 centimeters (cm). While this is sufficient for most clinical applications, it would be desirable if the FOV could be extended to image objects located outside this FOV. This may be particularly advantageous in applications such as oncology or CT / PET. For oncology applications, a larger FOV is desirable. The main reason is that in the case of X-ray treatment planning, there is the fact that the patient's limbs are often placed outside the scanning FOV in order to better place the tumor. Known CT reconstruction algorithms ignore truncated projections and produce images with strong artifacts. These artifacts can affect the accurate estimation of the attenuation path for treatment planning. FIG. 3 shows an example of a phantom, which shows truncation artifacts. In composite imaging systems such as CT / PET (Computerized Tomography / Positron Emission Tomography), the FOV of the PET system may not match the existing CT design. It is desirable to match the FOV between CT systems and other imaging systems, namely CT / PET, CT / NUC (CT / nuclear medicine) or CT / MR (CT / magnetic resonance). This correction can be used to adjust the FOV to match. In the case of PET, attenuation correction is improved by this method. Described herein is an algorithmic approach that extends the reconstructed FOV beyond the FOV limited by the detector hardware. This correction algorithm can be applied to a variety of reconstruction algorithms including, but not limited to, full scan, half scan / segment and cardiac sector algorithms. In addition, the system 10 is configured to employ the described algorithm.

FIG. 4 shows a plot of the total attenuation (accumulated over all channels) with a parallel sampling geometry as a function of projection angle for a chest phantom scan. Parallel sampling is obtained by rebinning the original fan beam data by techniques known in the art. Note that the curve is close to the horizon. However, this characteristic does not exist in the fan beam sampling geometry. Also, this characteristic is no longer effective when the scan target is located outside the scanning field of view (FOV). The amount of loss is equal to the portion of the object that is located outside the projection FOV. In almost all clinical cases, projection truncation occurs only at a portion of the projection angle, as shown in FIG. In this example, there is no truncation in the projection acquired at the 3 o'clock position, and the projection acquired at the 12 o'clock position is strongly truncated. Thus, depending on the untruncated projection (ie, for example, the position near 3 o'clock in FIG. 5), the amount of truncation for the truncated view (eg, the position near 12 o'clock in the example of FIG. 5) is estimated. be able to. One initial step in this correction method is to perform a software fan beam to parallel beam rebinning on the preprocessed projection. In one embodiment, this initial step is the first step. This method is well known in the art and does not require special data collection. Once rebinning is complete, the projections are integrated over all detector channels to obtain a total attenuation curve, as shown in FIG. Note that the sinking of the total attenuation curve corresponds to the view where truncation occurs. The flat part of the curve corresponds to a view in which no truncation has occurred. Once the total amount of objects located outside the FOV is estimated, the next step is to estimate the distribution of missing projections. To achieve this goal, in one embodiment, first, the boundary readings p _l as shown in Equation 1 below in the truncated projection, as shown in FIG. And _pr are calculated. In order to reduce noise, in one embodiment, an average of m samples is used. It has been empirically found that m = 3 is useful in reducing noise. In other embodiments, m is greater than 1 and less than 5.

  (Formula 1)

Where N is the number of detector channels and k is the projection view number. In addition, in one embodiment, to estimate the nearby slope s _l and s _r at both ends. Gradient estimation is performed by fitting n samples near both ends with a first order polynomial. Experience has shown that n = 5 is useful. In one embodiment, n is greater than 2 and less than 8. In other embodiments, n is greater than 3 and less than 7.

In order to further increase the reliability of the estimation, projections obtained from detector rows adjacent to each other are used. Since the anatomy of the human body typically does not change rapidly at a small distance (a few millimeters), the boundary samples and gradients estimated from adjacent rows typically do not change significantly. Thus, the estimated parameters (p _l , p _r , _sl and s _r ) can be a weighted average of the values calculated from several detector rows. Based on the boundary and gradient information, estimate the location and dimensions of the cylindrical water object that can best fit the truncated projection. The attenuation coefficient of water and mu _w, the radius of the cylinder and R, to represent the distance from the center of the cylinder and X, can be expressed projection value p (x) and the gradient p '(x) is the following formula.

  (Formula 2)

  Since both p (x) and p ′ (x) at the boundaries of the truncated projection are calculated, the target estimates R and x to obtain the size and position of the cylinder to be added to the missing projection. That is. The equations for estimating these parameters can be expressed as:

  (Formula 3)

  (Formula 4)

  Each variable represents an estimate of the position and size of the cylindrical object that needs to be extended from the truncated object. Once these parameters are determined, an extended projection can be calculated using equation (2). This process is illustrated in FIG. 8 for a truncated projection water-filled cylinder.

