Pet device, method and program

Abstract

To enable TOF calibrations to be carried out simply.SOLUTION: A PET device according to an embodiment comprises an acquisition unit, a reconstruction unit, an identification unit, and an estimation unit. The acquisition unit acquires list mode data. The reconstruction unit reconstructs a PET image on the basis of the list mode data. The identification unit identifies first coincidence data among a plurality of coincident data included in the list mode data that is simultaneously counted by a first detector and a second detector that is different from the first detector. The estimation unit estimates a deviation in time between the first and second detectors on the basis of the TOF information included in first coincident data and the reconstructed image.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The embodiments disclosed in the present specification and drawings relate to a PET device, a method, and a program. [Background technology]

[0002] In recent years, the time-of-flight (TOF) time resolution of positron emission computed tomography (PET) devices has improved, making it possible to perform reconstruction using TOF information with higher accuracy.

[0003] Here, in order to accurately perform reconstruction using TOF information, it is necessary to calibrate the time of the detector (hereinafter, TOF calibration), and for example, a method using an external radiation source is generally used as a TOF calibration method. As one example, a serviceman uses an external radiation source to identify the deviation in the detection time of the detector, and adjusts the time information of the corresponding detector based on the identified deviation. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2021-110737 A Summary of the Invention [Problem to be solved by the invention]

[0005] One of the problems to be solved by the embodiments disclosed in this specification and the drawings is to make it possible to easily perform TOF calibration. However, the problems to be solved by the embodiments disclosed in this specification and the drawings are not limited to the above problem. Problems corresponding to the effects of each configuration shown in the embodiments described later can also be positioned as other problems. [Means for solving the problem]

[0006] A PET device according to an embodiment includes an acquisition unit, a reconstruction unit, an identification unit, and an estimation unit. The acquisition unit acquires list mode data. The reconstruction unit reconstructs a PET image based on the list mode data. The identification unit identifies a first coincidence data, which is counted by a first detector and a second detector different from the first detector, from among a plurality of coincidence data included in the list mode data. The estimation unit estimates a time shift between the first detector and the second detector based on TOF information included in the first coincidence data and the reconstructed image. [Brief description of the drawings]

[0007] [Figure 1] FIG. 1 is a block diagram showing an example of the configuration of a PET apparatus according to the first embodiment. [Figure 2A] FIG. 2A is a diagram illustrating an example of count information according to the first embodiment. [Figure 2B] FIG. 2B is a diagram showing an example of a chronological list of coincidence counting information according to the first embodiment. [Diagram 3] FIG. 3 is a diagram for explaining a TOF depiction image according to the first embodiment. [Figure 4] FIG. 4 is a diagram for explaining the process of estimating the time offset according to the first embodiment. [Diagram 5] FIG. 5 is a flowchart showing a processing procedure performed by the PET apparatus according to the first embodiment. [Figure 6] FIG. 6 is a diagram for explaining an example of a pair of detector modules according to the first modification. [Figure 7] FIG. 7 is a diagram for explaining an example of a pair of detector modules according to the first modification. [Figure 8A] FIG. 8A is a diagram for explaining an example of constructing a trained model according to Modification Example 2. [Figure 8B] FIG. 8B is a diagram illustrating an example of a trained model according to the second modification. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0008] Hereinafter, embodiments of a PET device, a method, and a program will be described in detail with reference to the drawings. Note that the PET device, the method, and the program according to the present application are not limited to the embodiments described below.

[0009] (First embodiment) 1 is a block diagram showing an example of a configuration of a PET device 100 according to the first embodiment. As shown in FIG. 1, the PET device 100 according to the first embodiment includes a gantry device 10 and a console device 20.

[0010] The gantry device 10 includes a PET detector 101, a counting information generating circuit 102, a tabletop 103, a bed 104, and a bed driving unit 105, and generates counting information for reconstructing a PET image by detecting gamma rays emitted from biological tissue that has absorbed a positron-emitting nuclide administered to the subject P.

[0011] The PET detector 101 includes a plurality of detector modules. The plurality of detector modules detect gamma rays by detecting scintillation light (fluorescence) that is re-emitted when a substance that has become excited as a result of pair annihilation gamma rays emitted from positrons in the subject P interacting with a light emitter (scintillator) transitions back to the ground state. The plurality of detector modules detects radiation energy information of pair annihilation gamma rays emitted from positrons in the subject P. The plurality of detector modules are arranged to surround the subject P in a ring shape.

[0012] The detector module includes, for example, a scintillator array, a photodetector array, and a light guide.

[0013] The scintillator array includes a plurality of scintillators arranged two-dimensionally. The scintillator converts the annihilation gamma rays emitted from the positrons in the subject P and incident thereon into scintillation light (scintillation photons, optical photons), and outputs the scintillation light. The scintillator is formed of a scintillator crystal suitable for TOF measurement and energy measurement, such as LaBr3 (Lanthanum Bromide), LYSO (Lutetium Yttrium Oxyorthosilicate), LSO (Lutetium Oxyorthosilicate), or LGSO (Lutetium Gadolinium Oxyorthosilicate).

