Nuclear medicine diagnostic device, data processing method, and program

Abstract

To improve image quality.SOLUTION: A nuclear medicine diagnostic device comprises an acquisition section, a calculation section, an identification section, a measurement section, and a correction section. The acquisition section acquires first photon-number information detected by a first detector. The calculation section calculates a first light emission probability model corresponding to the first detector on the basis of the first photon-number information. The identification section identifies first timing at which detection probability becomes equal to or greater than a predetermined threshold value on the basis of the first light emission probability model. The measurement section measures detection timing of an event detected by the first detector. The correction section corrects the detection timing on the basis of the first timing.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] Embodiments disclosed herein and in the drawings relate to nuclear medicine diagnostic devices, data processing methods, and programs. [Background technology]

[0002] When reconstructing PET images, a Time of Flight (ToF) kernel is used. A ToF kernel is a probability density function of a detection event, expressed as a function of the detection time difference between signals detected by two detectors. For PET image reconstruction, it is desirable to use a ToF kernel with good accuracy. Furthermore, in PET image reconstruction, a narrower full width at half maximum (FWHM) of the ToF spectrum is preferable.

[0003] From this perspective, it is desirable to be able to accurately estimate the time from when gamma rays enter the scintillator until light emission occurs. One possible approach is to learn the relationship between the detected signal waveform and the detection time using a neural network. However, when using a neural network for learning, it is sometimes necessary to learn the relationship between the detected signal waveform and the detection time, which can require significant computational resources. [Prior art documents] [Non-patent literature]

[0004] [Non-Patent Document 1] Eric Berg and Simon R Cherry, "Using convolutional neural networks to estimate time-of-flight from PET detector waveforms", vol. 63, no 2, Phys. Med. Biol, 2018 [Overview of the Initiative] [Problems that the invention aims to solve]

[0005] One of the problems that the embodiments disclosed in this specification and drawings aim to solve is to improve image quality. However, the problems that the embodiments disclosed in this specification and drawings aim to solve 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] The nuclear medicine diagnostic apparatus according to the embodiment comprises an acquisition unit, a calculation unit, a specification unit, a measurement unit, and a correction unit. The acquisition unit acquires first photon count information detected by a first detector. The calculation unit calculates a first emission probability model corresponding to the first detector based on the first photon count information. The specification unit identifies a first timing at which the detection probability is greater than or equal to a predetermined threshold, based on the first emission probability model. The measurement unit measures the detection timing of an event detected by the first detector. The correction unit corrects the detection timing based on the first timing. [Brief explanation of the drawing]

[0007] [Figure 1] Figure 1 is a diagram showing an example of the configuration of a nuclear medicine diagnostic device according to an embodiment. [Figure 2] Figure 2 is a flowchart showing the processing flow of the nuclear medicine diagnostic device according to the embodiment. [Figure 3] Figure 3 is a flowchart that provides a more detailed explanation of an example of the process in step S30 of Figure 2. [Figure 4] Figure 4 is a diagram illustrating the processing performed by the nuclear medicine diagnostic device according to this embodiment. [Figure 5] Figure 5 is a flowchart that provides a more detailed explanation of an example of the process in step S300 of Figure 3. [Figure 6] Figure 6 is a diagram illustrating the processing performed by the nuclear medicine diagnostic device according to this embodiment. [Figure 7] Figure 7 is a diagram illustrating the processing performed by the nuclear medicine diagnostic device according to this embodiment. [Figure 8] Figure 8 is a diagram illustrating the processing performed by the nuclear medicine diagnostic device according to this embodiment. [Figure 9] Figure 9 is a flowchart illustrating in more detail another example of the process in step S300 of Figure 2. [Figure 10] Figure 10 is a diagram illustrating the processing performed by the nuclear medicine diagnostic device according to this embodiment. [Figure 11] Figure 11 is a diagram illustrating an example of the calculation results according to the embodiment. [Figure 12] Figure 12 illustrates an example of the calculation results according to this embodiment. [Figure 13] Figure 13 illustrates an example of the calculation results according to this embodiment. [Figure 14] Figure 14 is a diagram illustrating an example of the calculation results according to the embodiment. [Modes for carrying out the invention]

[0008] (Embodiment) The embodiments of the nuclear medicine diagnostic device, data processing method, and program will be described in detail below with reference to the drawings.

[0009] Figure 1 shows the configuration of a PET apparatus 100 as a nuclear medicine diagnostic device according to an embodiment. As shown in Figure 1, the PET apparatus 100 according to an embodiment comprises a stand device 1 and a console device 2. The stand device 1 comprises a detector 3, a front-end circuit 102, a top plate 103, a patient bed 104, and a patient bed drive unit 106.

[0010] Detector 3 is a detector that detects radiation by detecting scintillation light (fluorescence), which is light re-emitted when a substance, excited by the interaction of gamma rays generated when positrons emitted from a drug administered to and accumulated in a subject undergo annihilation with electrons in the surrounding tissue, and then transition back to its ground state. In this embodiment, detector 3 can also detect Cherenkov radiation. Detector 3 detects the energy information of the gamma ray radiation generated when positrons emitted from a drug administered to and accumulated in a subject undergo annihilation with electrons in the surrounding tissue. Multiple detectors 3 are arranged in a ring shape surrounding the subject P, and consist of, for example, multiple detector blocks.