  In this example, a cylindrical water phantom was used for simplicity. In practice, other object shapes, such as elliptical cylinders, may be used to increase flexibility. Of course, if a priori information about the characteristics of the scan target is available, this information may be used to select the shape of the additional target. Missing projection data can also be estimated using an iterative method.

Estimated cylinders located at both ends of the projection do not necessarily restore the total attenuation of the entire projection. This is because these objects are determined solely from gradient and boundary samples. Information derived from the total attenuation curve (FIG. 6) is not used. To ensure proper compensation for a total attenuation loss, based on the size of the p _l and p _r, to determine the attenuation distribution T _l vs. right attenuation distribution T _r of the left.

  (Formula 5)

In the equation, T is the total loss amount of attenuation determined from FIG. In addition, if the amount of attenuation below the expanded curve is insufficient to compensate for the attenuation loss, the estimated projection is stretched to fill the attenuation defect, as shown in FIG. At this time, the extended portion of the projection is scaled by the predicted total attenuation amount. In one embodiment, the calculation method is as follows. First, the ratio of the predicted total attenuation amount (shown in equation (5)) to the area under the extended projection curve (shown by the shaded area in FIG. 9) is calculated. If the ratio is greater than 1 unit, the x-axis is scaled by this ratio to further expand the initial estimated projection (indicated by the dotted line in FIG. 9) (indicated by the thick solid line in FIG. 9). Similarly, if the ratio is significantly less than one unit, the expanded projection may be compressed with respect to x.

  FIG. 10 shows an example of a reconstructed phantom image when no correction is performed and when correction is performed. The shoulder phantom was scanned in the axial scan mode with a 4 × 1.25 mm detector configuration. A 15 cm plastic phantom was attached to the shoulder phantom so that the edge of the plastic phantom was located near the 65 cm FOV boundary. The truncated object is approximately fully restored. 10A is reconstructed at 50 cm FOV without truncation correction (current product limitations), and FIG. 10B is reconstructed at 65 cm FOV using the methods and apparatus described herein. Please note that. For reference, a partially truncated phantom is shown in FIG.

Although the systems and methods described above estimate missing projection distributions using only total attenuation, boundary sample size, and gradient preservation, additional information may be used for estimation. For example, the above technique can be further refined using the Helgason-Ludwig condition (HL condition) for tomography. In addition, different thresholds can be set to ensure that the algorithm operates properly even under illegitimate measurement conditions. For example, an upper limit and a lower limit may be set for the expansion ratio described with reference to FIG. 9 to prevent an error increase due to a measurement value with low reliability. In addition, the slope calculations for s _l and s _r can be set so that the calculations are within reasonable limits. In addition, when it is known that the physical property of the scanning target is significantly different from that of water, the attenuation coefficient of a known substance (substituting for water) is used to obtain the equations (3) and (4). Dimension and position calculations can be performed. In addition, the information obtained from the other modality can be used to further refine the estimation of missing objects. For example, when some amount of radioactive material is ingested, the reconstructed PET image (without attenuation correction) can be used to estimate the boundary of the object . This information can be supplied to CT image reconstruction to further refine truncation correction.

  Since the interpolated data does not have the same image quality as the data inside the fully sampled FOV, it is useful to mark the image extrapolated from the FOV. FIG. 10 (D) shows the boundary marked with a dotted line. This marking can also be performed by shifting the color code or the CT number. Since the landmarks can affect the ability to observe the image data, a simple way to turn the landmarks on and off is provided. Then, the user of the system 10 can turn on or off the landmark.

  FIG. 11 is a top view of system 10 showing a first modality scan plane 60 and a second modality scan plane 62. In an example embodiment, the first modality is CT and the second modality is PET.