[0014] The scintillator is not limited to the above-mentioned examples. For example, the scintillator may be a lead compound such as GBO (Bismuth Germanium Oxide), lead glass (SiO2+PbO), lead fluoride (PbF2), or PWO (PbWO4).

[0015] The light guide is made of a plastic material or the like having excellent optical transparency, and transmits the scintillation light output from the scintillator to the photodetector. Specifically, the light guide transmits the scintillation light to a SiPM (Silicon Photomultiplier) to be described later.

[0016] The photodetector array includes a plurality of photodetectors (photodetection elements) arranged two-dimensionally. As the photodetectors, for example, SiPMs or the like are used.

[0017] The PET detector 101 includes a counting information generating circuit 102 for each detector module.

[0018] The counting information generating circuit 102 generates counting information (list mode data) by converting the output signal from the PET detector 101 into digital data. This counting information includes the detection position, energy value, and detection time of the pair annihilation gamma ray. For example, the counting information generating circuit 102 identifies a scintillator number (P) indicating the position of the scintillator on which the pair annihilation gamma ray is incident.

[0019] The counting information generating circuit 102 can use various methods to identify the scintillator position on which the pair annihilation gamma ray is incident. For example, when one scintillator corresponds to one SiPM, the counting information generating circuit 102 may identify the position of the scintillator corresponding to the SiPM from which the output is obtained as the scintillator position on which the pair annihilation gamma ray is incident. Also, for example, the counting information generating circuit 102 may identify the scintillator position on which the pair annihilation gamma ray is incident by performing a center of gravity calculation based on the position of each photodetector and the intensity of the output signal.

[0020] Furthermore, the counting information generating circuit 102 performs an integral calculation of the intensity of the electrical signal output from each photodetector to identify the energy value (E) of the pair annihilation gamma ray incident on the PET detector 101. Furthermore, the counting information generating circuit 102 identifies the detection time (T) at which the PET detector 101 detects the scintillation light caused by the pair annihilation gamma ray. Note that the detection time (T) may be an absolute time or may be the elapsed time from the start of imaging.

[0021] As described above, the counting information generating circuit 102 corresponding to each detector module generates counting information including a scintillator number (P), an energy value (E), and a detection time (T). Then, each counting information generating circuit 102 stores the generated counting information in the memory 130 of the console device 20. FIG. 2A is a diagram showing an example of the counting information according to the first embodiment. For example, as shown in FIG. 2A, the memory 130 stores the scintillator number (P), the energy value (E), and the detection time (T) in association with a module ID that identifies the detector module.

[0022] The count information generating circuit 102 is realized by, for example, a processor. The count information generating circuit 102 is an example of an acquiring unit.

[0023] Returning to FIG. 1 , the top board 103 is a bed on which the subject P is placed, and is placed on the bed 104. The bed driving unit 105 moves the top board 103 under the control of the system control function 106c of the processing circuitry 106. For example, the bed driving unit 105 moves the subject P into the imaging port of the gantry device 10 by moving the top board 103.

[0024] The console device 20 accepts operations of the PET device 100 by a user, controls the capture of PET images, and reconstructs (generates) PET images using counting information collected by the gantry device 10. As shown in Fig. 1, the console device 20 includes a processing circuit 106, an input interface 110, a display 120, and a memory 130. The processing circuit 106, the input interface 110, the display 120, and the memory 130 are connected via a bus.

[0025] The processing circuit 106 executes a coincidence information generating function 106a, a reconstruction function 106b, a system control function 106c, a specification function 106d, a generation function 106e, an estimation function 106f, and a calibration function 106g. Each of the coincidence information generating function 106a, the reconstruction function 106b, the system control function 106c, the specification function 106d, the generation function 106e, the estimation function 106f, and the calibration function 106g is stored in the memory 130 in the form of a program executable by a computer. The processing circuit 106 is a processor that reads each program from the memory 130 and executes each read program to realize each function corresponding to each program. In other words, the processing circuit 106 in a state in which each program has been read has each function shown in the processing circuit 106 of FIG. 1.

[0026] 1, the coincidence counting information generating function 106a, the reconstruction function 106b, the system control function 106c, the identification function 106d, the generating function 106e, the estimation function 106f, and the calibration function 106g are realized by a single processing circuit 106. However, the processing circuit 106 may be composed of multiple independent processors, and each processor may execute a program to realize the functions.

[0027] The reconstruction function 106b is an example of a reconstruction unit. The identification function 106d is an example of a identification unit. The generation function 106e is an example of a generation unit. The estimation function 106f is an example of an estimation unit. The calibration function 106g is an example of a calibration unit.

[0028] The coincidence counting information generating function 106a acquires the counting information stored in the memory 130, and generates coincidence counting information (coincidence data) based on the acquired counting information. Then, the coincidence counting information generating function 106a arranges the generated coincidence counting information in approximately chronological order based on the detection time (T) and stores it in the memory 130.