[0011] The detector 3 typically consists of a scintillator crystal and a photodetector surface made of a photodetector element.

[0012] As the material for the scintillator crystal, for example, a material suitable for generating Cherenkov light, such as bismuth germanate (BGO, Bismuth Germanium Oxide), lead glass (SiO2 + PbO), lead fluoride (PbF2), and lead compounds such as PWO (PbWO4), can be used. Alternatively, as another example, scintillator crystals such as LYSO (Lutetium Yttrium Oxyorthosilicate), LSO (Lutetium Oxyorthosilicate), LGSO (Lutetium Gadolinium Oxyorthosilicate), or BGO may be used. The photodetector element constituting the photodetector surface 3b consists of, for example, multiple pixels, each of which is composed of, for example, a SPAD (Single Photon Avalanche Diode). Furthermore, the configuration of the detector 3 is not limited to the above example; as an example, the photodetector element may be, for example, a SiPM (Silicon photomultiplier) or a photomultiplier tube.

[0013] The scintillator crystal may be a monolithic crystal, and the photodetector surface, which consists of a photodetector element, may be arranged on the six faces of the scintillator crystal, for example. In the following embodiments, we will first explain the case where the scintillator crystal of the detector 3 is not a monolithic crystal.

[0014] Furthermore, the mounting device 1 generates counting information from the output signal of the detector 3 using the front-end circuit 102, and stores the generated counting information in the storage unit 130 of the console device 2. The detector 3 is divided into multiple blocks and is equipped with the front-end circuit 102.

[0015] The front-end circuit 102 converts the output signal of the detector 3 into digital data and generates counting information. This counting information includes the detection position, energy value, and detection time of the annihilation gamma ray. For example, the front-end circuit 102 identifies multiple photodetectors that converted scintillation light into electrical signals at the same time. The front-end circuit 102 then identifies the scintillator number (P) that indicates the position of the scintillator into which the annihilation gamma ray was incident. The means for identifying the scintillator position into which the annihilation gamma ray was incident may be determined by performing a centroid calculation based on the position of each photodetector and the intensity of the electrical signal. Alternatively, if the size of the scintillator and the respective photodetectors correspond, for example, the scintillator corresponding to the photodetector that obtained the maximum output may be assumed to be the scintillator position into which the annihilation gamma ray was incident, and the final identification may be made by considering inter-scintillator scattering.

[0016] Furthermore, the front-end circuit 102 calculates the energy value (E) of the annihilation gamma ray incident on the detector 3 by integrally calculating the intensity of the electrical signal output from each photodetector or by measuring the time over which the electrical signal intensity exceeds a threshold. The front-end circuit 102 also identifies the detection time (T) at which the detector 3 detects the scintillation light from the annihilation gamma ray. Note that the detection time (T) may be an absolute time or the elapsed time from the start of imaging. In this way, the front-end circuit 102 generates counting information including the scintillator number (P), energy value (E), and detection time (T).

[0017] The front-end circuit 102 is implemented by circuits such as a CPU (Central Processing Unit), a GPU (Graphical Processing Unit), an Application Specific Integrated Circuit (ASIC), or 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 front-end circuit 102 is an example of a front-end section.

[0018] The top plate 103 is a bed on which the subject P is placed, and is positioned on the bed 104. The bed drive unit 106 moves the top plate 103 under the control of the control function 105g of the processing circuit 150. For example, the bed drive unit 106 moves the top plate 103 to move the subject P into the shooting opening of the mounting device 1.

[0019] The console device 2 receives input from the operator to the PET device 100, controls the acquisition of PET images, and reconstructs the PET images using the counting information collected by the rigging device 1. As shown in Figure 1, the console device 2 comprises a processing circuit 150, an input device 110, a display 120, and a storage unit 130. The various components of the console device 2 are connected via a bus. Details of the processing circuit 150 will be described later.

[0020] The input device 110 is a mouse, keyboard, or the like used by the operator of the PET scanner 100 to input various instructions and settings, and it transfers the input instructions and settings to the processing circuit 150. For example, the input device 110 is used to input the instruction to start imaging.

[0021] The display 120 is a monitor or similar device referenced by the operator, and under the control of the processing circuit 150, it displays the respiratory waveform and PET images of the subject, as well as a GUI (Graphical User Interface) for receiving various instructions and settings from the operator.

[0022] The memory unit 130 stores various data used in the PET apparatus 100. The memory unit 130 is composed of, for example, memory, and can be implemented by semiconductor memory elements such as RAM (Random Access Memory) and flash memory, or by hard disks, optical discs, etc. The memory unit 130 stores counting information, which is information that associates the scintillator number (P), energy value (E), and detection time (T); coincidence information, which is a sequential number of the coincidence information, which associates a set of counting information with a coincidence number; projection data obtained by aggregating the coincidence information; reconstructed PET images, etc.

[0023] The processing circuit 150 has an acquisition function 150a, a calculation function 150b, a specific function 150c, a measurement function 150d, a correction function 150e, a reconstruction function 150f, a control function 150g, a reception function 150h, an image generation function 150i, and a display control function 150j. The functions of acquisition function 150a, calculation function 150b, specific function 150c, correction function 150e, and reconstruction function 150f will be explained in detail later.