  FIG. 12 shows the transaxial imaging field of the first and second modalities. This cross-axis imaging configuration shows a PET detector 70 positioned around the patient aperture 72 and captures photons emitted from the patient 22 or other test object positioned within the patient aperture 72. Make an image. The source 14 includes an X-ray tube (not shown) with a focal spot 74 located at the focal point of the array 18 of X-ray detectors, which X-ray detector array 18 transmits X through the patient 22. Measure the line strength. The x-ray tube and detector 18 are firmly held together in a frame that rotates around the patient aperture 72. In the course of rotation, measurements are continuously formed within the “fully sampled field of view” 76. X-ray attenuation across any area of the subject 22 positioned between the fully sampled field of view 76 and the patient aperture 72 is measured over a limited range of rotation angles, and this region is referred to as “ Called the “partially sampled field of view” region. In other words, the part of the field of view 76 that is fully sampled is placed inside the fan 78 so that measurements can be taken at all gantry angles, and the collected data is fully sampled. Defined as visual field data. However, there are portions that are inside the fan 78 depending on the angle but outside the fan 78 at other angles, and the data collected for these portions is defined as partially sampled field of view data.

  FIG. 13 shows a typical reconstructed image of the CT detector, which is completely reconstructed even when the patient 22 (not shown in FIG. 13) extends outside the FOV. Limited to the field of view 76 being sampled. Typically, the CT reconstruction method reconstructs only the fully sampled field of view 76, resulting in an image similar to FIG. 13 where the object extends to the partially sampled field of view. Or there is no target part at all. Another artifact observed in several truncated CT reconstructions is the apparent increase in attenuation at the intersection where there is a large amount of truncated attenuation as seen in FIG. 16 (F). is there. When measuring patient attenuation using truncated images, attenuation is underestimated due to missing subjects and overestimated at some lines of response due to overshoot at the margin of CT FOV There is a case.

  FIG. 14 shows a CT-FI (functional image) reconstruction that uses extensible field of view data from CT to correct attenuation of other functional images such as PET images or single photon emission computed tomography (SPECT). A configuration flow process 80 is shown. The CT-FI scan setting step 82 defines a scan space of a functional image (FI) and a plurality of reconstruction parameters, and a corresponding CT image for synthesis when a diagnostic CT is later optionally selected as desired. To do. After an optional CT scout 84, a low dose CT scan 86 is performed and reconstructed twice. The first time forms a composite CT image 87 using the set reconstruction parameters, and the second time forms an attenuation correction (AC) CT image 88 (CTAC) using the above-mentioned extended FOV. In CTAC step 90, the CT scan is converted to an attenuated raw data file (RDF) using FI scan settings 92 (reconstruction definition) and matched to the file for each FI slice position and the FI detector readings. Create CT-FI alignment calibration data to align attenuation measurements. In FI acquisition step 94, release data in one or more FI FOVs are acquired using a resource description framework (RDF) model. In a FI reconstruction step 96, the emission data is corrected for attenuation using a CT-FI raw data file (RDF) to form a corrected functional image 98 and a selected optional image 100 showing attenuation. The CT-FI synthesis step 102 receives both panned, zoomed and filtered CT and PET images that are essentially aligned and specified in the CT-FI scan settings 82.

  In CTAC step 90, the CT image is converted into an attenuation correction file for emission attenuation correction. In the following, a method for deriving the conversion from the CT number to attenuation with the required emission energy will be described. The CT image is calibrated in Hansfield units representing the attenuation of the X-ray beam relative to the attenuation of air and water. When μ represents a linear attenuation coefficient, CT [material], which is the CT number of a specific material, is calculated as follows.

CT [substance] = 1000 * {μ [substance] −μ [water]}
/ {Μ [water] −μ [air]}
The CT machine is calibrated at each kV setting to give a CT number of 0 for water and a CT number of -1000 for air. Some materials, such as bone (and to a lesser extent fat) have different energy dependencies for attenuation, and the CT number of these materials varies with energy. Two different scaling algorithms are used to convert the tissue range to the emission attenuation factor.

  First, if the CT value is less than 0, the substance is assumed to have an energy dependency similar to water (for example, water and tissue), and the attenuation value at the required emission energy keV is as follows: Desired.

μ [substance, keV] −μ [air, keV]
= {Μ [water, keV] −μ [air, keV]} * {CT substance + 1000}
/ 1000
Neglecting the attenuation of air, this conversion requires only knowledge of the attenuation of water at the relevant emission energy. Since the scanner is calibrated to give the same soft tissue CT number regardless of the scanning technique, the scanner's effective energy is not necessary. The emitted energy keV is derived from knowledge about the radioisotope and the type of detection. In the case of a PET detector, the emission energy is 511 keV, and in the case of a SPECT detector, the emission energy depends on the isotope and detector energy tolerance settings. For this reason, the PET detector can use a fixed value for attenuation of water at 511 keV. In the SPECT detector, a table of attenuation values of keV in a certain range can be used.