[0029] 2B is a diagram showing an example of a time series list of coincidence counting information according to the first embodiment. As shown in FIG. 2B, memory 130 stores sets of counting information in association with "coincidence No.", which is the serial number of the coincidence counting information. In the first embodiment, the time series list of coincidence counting information is arranged in approximate chronological order based on the detection time (T) of the counting information. In addition, hereinafter, the time difference between the detection times (T) in a set of counting information is referred to as TOF information.

[0030] The reconstruction function 106b reconstructs a PET image (reconstructed image). Specifically, the reconstruction function 106b reconstructs the PET image based on the list mode data. For example, the reconstruction function 106b acquires a time series list of coincidence counting information stored in the memory 130, and reconstructs the PET image using the acquired time series list of coincidence counting information. Then, the reconstruction function 106b stores the reconstructed PET image in the memory 130.

[0031] The system control function 106c controls the entire PET device 100 by controlling the gantry device 10 and the console device 20. For example, the system control function 106c controls imaging in the PET device 100. In addition, the system control function 106c controls the bed drive unit 105 to control the movement of the top board 103.

[0032] Furthermore, the system control function 106c causes various images and various information to be displayed on the display 120. For example, the system control function 106c causes a PET image to be displayed on the display 120. Furthermore, the system control function 106c causes the display 120 to display a GUI (Graphical User Interface) for receiving various instructions and various settings from a user (operator) of the PET device 100.

[0033] The identifying function 106d identifies predetermined coincidence information from the coincidence information stored in the memory 130. The process performed by the identifying function 106d will be described in detail later.

[0034] The generating function 106e generates a TOF depicting image based on the TOF information included in the predetermined coincidence counting information identified by the identifying function 106d. Note that the processing by the generating function 106e will be described in detail later.

[0035] The estimation function 106f estimates the amount of time shift between the detector modules based on the TOF information and the reconstructed image. Note that the processing by the estimation function 106f will be described in detail later.

[0036] The calibration function 106g calibrates the time information in the detector module based on the time offset estimated by the estimation function 106f. Note that the process performed by the calibration function 106g will be described in detail later.

[0037] The input interface 110 accepts various instructions and settings input from a user, and outputs the accepted instructions and settings to the processing circuit 106. For example, the input interface 110 converts the input operation accepted from the user into an electric signal, and transmits the electric signal to the processing circuit 106. For example, the input interface 110 is realized by a trackball, a switch button, a mouse, a keyboard, a touch pad that performs an input operation by touching the operation surface, a touch screen in which a display screen and a touch pad are integrated, a non-contact input circuit using an optical sensor, and a voice input circuit. Note that in this specification, the input interface 110 is not limited to only those that have physical operation parts such as a mouse and a keyboard. For example, an electric signal processing circuit that receives an electric signal corresponding to an input operation from an external input device provided separately from the PET device 100 and transmits the electric signal to the processing circuit 106 is also included as an example of the input interface 110.

[0038] The display 120 is connected to the processing circuitry 106 and displays various information and images. For example, the display 120 converts information and data transmitted from the processing circuitry 106 into an electric signal for display and outputs it. To give a specific example, under the control of the system control function 106c, the display 120 displays a PET image and a GUI for receiving various instructions and settings from a user. For example, the display 120 is realized by a liquid crystal monitor, a CRT (Cathode Ray Tube) monitor, a touch panel, or the like.

[0039] The memory 130 stores various data used in the PET device 100. The memory 130 is realized by, for example, a semiconductor memory element such as a random access memory (RAM) or a flash memory, a hard disk, an optical disk, etc. The memory 130 stores counting information, a time series list of coincidence counting information, a reconstructed PET image, etc.

[0040] An example of the configuration of the PET device 100 according to this embodiment has been described above. For example, the PET device 100 is installed in a medical institution such as a hospital or a clinic, and is used for various image diagnoses using PET images generated by the PET device 100, with a patient who is admitted to or visits the medical institution as a subject P. Here, the PET device 100 according to this embodiment makes it possible to easily perform TOF calibration.

[0041] As described above, in the TOF-PET device, in order to perform accurate reconstruction using TOF information, it is required to periodically calibrate the time of the detector (TOF calibration). However, in general TOF calibration, a serviceman must use an external radiation source to identify the detection time shift in the detector and adjust the time information of the corresponding detector based on the identified shift, which is time-consuming. Therefore, the PET device 100 according to this embodiment estimates the amount of time shift of the detector using the TOF information and the reconstructed image, thereby reducing the time required for TOF calibration and making it possible to perform TOF calibration easily. Furthermore, the PET device 100 according to this embodiment can reduce the radiation exposure of the serviceman by using list mode data collected from the subject P, rather than using list mode data collected from an external radiation source. Details of the processing of the PET device 100 will be described below.

[0042] The identification function 106d identifies a first coincidence data counted by a first detector and a second detector different from the first detector among a plurality of coincidence data included in the list mode data. Specifically, the identification function 106d identifies TOF information used to estimate the time shift amount. In other words, the identification function 106d identifies a detector module to be subjected to TOF calibration.