[0024] In this embodiment, the system includes an acquisition function 150a, a calculation function 150b, a specific function 150c, a measurement function 150d, a correction function 150e, a reconstruction function 150f, a control function 150g, a reception function 150h, an image generation function 150i, and a display control function 150j. The processing functions performed by the acquisition function 150a, calculation function 150b, specific function 150c, correction function 150e, and reconstruction function 150f are stored in the storage unit 130 in the form of programs that can be executed by a computer. The processing circuit 150 is a processor that reads programs from the storage unit 130 and executes them to realize the functions corresponding to each program. In other words, the processing circuit 150 in the state where each program has been read will have the functions shown in the processing circuit 150 of Figure 1.

[0025] In Figure 1, the processing functions performed by the acquisition function 150a, calculation function 150b, identification function 150c, measurement function 150d, correction function 150e, reconstruction function 150f, control function 150g, reception function 150h, image generation function 150i, and display control function 150j are described as being realized by a single processing circuit 150. However, it is also possible to configure the processing circuit 150 by combining multiple independent processors, and each processor can realize the functions by executing a program. In other words, each of the above functions may be configured as a program, and a single processing circuit 150 may execute each program. As another example, a specific function may be implemented in a dedicated, independent program execution circuit.

[0026] In the above description, the term "processor" refers to circuits such as a CPU (Central Processing Unit), a GPU (Graphical Processing Unit), an Application Specific Integrated Circuit (ASIC), or 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 functions by reading and executing programs stored in the memory unit 130.

[0027] In Figure 1, the acquisition function 150a, calculation function 150b, identification function 150c, measurement function 150d, correction function 150e, reconstruction function 150f, control function 150g, reception function 150h, image generation function 150i, and display control function 150j are examples of the acquisition unit, calculation unit, measurement unit, correction unit, reconstruction unit, control unit, reception unit, image generation unit, and display control unit, respectively.

[0028] Alternatively, instead of the processing circuit 150, the front-end circuit 102 may handle the processing of the measurement unit, the identification unit, etc.

[0029] The processing circuit 150 reconstructs the PET image based on the data acquired from the front-end circuit 102 using the reconstruction function 150f, and generates the image using the image generation function 150i.

[0030] The processing circuit 150 controls the PET apparatus 100 as a whole by controlling the gantry device 1 and the console device 2 using the control function 150g. For example, the processing circuit 150 controls imaging in the PET apparatus 100 using the control function 150g. Also, the processing circuit 105 controls the patient table drive unit 106 using the control function 150g.

[0031] The processing circuit 150 receives information input from the user via the input device 110 through its reception function 150h. The processing circuit 150 also displays PET images and other data on the display 120 through its display control function 150j. Furthermore, the processing circuit 150 generates various images through its image generation function 150i.

[0032] Next, I will briefly explain the background of this embodiment.

[0033] When reconstructing PET images, a Time of Flight (ToF) kernel is used. A ToF kernel is a probability density function of a detection event, expressed as a function of the detection time difference between signals detected by two detectors. For PET image reconstruction, it is desirable to use a ToF kernel with good accuracy. Furthermore, in PET image reconstruction, a narrower full width at half maximum (FWHM) of the ToF spectrum is preferable.

[0034] From this perspective, it is desirable to be able to accurately estimate the time from when gamma rays enter the scintillator until light emission occurs. One possible approach is to learn the relationship between the detected signal waveform and the detection time using a neural network. However, when using a neural network for learning, it is sometimes necessary to learn the relationship between the detected signal waveform and the detection time, which can require significant computational resources.

[0035] Therefore, in this embodiment, the delay time from when gamma rays enter the scintillator until emission occurs is estimated based on the emission model, and the ToF kernel is corrected based on the estimation result. This makes it possible to sharpen the ToF spectrum and improve the image quality of the reconstructed PET image.

[0036] Specifically, the nuclear medicine diagnostic apparatus according to the embodiment comprises an acquisition unit, a calculation unit, a specification unit, a measurement unit, and a correction unit. The acquisition unit acquires first photon count information detected by the first detector. The calculation unit calculates a first emission probability model corresponding to the first detector based on the first photon count information. The specification unit identifies a first timing at which the detection probability exceeds a predetermined threshold, based on the first emission probability model. The measurement unit measures the detection timing of an event detected by the first detector. The correction unit corrects the detection timing based on the first timing.

[0037] Furthermore, the data processing method according to the embodiment acquires first photon count information detected by the first detector, calculates a first emission probability model corresponding to the first detector based on the first photon count information, identifies a first timing at which the detection probability exceeds a predetermined threshold based on the first emission probability model, measures the detection timing of an event detected by the first detector, and corrects the detection timing based on the first timing.

[0038] Furthermore, the program according to the embodiment acquires first photon count information detected by the first detector, calculates a first emission probability model corresponding to the first detector based on the first photon count information, identifies a first timing at which the detection probability exceeds a predetermined threshold based on the first emission probability model, measures the detection timing of an event detected by the first detector, and causes the computer to perform a process to correct the detection timing based on the first timing.