In the case of bone scaling, a CT value greater than 0 is treated as a mixture of bone and water, and the attenuation value is from the measured value at the X-ray effective energy kV _eff to the attenuation value at the required emission energy keV It is converted as follows.

μ [substance, keV] = μ [water, keV]
+ [CT [kVp] * μ [water, kV _eff ]
* {Μ [bone, keV] −μ [water, keV]}]
/ 1000 * {μ [bone, kV _eff ] −μ [water, kV _eff ]}
In the formula, CT [kVp] is the CT number of the substance measured at high voltage setting kVp (kilovolt potential). This equation requires bone and water attenuation values for both the effective and emitted energy of the CT scanner. These values can be given in a table of the form: A table of bone and water attenuation at each kVp setting (derived from the measurement of effective energy) and a table of bone and water attenuation at each emission energy (511 eV in PET).

  The conversion of CT numbers to attenuation values can be done by applying the above formula and / or by using a look-up table that contains entries for attenuation corresponding to each CT number. FIG. 15 shows a graph of a conversion table for converting measured values at different CT kVp settings into attenuation coefficients at 511 keV.

  After converting the CT value to an attenuation value corresponding to a photon energy of 511 keV, the PET reconstruction proceeds as follows. Smooth the attenuation map to match the resolution of the functional image. The attenuation line integral is calculated by the smoothed attenuation map and sorted as a sinogram to match the functional emission sinogram. The function release data is corrected for attenuation by multiplying the attenuation correction factor. The corrected functional data is reconstructed using tomographic reconstruction such as filtered backprojection (FBP) or sequential subset expectation maximization (OSEM).

  FIG. 16 shows the PET / CT system 10 when the phantom arranged inside the 50 cm CT FOV is located on the left side and the phantom arranged outside the 50 cm CT FOV is located on the right side. An example image from (shown in FIGS. 1 and 2) is shown. A and B represent PET emission reconstructions without attenuation correction. C and D in the middle row represent the PET emission reconstruction with attenuation correction from CT, E in the lower row is a CT image of the centrally located phantom, and F is a CT image from the displaced phantom. .

  Two 20 cm diameter radioactive phantoms were imaged with both PET and CT. The attenuation map derived from the standard 50 cm FOV image has zero attenuation outside the 50 cm diameter as shown in FIG. 16 (F), and these attenuation maps are used to correct PET emission for attenuation, and so on. The release reconstruction was formed. FIG. 17 shows that the radioactivity reconstructed in the truncated attenuation region is smaller than in the fully maintained region (ie, partially sampled data). FIG. 18 shows a CT image reconstructed using the above-described detector extrapolation to form an expanded field reconstruction. The second set of PET reconstructions uses an attenuation map derived from CT data extending beyond 65 cm FOV. FIG. 19 shows a PET emission scan reconstructed with attenuation correction derived from the expanded CT image.

In at least some examples in this document, the phrase “projection view” is used to refer to a set of image data or attenuation measurements that correspond to parallel trajectories through the FOV, where each view is Contains attenuation measurements from beginning to end corresponding to parallel trajectories from beginning to end. In addition, the phrase “enhanced projection view” typically adds additional attenuation measurements corresponding to trajectories adjacent to either the first or last trajectory of the original view. (See again FIGS. 8 and 9 in which the curves corresponding to the projection views are augmented or expanded). Similarly, the phrase “unenhanced projection view” is used to refer to a projection view to which no additional measurements have been added. The phrase “view attenuation measurement” (see FIG. 6 showing view attenuation measurement as a function of CT projection angle) is used to refer to the total attenuation measurement from a single projection view. Yes. The phrase “attenuated projection view” is used to refer to a forward projection set of views derived from an attenuation map, where a 2D image is divided into multiple views. The phrase “attenuation curve” is used to refer to a curve as shown in FIG. 7, in which the attenuation measurement corresponding to a single projection view is represented by the first corresponding attenuation measurement for the curve. so as to extend between the attenuation measurements of the values and the last corresponding, also the first and second gradient s _l and s _r is present near each of the first attenuation measurements and last attenuation measurements So that it is plotted.

  While the invention has been described in terms of various specific embodiments, those skilled in the art will recognize that the invention can be practiced with modification within the spirit and scope of the claims.