[0043] For example, the identifying function 106d identifies TOF information used to estimate the time shift amount from the coincidence data shown in Fig. 2B. As an example, the identifying function 106d identifies the TOF information of "Coincidence No. 1" as the TOF information used to estimate the time shift amount. As a result, the detector module D1 corresponding to "scintillator number (P): P11" and the detector module D2 corresponding to "scintillator number (P): P22" become targets for TOF calibration.

[0044] Note that the identifying function 106d may identify TOF information of all coincidence data, or may identify TOF information of predetermined coincidence data (ie, coincidence data corresponding to a predetermined detector module).

[0045] The generating function 106e generates a TOF depicting image in which the pair annihilation point position is estimated based on the TOF information. Specifically, the generating function 106e generates a TOF depicting image in which the pair annihilation point position is depicted, using the TOF information of the coincidence data identified by the identifying function 106d.

[0046] Fig. 3 is a diagram for explaining a TOF depiction image according to the first embodiment. Here, Fig. 3 shows an example of generating a TOF depiction image of the TOF information of "coincidence No. 1", but in reality, the generating function 106e generates a TOF depiction image based on the TOF information of all coincidence data identified by the identifying function 106d. Also, Fig. 3 is shown in two dimensions for convenience of explanation, but in reality, the generating function 106e generates a three-dimensional TOF depiction image.

[0047] In the TOF-PET device, the coordinates of the pair annihilation point are estimated from the difference in detection time (T) between detector modules, and the distribution having a half-width of the distance according to the time resolution of TOF along the LOR (Line-of-Response) is used as the position information of the pair annihilation point, as shown in the diagram of the raw data area on the left side of Fig. 3. The generation function 106e generates a TOF depiction image showing points corresponding to the position information of the pair annihilation point on a three-dimensional coordinate system, as shown in the diagram of the image data area on the right side of Fig. 3.

[0048] Here, the TOF depiction image generated by the generation function 106e depicts coincidence events as points without reconstruction, so the positions of the points respond sensitively to the time information in the detector modules. That is, the positions of the points in the TOF depiction image are information that more closely reflects the detection time (T) in each detector module. Note that the higher the time resolution of the TOF, the more accurate the positions of the points in the TOF depiction image.

[0049] The estimation function 106f estimates the amount of time shift between the first detector and the second detector based on the TOF information and the reconstructed image included in the first coincidence data. Specifically, the estimation function 106f estimates the amount of time shift based on the TOF depicting image and the reconstructed image. More specifically, the estimation function 106f estimates the amount of time shift based on a result of superimposing and comparing the TOF depicting image and the reconstructed image.

[0050] Fig. 4 is a diagram for explaining the estimation process of the time shift amount according to the first embodiment. Here, Fig. 4 shows the positional relationship of the corresponding pair annihilation points when the TOF depiction image and the PET image (normal image in the figure) shown in Fig. 3 are superimposed. For example, the estimation function 106f estimates the time shift amount between the detector module D1 and the detector module D2 from the positional shift between the pair annihilation point in the normal image shown in Fig. 4 and the pair annihilation point in the TOF depiction image.

[0051] As described above, the TOF depiction image depicts coincidence events as points without reconstruction, so the detection time (T) at each detector module is reflected in the position. On the other hand, the PET image is reconstructed based on coincidence data collected at various angles, so the position of the annihilation point is closer to the true value. Therefore, by comparing the TOF depiction image with the PET image to detect the deviation, it is possible to detect what deviation has occurred in the detection time (T) at each detector module.

[0052] Here, the PET image to be compared with the TOF depiction image may be generated by reconstruction using TOF information, or may be generated by non-TOF reconstruction without using TOF information. That is, the reconstruction function 106b reconstructs a PET image without using the TOF information of each of the multiple coincidence data included in the list mode data, and the estimation function 106f estimates the time shift amount using the PET image.

[0053] Since PET images generated by non-TOF reconstruction are reconstructed without using TOF information, they are not affected by the time shift contained in the TOF information and are closer to the true value than PET images generated by reconstruction using TOF information. In other words, by using PET images generated by non-TOF reconstruction, it is possible to estimate the time shift more accurately.

[0054] 4, since the points of the TOF depiction image are shifted toward the detector module side with respect to the normal image, the estimation function 106f determines that there is a time shift between the detector module D1 and the detector module D2. Then, the estimation function 106f estimates the amount of time shift from the amount of shift of the points.

[0055] Here, a priority is set between the detector modules, and it is determined which detector module is used as a reference. For example, in the state shown in FIG. 4, if the detector module D1 is set to have a higher priority than the detector module D2, the estimation function 106f determines that the time of the detector module D2 is shifted earlier. In other words, the estimation function 106f estimates that the time of the detector module D2 is earlier than that of the detector module D1 by an amount corresponding to the image shift.

[0056] As described above, the estimation function 106f estimates the time shift amount of the detector module by comparing the TOF depiction image and the PET image. Here, the estimation function 106f can arbitrarily set the area to be compared in the image comparison.