[0039] The processing performed by the nuclear medicine diagnostic device according to this embodiment will be explained in detail below using Figures 2 to 10. Figure 2 shows the processing flow performed by the PET device 100 as a nuclear medicine diagnostic device according to this embodiment. For simplicity, the following flowchart only describes the case where a probability distribution model (ToF kernel) for the first detector l and the second detector m is generated. However, typically, in image reconstruction, the processing circuit 150 generates a probability distribution model for all pairs of l and m and then performs the image reconstruction processing in step S70. In this case, the processing circuit 150 repeats the processing from steps S10 to S60 for all pairs of l and m and then performs the processing in step S70.

[0040] In step S10, the processing circuit 150 selects values ​​for l and m that indicate a pair of detectors to be processed. That is, the processing circuit 150 selects values ​​for l and m that characterize the first detector l (the l-th detector), which is the detector that detects the first event, and the second detector m (the m-th detector), which is the detector that detects the second event. The first event and the second event here constitute one simultaneous counting event. In other words, each simultaneous counting event consists of two events: the first event and the second event.

[0041] Next, in step S20, the processing circuit 150 sets the value of i, which indicates the number of the coincidence event, to 1. Here, i=1 means that the processing circuit 150 will process the first coincidence event.

[0042] Next, in step S30, the processing circuit 150 uses the calculation function 150b to calculate the probability distribution model p related to coincidence for the i-th coincidence event among the coincidence events between the event detected by the l-th detector and the event detected by the m-th detector. ct;l、m i、(doi) (θ l i ,θ m i Calculate t). Here, θ li , θ m i represents the parameters obtained for the l-th detector and the m-th detector, respectively, in the i-th coincidence event. Also, the symbol (doi) means that the uncertainty of the light emission position in the scintillator may or may not be considered. In an embodiment where the uncertainty of the light emission position in the scintillator is considered, the processing circuit 150 uses the calculation function 150b to calculate the probability distribution model p ct;l、m i、doi (θ l i , θ m i , t) regarding coincidence considering the uncertainty of the light emission position in the scintillator. In an embodiment where the uncertainty of the light emission position in the scintillator is not considered, it means calculating the probability distribution model p ct;l、m i (θ l [[ID=I8]] i , θ m i , t) regarding coincidence without considering the uncertainty of the light emission position in the scintillator.

[0043] Figure 3 shows a more detailed processing flow for the processing in step S30 of Figure 2. That is, steps S100 to S500 in Figure 3 explain step S30 in Figure 2 in more detail.

[0044] First, in step S100, the processing circuit I50 uses the acquisition function 150a to acquire the first photon number information regarding the first event detected by the first detector l for the i-th coincidence event.

[0045] Here, the first photon number information is, for example, the number of scintillation photons N i S;l for the event. Also, as another example, the first photon number information is, for example, the number of Cherenkov photons N i C;lAs another example, the first photon number information refers to the scintillation photon number N for the event in question. i S;l and Cherenkov photon number N i C;l That is, the processing circuit 150, through the acquisition function 150a, acquires the number of scintillation photons N i S;l , Cherenkov photon number N i C;l , or the number of scintillation photons N i S;l and Cherenkov photon number N i C;l This is obtained as the first photon count information.

[0046] The first photon count information (and second photon count information) acquired by the processing circuit 150 using the acquisition function 150a is typically acquired for each event. However, the embodiment is not limited to this, and the processing circuit 150 may acquire, for example, the average number of photons detected by the first detector l as the first photon count information (and second photon count information) using the acquisition function 150a.

[0047] Furthermore, the first photon count information (and second photon count information) acquired by the processing circuit 150 through the acquisition function 150a is typically acquired for each detector, for example. However, the embodiment is not limited to this, and the processing circuit 150 may acquire the same photon count information across all detectors, for example, as the first photon count information (and second photon count information), etc., through the acquisition function 150a.

[0048] Furthermore, the PET apparatus 100 may acquire first photon number information based on the energy spectrum acquired by the first detector l.

[0049] Figure 4 shows an explanation of the emission model within the detector 3 and the photon spectrum acquired by the processing circuit 150 using the acquisition function 150a. When gamma rays 10 are incident on the scintillator 11 of the detector 3, the gamma rays 10 impart energy to the scintillator 11 in the form of photoelectric absorption or Compton scattering, and high-energy electrons 12 are generated. Curve 16 shows the energy E of the high-energy electrons 12. e This is expressed as a function of time t. The high-energy electron 12 loses energy as it travels through the scintillator 11, but the energy threshold E th While larger, it emits Cherenkov light-13 into the surroundings.

[0050] Curve 17 represents the photon density n of Cherenkov light. c (t) is approximated as a Gaussian function of time t. Specifically, the photon density n of Cherenkov radiation. c (t) is N c The number of photons in Cherenkov light, σ C Let be the full width at half maximum of the Cherenkov light spectrum and t0 be the peak time of the Cherenkov light. The result is given by the following equation (1).