1 is a sketch of an embodiment of a CT imaging system. It is a block schematic diagram of the system shown in FIG. FIG. 6 shows a truncated artifact. It is a graph which shows the total attenuation amount integrated | accumulated over all the channels about the chest phantom as a function of a projection angle. FIG. 2 is a diagram of truncation in a clinical environment. It is a graph which shows the influence which truncation projection has on the total amount of attenuation. FIG. 6 is a diagram of gradient and boundary estimation. FIG. 6 is a view of a water cylinder fitted to a truncated projection. It is the figure of the projection expansion scaled by the prediction total attenuation amount. It is a figure which shows a some image. FIG. 3 is a top view of the system shown in FIGS. 1 and 2 showing a first modality scan plane and a second modality scan plane. It is a figure which shows the imaging field of the transaxial direction of a 1st and 2nd modality. FIG. 4 shows a typical reconstructed image of a CT detector restricted to a fully sampled field of view. It is a flowchart of CT-FI (functional image) reconstruction. It is a graph figure of the some conversion table which converts the measured value in a different CT kVp setting into the attenuation coefficient in 511 keV. FIG. 3 is an example of an image from the PET / CT system shown in FIG. 1 and FIG. 2, the phantom arranged inside the 50 cm CT FOV is located on the left side, and the phantom arranged outside the 50 cm CT FOV is located on the right side. FIG. FIG. 5 shows that the reconstructed radioactivity in the truncated attenuation region is less than in the fully maintained region. It is the figure of CT image reconstructed using the extrapolation of the detector described in this book. FIG. 5 is a diagram of a PET emission scan reconstructed with attenuation correction derived from an expanded CT image.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 CT system 12 Gantry 14 X-ray source 16 X-ray cone beam 18 Detector array 20 Detector element 22 Patient 24 Center of rotation 26 Control mechanism 42 Display 46 Motorized table 48 Gantry opening 50 Media reader 52 Media 60 1st Modality scan plane 62 Second modality scan plane 70 PET detector 72 Patient aperture 74 Focus spot 76 Fully sampled field of view 78 Fan 80 CT-FI reconstruction process

Claims (6)

An X-ray source (14) and a detector (18) responsive to the X-rays arranged to receive fan-shaped X-rays (16) emitted from the source toward the subject . A computed tomography (CT) system;
a positron emission tomography (PET) system including a detector (70) responsive to gamma rays;
A computer (36) coupled in operation to the CT system and the PET system,
Receiving data from the subject's PET scan and reconstructing the subject's PET image;
Includes fully sampled view (76) data and partially sampled view data from a CT scan of the object (22) over all gantry angles using the fan-shaped X-ray (16) Receive data,
Enhance the view data that is pre-Symbol received partially sampled,
A computer (36) configured to reconstruct the CT image of the object by using the view data that is pre-Symbol enhanced partially sampled, with a
The augmentation of the view data
Estimating the distribution of missing projections outside the detector (18);
Done by adding additional attenuation measurements corresponding to trajectories adjacent to either the first or last trajectory of the original view to satisfy the estimated missing projection;
Estimating a boundary between the area of the fully sampled view (76) and the area of the partially sampled view with the aid of the reconstructed PET image, An imaging device (10) for performing addition of the additional attenuation measurement value based on the information . The fully sampled view area (76) is the area located inside the fan shape at all gantry angles;
The partially section view being sampled is a region which is disposed within the fan-shaped part of the gantry angle only,
The computer (36) further includes a boundary between the area of the fully sampled view (76) and the area of the partially sampled view and the area of the partially sampled view. The apparatus (10) of claim 1, wherein the apparatus (10) is configured to form a total attenuation curve by integrating projections across all detector channels to estimate a total amount of missing attenuation measurements at. . The computer (36) further includes
Receiving a signal representing the x-ray tube voltage,
The apparatus (10) according to claim 1 or 2 , configured to convert one or more CT numbers of the CT image into PET attenuation numbers based on the X-ray tube voltage. The computer (36) further includes
The apparatus (10) of claim 3, configured to match the resolution of the PET image by smoothing a map of attenuation numbers after converting the CT numbers to PET attenuation numbers. The computer (36) further includes
Read the boundary readings pl and pr
Is configured to calculate according to
The apparatus (10) of claim 2, wherein m is a predetermined variable, N is the number of detector channels, and k is a projection view number. The computer (36) further includes
An attenuation distribution Tl outside the boundary reading value pl and an attenuation distribution Tr outside the boundary reading value pr.
Is configured to calculate according to
The apparatus (10) according to claim 5, wherein T is the total amount of the attenuation measurements.