[0057] For example, the estimation function 106f calculates a first center of gravity position indicating the center of gravity of the image information in the entire TOF depiction image and a second center of gravity position indicating the center of gravity of the image information in the entire reconstructed image (PET image), and estimates the time shift amount based on the first center of gravity position and the second center of gravity position. That is, the estimation function 106f simply estimates the time shift amount from the shift amount of the center of gravity of the entire image. Here, the estimation function 106f identifies the detector module whose time information is shifted based on the direction in which the center of gravity is shifted.

[0058] In this way, by estimating the time shift amount using the center of gravity of the entire image, it is possible to identify the detector module whose time information is shifted from among all the detector modules, and estimate the time shift amount of the identified detector module, thereby performing TOF calibration of the entire PET detector 101.

[0059] Also, for example, the estimation function 106f extracts a first region showing a relatively high count in the TOF depiction image and a second region showing a relatively high count in the reconstructed image (PET image), and estimates the time shift amount based on the first region and the second region. That is, the estimation function 106f compares the region (HOT part) where the annihilation points are concentrated, and estimates the time shift amount from the shift amount of the HOT part. Here, the estimation function 106f identifies the detector module whose time information is shifted based on the direction in which the HOT part is shifted.

[0060] In this way, by estimating the time shift amount using the HOT parts in the image, it is possible to limit the detector modules to be calibrated and perform highly accurate TOF calibration on the limited detector modules.

[0061] The calibration function 106g calibrates the time information of the first detector or the second detector based on the time shift amount. Specifically, the calibration function 106g calibrates the time information of the target detector module based on the time shift amount estimated by the estimation function 106f. For example, in the state shown in FIG. 4, the calibration function 106g adjusts the time information for the counting information generation circuit 102 corresponding to the detector module D2 based on the estimation result of the estimation function 106f that "the time of the detector module D2 is ahead of the detector module D1 by the time corresponding to the image shift."

[0062] In this way, the calibration function 106g performs time adjustment for the counting information generating circuit 102 corresponding to the detector module identified by the estimation function 106f as having a shift in time information by the amount of time offset estimated by the estimation function 106f. Here, an upper limit may be set for the amount of time offset calibrated by the calibration function 106g. That is, the calibration function 106g calibrates the time information of the first detector or the second detector on the condition that the amount of time offset is within a threshold value.

[0063] For example, 1% of the time resolution of the TOF in the PET detector 101 may be set as the upper limit, and if this limit is exceeded, TOF calibration by the calibration function 106g may not be performed. In such a case, the system control function 106c notifies the user that TOF calibration by the PET device 100 has not been performed by displaying information on the display 120 that the time offset has exceeded the upper limit. The user who has confirmed this display may, for example, request a serviceman to perform TOF calibration.

[0064] As described above, PET device 100 performs TOF calibration using list mode data. Here, TOF calibration by PET device 100 is performed at any timing. For example, PET device 100 may perform TOF calibration every time list mode data is collected, or may perform TOF calibration at a set time (e.g., overnight) using list mode data collected up to that point.

[0065] Next, an example of processing by the PET device 100 will be described with reference to Fig. 5. Fig. 5 is a flowchart showing a processing procedure by the PET device 100 according to the first embodiment.

[0066] 5, in this embodiment, the system control function 106c and the coincidence counting information generating function 106a start data collection (step S101) and end the data collection (step S102). This process is realized, for example, by the processing circuit 106 reading out from the memory 130 and executing a program corresponding to the system control function 106c and the coincidence counting information generating function 106a.

[0067] Next, the reconstruction function 106b generates a normal image using the collected coincidence data (step S103). This process is realized, for example, by the processing circuitry 106 reading out a program corresponding to the reconstruction function 106b from the memory 130 and executing it.

[0068] In addition, the identifying function 106d identifies coincidence data, and the generating function 106e generates a TOF depiction image based on the TOF information in the identified coincidence data (step S104). This process is realized, for example, by the processing circuitry 106 reading out from the memory 130 and executing a program corresponding to the identifying function 106d and the generating function 106e.

[0069] Next, the estimation function 106f superimposes the normal image and the TOF depiction image to calculate the time shift amount (step S105). This process is realized, for example, by the processing circuitry 106 reading out from the memory 130 and executing a program corresponding to the estimation function 106f.

[0070] Next, the calibration function 106g calibrates the time information of the corresponding detector module based on the time shift amount (step S106). This process is realized, for example, by the processing circuit 106 reading out a program corresponding to the calibration function 106g from the memory 130 and executing it.

[0071] In the flowchart shown in FIG. 5, an example is shown in which a TOF depicting image is generated after data collection is completed, but a TOF depicting image may be generated during data collection.