[0051]

number

[0052] Meanwhile, as the high-energy electrons 12 travel, they excite valence electrons 14 within the scintillator 11, and when the excited valence electrons 14 return to the ground state, scintillation light 15 is emitted. Curve 19 shows the excitation density n of the valence electrons. ex (t) is approximated by a Gaussian function at time t. Valence electron excitation density n ex (t) is N s The number of photons in scintillation light, σ S Let be the full width at half maximum of the scintillation light spectrum and t1 be the peak time of the scintillation light. The result is given by the following equation (2).

[0053]

number

[0054] Curve 21 shows how excited valence electrons transition back to their original state at time 20. Scintillation light is generated as excited valence electrons transition back to their original state. Curve 22 shows the photon density n of the scintillation light. s (t) is shown as a function of time t. The photon density n of the scintillation light. s (t) is the relaxation constant τ S As such, the excitation density of valence electrons n ex Using (t), it is given by the following equation (3).

[0055]

number

[0056] Here, the excitation density of valence electrons n ex If we assume that the functional form of (t) is given by equation (2), then substituting equation (2) into equation (3) yields the following equation (4).

[0057]

number

[0058] In other words, as an example, the PET apparatus 100 may acquire energy spectra obtained from the first detector l and the second detector m, generate spectra shown in curves 17 and 22 based on these energy spectra, and acquire first photon count information and second photon count information based on the generated spectra.

[0059] Note that the emission distribution function p ph (t) is the photon density n of Cherenkov light. c (t) and the photon density n of scintillation light s Since it is proportional to the sum with (t), the following equation (5) holds true.

[0060]

number

[0061] Here, the luminescence distribution function p ph (t) is normalized as shown in equation (6) below.

[0062]

number

[0063] In step S100, the processing circuit 150 may acquire information other than the number of photons as a parameter set for the first detector l using the acquisition function 150a. For example, the processing circuit 150 may acquire the time resolution σ of the photosensor with respect to Cherenkov light using the acquisition function 150a. C The parameter set θ obtained in the first detector l is either in place of the first photon number information or in addition to the first photon number information. l As one example, it may be acquired. Another example is that the processing circuit 150, by acquisition function 150a, acquires the time resolution σ of the photosensor with respect to scintillation light. S The parameter set θ obtained in the first detector l is either in place of the first photon number information or in addition to the first photon number information. l As one example, it may be acquired. Another example is that the processing circuit 150 acquires the probability of excessive noise generation at the first detector l using the acquisition function 150a, and the parameter set θ l It may be acquired as one of the methods. The probability of excessive noise occurring here refers to the probability of APD (Avalanche photodiode) excessive noise occurring, such as optical crosstalk or afterpulse.

[0064] Similar to step S100, in step S200, the processing circuit 150, using the acquisition function 150a, acquires second photon count information for the i-th simultaneous counting event, relating to the second event detected by a second detector m, which is different from the first detector l.

[0065] Here, the second photon number information refers, for example, to the number of scintillation photons N for the corresponding coincidence counting event. i S;m As another example, the second photon number information is, for example, the Cherenkov photon number N for the corresponding coincidence event. i C;m As another example, the second photon number information refers to the scintillation photon number N for that coincidence counting event. i S;m and Cherenkov photon number N i C;m That is, the processing circuit 150, through the acquisition function 150a, acquires the number of scintillation photons N i S;m , Cherenkov photon number N i C;m , or the number of scintillation photons N i S;m and Cherenkov photon number N i C;m This is acquired as second photon count information. Here, the second photon count information measured by the processing circuit 150 using the acquisition function 150a is typically acquired for each event.

[0066] The PET apparatus 100 may acquire second photon number information based on the energy spectrum acquired by the second detector m.

[0067] Furthermore, the processing circuit 150, through the acquisition function 150a, acquires the time resolution σ of the photosensor for Cherenkov light. C , the time resolution of the photosensor for scintillation light σ S The parameter set θ is used to acquire the probability of excessive noise generation, etc., in the second detector m. m You may acquire it as such.

[0068] Following step S100, in step S300, the processing circuit 150, using the calculation function 150b, calculates the first emission probability model p for the first detector l in the i-th simultaneous counting event, based on the first photon count information obtained in step S100. t;l i,(doi)Calculate it. Note that the processing circuit 150 determines the first light emission probability model p for the first detector l by the calculation function 150b. t;l i,(doi) Typically, it is determined for each coincidence counting event. In the following embodiments, the processing circuit 150 determines the first light emission probability model p for the first detector l by the calculation function 150b. t;l i,(doi) It will be described as being determined for each coincidence counting event. However, the embodiment is not limited to this, and the processing circuit 150 may determine the first light emission probability model for the first detector l only on a detector-by-detector basis by the calculation function 150b.

[0069] An example of the processing of step S300 is described in more detail in FIG. 5. The flowchart of FIG. 5 is a flowchart that describes an example of the processing of step S300 in more detail.