[0072] As described above, according to the first embodiment, the counting information generating circuit 102 acquires list mode data. The reconstruction function 106b reconstructs a PET image based on the list mode data. The identification function 106d identifies the first coincidence data, which is counted by the first detector and the second detector different from the first detector, among a plurality of coincidence data included in the list mode data. The estimation function 106f estimates the time shift between the first detector and the second detector based on the TOF information included in the first coincidence data and the reconstructed image. Therefore, the PET device 100 according to the first embodiment can perform TOF calibration using collected data, making it possible to easily perform TOF calibration. In addition, according to the first embodiment, TOF configuration can be performed without using an external radiation source, making it possible to reduce the radiation exposure of service personnel.

[0073] According to the first embodiment, the calibration function 106g calibrates the time information of the first detector or the second detector based on the time shift amount. Therefore, the PET device 100 according to the first embodiment can automatically perform TOF calibration, and allows the TOF calibration to be easily performed.

[0074] According to the first embodiment, the generating function 106e generates a TOF depicting image in which the annihilation point position is estimated based on the TOF information. The estimating function 106f estimates the time shift amount based on the TOF depicting image and the reconstructed image. Therefore, the PET device 100 according to the first embodiment can estimate the time shift amount of the detector module by comparing images, and enables easy TOF calibration.

[0075] According to the first embodiment, the estimation function 106f calculates a first center position indicating the center of gravity of the image information in the entire TOF depiction image and a second center position indicating the position of the center of gravity of the image information in the entire reconstructed image, and estimates the time shift amount based on the first center position and the second center position. Therefore, the PET device 100 according to the first embodiment makes it possible to perform TOF calibration of the entire PET detector 101.

[0076] According to the first embodiment, the estimation function 106f extracts a first region showing a relatively high count in the TOF depiction image and a second region showing a relatively high count in the reconstruction image, and estimates the time shift amount based on the first region and the second region. Therefore, the PET device 100 according to the first embodiment makes it possible to perform highly accurate TOF calibration for a limited detector module.

[0077] According to the first embodiment, the reconstruction function 106b reconstructs a PET image without using the TOF information of each of the coincidence data included in the list mode data. Therefore, the PET device 100 according to the first embodiment makes it possible to estimate a more accurate time shift amount.

[0078] According to the first embodiment, the calibration function 106g calibrates the time information of the first detector or the second detector on condition that the time offset is within a threshold value. Therefore, the PET device 100 according to the first embodiment makes it possible to perform TOF calibration within a desired range.

[0079] (Variation 1) In the above-described embodiment, the case where all pairs of detector modules are targeted has been described. However, the embodiment is not limited to this, and may be, for example, a case where the pairs of detector modules are limited.

[0080] Specifically, the identification function 106d of the first variant example limits at least one of the first detector and the second detector to a specified detector, and identifies the first coincidence data counted simultaneously by the first detector and the second detector.

[0081] 6 and 7 are diagrams for explaining an example of a pair of detector modules according to Modification 1. For example, as shown in Fig. 6, the identifying function 106d limits the angle to be calibrated in the FOV to a 60° direction, and identifies coincidence data detected by a detector module paired in the limited direction as coincidence data to be used for TOF calibration.

[0082] Also, for example, the identification function 106d limits the detector module D3 as shown in FIG. 7, and identifies the coincidence data detected by the detector module paired with the limited detector module D3 as the coincidence data to be used for TOF calibration.

[0083] As described in the first embodiment, if all pairs of detector modules are targeted, the calculations become complicated and may not converge. Therefore, by limiting the pairs of detector modules as described above, the calculations related to the TOF calibration are simplified, and it is possible to ensure that the calculations related to the TOF calibration converge.

[0084] (Variation 2) In the above-described embodiment, a case has been described in which a PET image and a TOF depiction image are superimposed and a time shift amount is estimated from a position shift. However, the embodiment is not limited to this, and for example, a time shift amount may be estimated using a trained model. Specifically, the estimation function 106f according to the second modification estimates the time shift amount using a trained model trained to output a time shift amount according to input of TOF information and a reconstructed image.

[0085] Fig. 8A is a diagram for explaining an example of construction of a trained model according to Modification 2. Here, construction of the trained model shown in Fig. 8A is executed by an information processing device such as a workstation using data collected by a PET device. Then, the constructed trained model is stored in the memory 130 of the PET device 100.

[0086] For example, the trained model according to the second modification is constructed by machine learning using a plurality of sets of data (training data) of a TOF depiction image, a normal image (PET image), and a deviation amount, as shown in Fig. 8A. Here, the training data used to construct the trained model is generated based on ideal raw data, as shown in Fig. 8A.

[0087] Ideal raw data is list mode data with no deviation in TOF information (no deviation in time information of detector modules), and is collected, for example, by a PET device with a highly accurate TOF calibration. Here, a highly accurate TOF calibrated PET device is, for example, a PET device immediately after TOF calibration is performed by a serviceman.

[0088] Then, a known time shift is given to the time information of the detector by adding the shift amount A to the time information of the detector of the above ideal raw data. This allows raw data with time information shifted by A to be acquired. Then, a TOF depicting image and a normal image are generated using the acquired raw data, which become learning data together with the time shift amount A. In constructing a trained model according to the second modification, the above shift amount A is changed in various ways to generate a plurality of learning data.