[0070] In step S310, the processing circuit 150 generates a light emission distribution function model p(t) at each time based on the first photon number information in the first detector l acquired in step S100 by the calculation function 150b. The light emission distribution function p(t) is a distribution function indicating the probability of light emission occurring at time t. The light emission distribution function model p(t) integrates to 1 over all time with respect to time. Also, the light emission distribution function model p(t) is determined for each detector and each coincidence counting event. i ph,l (t) ph [ (t) i ph,l (t) <0(000093> ph,l (t)

[0071] The processing circuit 150 generates the light emission distribution function model p(t) based on the first photon number information by the calculation function 150b, for example, using equations (1), (4), (5), (6), etc. Since the first photon number information is acquired for each detector and each coincidence counting event, the processing circuit 150 generates the light emission distribution function model p(t) by the calculation function 150b. i ph,l (t) ph,l i(t) is calculated for each detector and for each simultaneous counting event. Figure 6 shows the emission distribution function p ph An example of (t) is shown. In Figure 6, curve 30 is the emission distribution function p ph (t) is shown. The photon threshold (N') for defining the detection time is also indicated.

[0072] Next, in step S320A, the processing circuit 150 calculates the emission distribution function model p using the calculation function 150b. i ph,l (θ i l Based on t), the first emission probability model p i t,l (≧N',θ i l, Calculate t). Here, the first emission probability model p i t,l (≧N',θ i l, t) is a model relating to the probability density of detecting the first N' or more photons among multiple photons included in the detection event. In other words, the first emission probability model p t,l (≧N',t) is a model relating to the probability density of detecting a predetermined number of photons from among multiple photons included in the detected event. Figure 7 shows the first emission probability model p t,l (Δt) is shown. Here, the photon detection threshold N' is fixed to a certain value and is therefore not shown. Curve 31 in Figure 7 represents the first emission probability model p t,l The general shape of (Δt) is shown. Note that the first luminescence probability model p t,l (Δt) is determined for each detector and for each coincidence event. The following are the subscripts for the detector, the subscript representing the number of the coincidence event, and the argument θ representing the set of parameters to be acquired, as needed. i l Details such as the above will be omitted as appropriate. Note that curve 32 represents the first luminescence probability model p t,l This curve represents the cumulative distribution P(t) of (t). Curve 32 will be discussed later.

[0073] Next, the first luminescence probability model p t,lThe method for calculating (≧N',t) will be explained. The first emission probability model p is the probability density for detecting the first N' or more photons out of multiple photons included in the detected event. t (≧N',t) can be calculated as follows:

[0074] First, we derive an equation for the probability P'(≧N',t) of detecting fewer than N' photons by time t. This is the probability that at most N'-1 photons have been detected by time t, and can be expressed by equation (7) below using the probability Pd(k,t) of detecting k photons by time t.

[0075]

number

[0076] The probability of detecting k photons by time t can be expressed in differential form as shown in equation (8) below, by evaluating the photon detection events using a binomial distribution.

[0077]

number

[0078] Solving this differential equation sequentially for k=0, 1, 2, ..., we inductively obtain equation (9) below.

[0079]

number

[0080] From equation (7), the probability P'(≧N',t) that fewer than N' photons are detected by time t can be written as shown in equation (10).

[0081]

number

[0082] Substituting this into equation (8), we obtain the first emission probability model p, which is the probability density for detecting the first N' photons among the multiple photons included in the detected event. t (≧N',t) can be expressed as shown in equation (11) below.

[0083]

number

[0084] Here, the function F(t) is the emission distribution function p ph This is the cumulative distribution function of (t), and can be written as shown in equation (12) below.

[0085]

number

[0086] In the following examples, for simplicity, we will take the case where the probability distribution evaluated at the time of the first photon detection, i.e., N'=0, as an example. That is, the first emission probability model, which is the probability density of detecting the first photon, is p t Let (t) be the value, and express it explicitly as shown in equation (13) below.

[0087]

number

[0088] In other words, the processing circuit 150 uses the calculation function 150b to calculate the light emission distribution function p ph Based on (t), the first emission probability model p t (t) can be calculated. In this way, the processing circuit 150, using the calculation function 150b, can calculate the distribution function model of light emission in the first detector p in equations (12) and (13). ph By substituting (t) and the first photon count information into N on the right side of equation (13) and evaluating the right side of equation (13), we obtain the first photon count information and the emission distribution function model p at the first detector, which is defined for each simultaneous counting event. phBased on (t), the first luminescence probability model p for the first event t Calculate (t).

[0089] The above-mentioned emission distribution function p ph (t) and the first first luminescence probability model p t (t) can be determined for each detector and for each coincidence event. Therefore, if the subscripts for the detectors and the subscripts for the coincidence events are explicitly written, equation (13) can also be written as equation (14) below.

[0090]

number

[0091] Here, p i t;l (t) is the first emission probability model at detector l for the i-th simultaneous counting event, p i ph;l (t) is the emission distribution function at detector l for the i-th simultaneous counting event, N i s;l N is the number of photons of scintillation light in detector l for the i-th simultaneous counting event. i c;l This indicates the number of Cherenkov photons in detector l for the i-th simultaneous counting event.

[0092] Furthermore, when calculating the first emission probability model, the processing circuit 150 can also perform calculations that take into account the uncertainty of the emission position within the scintillator using the calculation function 150b. For example, as shown in Figure 8, if a gamma ray 40 is emitted at point 42 at a depth x in the scintillator 11, and the scintillation light propagates through a medium with refractive index n at the speed of light c / n in the material and reaches the edge 44 of the scintillator 11, as indicated by the arrow 43, the time from when the gamma ray 40 enters the scintillator until it reaches the edge of the scintillator 11 will differ depending on the emission position within the scintillator, and therefore the time at which the signal is detected will differ.