[0089] Then, machine learning is performed using multiple TOF depiction images and normal images generated using raw data with time information shifted by A, and training data including the known time shift amount A, to construct a trained model.

[0090] For example, as shown in Fig. 8B, the estimation function 106f reads out the trained model stored in the memory 130, and inputs a TOF depiction image and a normal image based on newly collected list mode data to the read trained model to obtain the time shift amount. Note that Fig. 8B is a diagram showing an example of a trained model according to Modification Example 2.

[0091] As described above, the estimation function 106f estimates the time shift amount by using the trained model. Therefore, the PET device 100 according to the second modification makes it possible to easily estimate the time shift amount.

[0092] (Variation 3) In the above embodiment, a case has been described in which a TOF depicting image is generated from TOF information and the time shift amount is estimated using the generated TOF depicting image. However, the embodiment is not limited to this, and the time shift amount may be estimated using the TOF information without generating a TOF depicting image.

[0093] In such a case, the estimation function 106f according to the third modification specifies position information from the TOF information, and estimates the amount of time shift by comparing the specified position information with the PET image. That is, the estimation function 106f specifies the position on the PET image of the position information based on the TOF information (the position of the pair annihilation point corresponding to the TOF information), and estimates the amount of time shift from the specified position and the position of the pair annihilation point in the PET image.

[0094] As a result, the PET device 100 according to the third modification can omit the process related to the generation of the TOF depiction image, and can reduce the processing load related to the TOF calibration. When constructing a trained model that estimates the time shift amount using the TOF information, the TOF information is used instead of the TOF depiction image in constructing the trained model shown in FIG. 8A.

[0095] (Variation 4) In the above-described embodiment, a case has been described in which TOF calibration is performed using list mode data acquired from a subject P. However, the embodiment is not limited to this, and for example, TOF calibration according to the present application may be performed using list mode data acquired from an external radiation source.

[0096] The term "processor" used in the above description means a circuit such as a CPU (Central Processing Unit), a GPU (Graphical Processing Unit), an Application Specific Integrated Circuit (ASIC), a programmable logic device (e.g., a Simple Programmable Logic Device (SPLD), a Complex Programmable Logic Device (CPLD), and a Field Programmable Gate Array (FPGA)). The processor reads a program stored in memory 130 and executes the read program to realize a function.

[0097] Here, the program executed by the processor is provided in advance in a ROM (Read Only Memory) or a storage circuit. The program may be provided by being recorded in a computer-readable non-transitory storage medium such as a CD (Compact Disk)-ROM, a FD (Flexible Disk), a CD-R (Recordable), or a DVD (Digital Versatile Disk) in a format that can be installed in these devices or in a format that can be executed. The program may also be provided or distributed by being stored on a computer connected to a network such as the Internet and downloaded via the network. For example, the program is composed of modules including each of the above-mentioned processing functions. As actual hardware, a CPU reads out the program from a storage medium such as a ROM and executes it, so that each module is loaded onto a main storage device and generated on the main storage device.

[0098] According to at least one of the embodiments described above, it is possible to easily perform TOF calibration.

[0099] Although some embodiments have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their modifications are included in the scope of the invention and its equivalents as described in the claims, as well as in the scope and spirit of the invention. [Explanation of symbols]

[0100] 100 PET equipment 102 Counting information generating circuit 106b Reconfiguration function 106d Specific Functions 106e generation function 106f Estimation Function 106g Calibration Function

Claims

1. an acquisition unit for acquiring list mode data; a reconstruction unit that reconstructs a PET image based on the list mode data; an identifying unit that identifies a first coincidence data item that is counted by a first detector and a second detector different from the first detector, from among a plurality of coincidence data items included in the list mode data; a generation unit that generates a TOF depiction image in which a pair annihilation point position is estimated based on TOF information included in the first coincidence data; an estimation unit that estimates a time lag amount indicating a lag between a reaction time of the first detector to incident gamma rays and a reaction time of the second detector to incident gamma rays, based on TOF information included in the first coincidence data and the PET image; Equipped with the estimation unit calculates a first centroid position indicating a centroid of image information in the entire TOF depiction image and a second centroid position indicating a position of a centroid of image information in the entire PET image, and estimates the time shift amount using a shift amount between the first centroid position and the second centroid position and a positional relationship between the first detector and the second detector. PET device.

2. The PET device according to claim 1 , further comprising a calibration unit that calibrates time information of the first detector or the second detector based on the amount of time shift.

3. The PET apparatus according to claim 1 , wherein the reconstruction unit reconstructs the PET image without using TOF information of each of the plurality of coincidence data included in the list mode data.

4. 2. The PET device according to claim 1, wherein the identifying unit limits at least one of the first detector and the second detector to a predetermined detector, and identifies first coincidence data counted by the first detector and the second detector.

5. The PET device according to claim 2 , wherein the calibration unit calibrates the time information of the first detector or the second detector on condition that the time deviation is within a threshold value.

6. The PET device according to any one of claims 1 to 5, wherein the estimation unit estimates the time shift amount using a trained model that has been trained to output a time shift amount in response to input of TOF information and a reconstructed image.

7. Get listmode data, reconstructing a PET image based on the list-mode data; Identifying a first coincidence data item among a plurality of coincidence data items included in the list mode data, the first coincidence data item being counted by a first detector and a second detector different from the first detector; generating a TOF depiction image in which the position of the annihilation point is estimated based on the TOF information included in the first coincidence data; estimating a time lag amount indicating a lag between a reaction time of the first detector to the incident gamma ray and a reaction time of the second detector to the incident gamma ray based on TOF information included in the first coincidence data and the PET image; This includes: The method includes calculating a first centroid position indicating the center of gravity of image information in the entire TOF depiction image and a second centroid position indicating the position of the center of gravity of image information in the entire PET image, and estimating the time shift amount using the shift amount between the first centroid position and the second centroid position and the positional relationship between the first detector and the second detector.

8. Get listmode data, reconstructing a PET image based on the list-mode data; Identifying a first coincidence data item among a plurality of coincidence data items included in the list mode data, the first coincidence data item being counted by a first detector and a second detector different from the first detector; generating a TOF depiction image in which the position of the annihilation point is estimated based on the TOF information included in the first coincidence data; estimating a time lag amount indicating a lag between a reaction time of the first detector to the incident gamma ray and a reaction time of the second detector to the incident gamma ray based on TOF information included in the first coincidence data and the PET image; Each process is executed by a computer, The process of estimating the time shift amount includes calculating a first center of gravity position indicating the center of gravity of image information in the entire TOF depiction image and a second center of gravity position indicating the position of the center of gravity of image information in the entire PET image, and estimating the time shift amount using the shift amount between the first center of gravity position and the second center of gravity position and the positional relationship between the first detector and the second detector.

9. An acquisition unit for acquiring list mode data; a reconstruction unit that reconstructs a PET image based on the list mode data; an identifying unit that identifies a first coincidence data item that is counted by a first detector and a second detector different from the first detector, from among a plurality of coincidence data items included in the list mode data; a generation unit that generates a TOF depiction image in which a pair annihilation point position is estimated based on TOF information included in the first coincidence data; an estimation unit that estimates a time lag amount indicating a lag between a reaction time of the first detector to incident gamma rays and a reaction time of the second detector to incident gamma rays, based on TOF information included in the first coincidence data and the PET image; Equipped with The estimation unit extracts a first region showing a relatively high count in the TOF depiction image and a second region showing a relatively high count in the PET image, and estimates the time shift amount from the shift amount between the position of the first region and the position of the second region and the positional relationship between the first detector and the second detector.

10. A PET device as described in claim 9, further comprising a calibration unit that calibrates the time information of the first detector or the second detector based on the time shift amount.

11. A PET device as described in Claim 9, wherein the reconstruction unit reconstructs the PET image without using TOF information for each of the multiple coincidence data included in the list mode data.

12. A PET device as described in Claim 9, wherein the identification unit limits at least one of the first detector and the second detector to a predetermined detector and identifies first coincidence data counted simultaneously by the first detector and the second detector.

13. A PET device as described in claim 10, wherein the calibration unit calibrates the time information of the first detector or the second detector, provided that the time shift amount is within a threshold value.

14. A PET device described in any one of claims 9 to 13, wherein the estimation unit estimates the time shift amount using a trained model trained to output the time shift amount in response to input TOF information and a reconstructed image.

15. Acquire list mode data, reconstructing a PET image based on the list-mode data; identifying a first coincidence data item counted by a first detector and a second detector different from the first detector among a plurality of coincidence data items included in the list mode data; generating a TOF depiction image in which the position of the annihilation point is estimated based on the TOF information included in the first coincidence data; estimating a time lag amount indicating a lag between a reaction time of the first detector to the incident gamma ray and a reaction time of the second detector to the incident gamma ray based on TOF information included in the first coincidence data and the PET image; This includes: The method includes extracting a first region showing a relatively high count in the TOF depiction image and a second region showing a relatively high count in the PET image, and estimating the time shift amount from the shift amount between the position of the first region and the position of the second region and the positional relationship between the first detector and the second detector.

16. Acquire list mode data, reconstructing a PET image based on the list-mode data; identifying a first coincidence data item counted by a first detector and a second detector different from the first detector among a plurality of coincidence data items included in the list mode data; generating a TOF depiction image in which the position of the annihilation point is estimated based on the TOF information included in the first coincidence data; estimating a time lag amount indicating a lag between a reaction time of the first detector to the incident gamma ray and a reaction time of the second detector to the incident gamma ray based on TOF information included in the first coincidence data and the PET image; Each process is executed by a computer, The process of estimating the time shift amount includes extracting a first region showing a relatively high count in the TOF depiction image and a second region showing a relatively high count in the PET image, and estimating the time shift amount from the shift amount between the position of the first region and the position of the second region and the positional relationship between the first detector and the second detector.

Images