[0093] In other words, the time t1 required for the gamma ray 40 to reach a depth x after being incident on the scintillator 11 is given by equation (15) as follows:

[0094]

number

[0095] Furthermore, the time t2 required for the light emitted at point 42 to reach the edge of scintillator 11 from depth x is given by equation (16) as follows.

[0096]

number

[0097] Therefore, the time t' required from the time the gamma ray 40 enters the scintillator 11 until the scintillation light reaches the edge of the scintillator 11 is given by equation (17) as follows.

[0098]

number

[0099] From equation (17), it can be seen that the time at which the signal is detected differs depending on the light emission position x within the scintillator.

[0100] Based on the above, if we consider a formulation that takes into account the luminescence position x within the scintillator, based on equation (13), then when considering the treatment that takes into account the luminescence position x within the scintillator, the following equation (18) holds.

[0101]

number

[0102] Here, p t、l i,doi(t) is a first luminescence probability model that takes into account the uncertainty of the luminescence position within the scintillator, which is determined for each simultaneous counting event, and p ph、l i,doi (t) is a luminescence distribution function model that takes into account the uncertainty of the luminescence position at the scintillator's position, which is determined for each simultaneous counting event.

[0103] Equation (18) represents the parameter θ acquired by the processing circuit 150 in step S100 by the acquisition function 150a. l If we explicitly express it as a function argument, it can also be written as shown in equations (19) and (20) below.

[0104]

number

[0105]

number

[0106] Here, θ i l This represents the parameter set obtained at detector l, which is determined for each simultaneous counting event. For example, the number of scintillation photons N is determined for each simultaneous counting event. i s;l、 Cherenkov photon number N i c;l , the time resolution of the photosensor for Cherenkov light σ i C , the time resolution of the photosensor for scintillation light σ i S This represents a set of parameters such as those listed above.

[0107] Returning to equation (18), we will now explain the process in step S300 when considering the uncertainty of the emission position within the scintillator. Figure 9 is a flowchart showing the calculation flow of the emission probability model when considering the uncertainty of the emission position within the scintillator.

[0108] Similar to the case shown in Figure 5, in step S310, the processing circuit 150 uses the calculation function 150b to calculate parameters such as the first photon number information in the first detector l. l Based on this, the emission distribution function model p ph,l Generate (t).

[0109] Next, in step S311, the processing circuit 150 calculates the emission probability density p for each position x within the scintillator using the calculation function 150b. doi (x) is calculated based on the value of the attenuation coefficient μ. The luminescence probability density p at position x. doi (x) is expressed by the following equation (21).

[0110]

number

[0111] Note that the emission probability density p doi (x) is normalized as shown in equation (22) below.

[0112]

number

[0113] In Figure 8, graph 45 shows the attenuation rate μexp(-μx).

[0114] Returning to Figure 9, in step S312, the processing circuit 150, using the calculation function 150b, calculates the distribution function model p, which is the distribution function of light emission defined for each simultaneous counting event generated in step S310. i ph、l (t) and the emission probability density p doi Based on (x), a distribution function model of light emission p that takes into account the uncertainty of the light emission position determined for each simultaneous counting event is formed. ph;l i,doi (t) is calculated as the first luminescence probability model.

[0115] Specifically, the processing circuit 150 uses the calculation function 150b to calculate the emission distribution function model p, which takes into account the uncertainty of the emission position. i,doi ph;l (t) is the distribution function model p generated in step S310. i ph、l (t) and the emission probability density p calculated in step S311 doi Based on (x), the following equation (23) is used to calculate it.

[0116]

number

[0117] Here, L is the length of the scintillator, x is the emission position, and t' is the delay time, as explained in equation (17), from when the gamma ray enters the scintillator until the scintillation light reaches the edge of the scintillator.

[0118] Next, in step S320B, the processing circuit 150 calculates the emission distribution function model p, which takes into account the uncertainty of the emission position, using the calculation function 150b. i,doi ph;l Based on (t), the emission probability model p considers the same emission position uncertainty. i,doi t;l (θ l The t) is calculated. Specifically, as shown in equations (24) and (25), the processing circuit 150, using the calculation function 150b, calculates the emission distribution function model p that takes into account the uncertainty of the emission position, similar to equation (13). i,doi ph Based on (t), the emission probability model p considers the uncertainty of the emission position. i,doi t Calculate (t).

[0119]

number

[0120]

number

[0121] In addition, in Equation (23), when the variable x of the integrand is transformed into t' based on Equation (18), the following Equation (26) is obtained. Therefore, the processing circuit 150 may perform the calculation using Equation (26) in the evaluation of the right side of Equation (25) by the calculation function 150b.

[0122] [Number]

[0123] In this way, the processing circuit 150 calculates the first emission probability model and the second emission probability model by the calculation function 150b, taking into account the uncertainty of the emission position in the scintillator.

[0124] Returning to FIG. 3, in step S400, the processing circuit 150 performs the same processing as in step S300 by the calculation function 150b, and based on the second photon number information acquired in step S200, for the second detector m, the second emission probability model p defined for each coincidence counting event i、(doi) t;m is calculated.

[0125] Specifically, the processing circuit 150, by the calculation function 150b, based on the second photon number information defined for each coincidence counting event in the second detector m acquired in step S200, uses equations corresponding to, for example, equations (1), (4), (5), (6), etc., and generates the emission distribution function model p i ph;m (t).

[0126] Subsequently, the processing circuit 150, by the calculation function 150b, based on the second photon number information and the emission distribution function model p i ph;m (t) in the second detector m, calculates the second emission probability model p i t;m (t) defined for each coincidence counting event using an equation corresponding to, for example, equation (14), etc.

[0127] Furthermore, when calculating the second emission probability model, the processing circuit 150 can also use calculation function 150b to perform calculations that take into account the uncertainty of the emission position within the scintillator, using equations similar to those in equations (17) and (21) to (26). In this case, the processing circuit 150 uses calculation function 150b to calculate the emission probability density p for each position x within the scintillator. doi (x) is calculated based on the value of the attenuation coefficient μ. Subsequently, the processing circuit 150 uses the calculation function 150b to determine the distribution function model p for each simultaneous counting event. i ph;m (t) and the emission probability density p doi Based on (x), a distribution function model of light emission p that takes into account the uncertainty of the light emission position determined for each simultaneous counting event is formed. i、doi ph;m (t) is calculated. Next, the processing circuit 150 calculates the distribution function model p of the light emission, which takes into account the uncertainty of the light emission position determined for each simultaneous counting event, using the calculation function 150b. i、doi ph;m Based on (t), a probability model of light emission p that takes into account the uncertainty of the light emission position determined for each simultaneous counting event is formed. i、doi t;m Calculate (t).

[0128] In this way, the processing circuit 150, using the calculation function 150b, takes into account the uncertainty of the light emission position within the scintillator and calculates a second light emission probability model p determined for each simultaneous counting event. t,m Calculate (t).

[0129] Next, in step S410, the processing circuit 150 uses its specific function 150c to identify a first timing in which the detection probability of light emission is greater than or equal to a predetermined threshold, based on the first light emission probability model. As an example, the processing circuit 150 uses its specific function 150c to determine the estimated detection delay time Δt for each event, based on the first light emission probability model calculated in step S300. l、i p This is identified as the first timing.

[0130] Here, the estimated detection delay time is an estimated value of the time it takes from, for example, when X-rays enter the scintillator until the emission signal is detected. The processing circuit 150 can estimate the estimated detection delay time for each detector and for each simultaneous counting event using the specific function 150c. By estimating the estimated detection delay time and subtracting it from the detection time, the processing circuit 150 can sharpen the ToF spectrum and improve image quality.

[0131] Returning to Figure 7, let's explain the identification of the first timing. As a first method for identifying the first timing in which the detection probability of light emission is above a predetermined threshold, the processing circuit 150 uses the identification function 150c to determine the first light emission probability model p i,doi t;l (θ l Based on the cumulative distribution P(t) of the distribution shown by ;t), a first timing is identified in which the detection probability of luminescence exceeds a predetermined threshold. Here, the first luminescence probability model p i,doi t;l (θ l The cumulative distribution P(t) of the distribution shown by t is given, for example, by the following equation (27).

[0132]

number

[0133] Here, the processing circuit 150 uses a specific function 150c to determine the first luminescence probability model p i,doi t;l (θ l The time at which the cumulative distribution P(t) of the distribution shown by t) reaches a predetermined threshold p, i.e., the time 34 at the intersection of curve 32 and line 33, is defined as the first timing Δt p l,i This is identified as follows: Expressed as the first timing Δt p l,i This is given by equation (28) below.

[0134]

number

[0135] Furthermore, as a second method for identifying a first timing in which the detection probability of light emission exceeds a predetermined threshold, the processing circuit 150 uses the specific function 150c to determine the first light emission probability model p i,doi t;l (θ l The time 35 in which the distribution shown by ;t) takes an extremum is the first timing Δt p l,i This is identified as follows: Expressed as the first timing Δt p l,i This is given by equation (29) below.

[0136]

number

[0137] The processing circuit 150, through specific function 150c, determines the first timing Δt p l,i Typically, this is identified for each simultaneous counting event and for each detector. However, the embodiments are not limited thereto, and the processing circuit 150 may, by the identification function 150c, identify a first timing in which the detection probability of light emission is greater than or equal to a predetermined threshold, other than on an event-by-event or detector-by-detector basis.

[0138] Returning to Figure 3, in step S420, similar to step S410, the processing circuit 150 uses its specific function 150c to identify a second timing in which the detection probability of light emission exceeds a predetermined threshold, based on the second light emission probability model. As an example, the processing circuit 150 uses its specific function 150c to determine the estimated detection delay time Δt for each event, based on the second light emission probability model calculated in step S300. m、i p This is identified as the second timing.

[0139] As a first method for identifying a second timing at which the light emission detection probability is equal to or greater than a predetermined threshold, the processing circuit 150 uses a specific function 150c to determine a first light emission probability model p i,doi t;m (θ m ;t) to identify a first timing at which the light emission detection probability is equal to or greater than a predetermined threshold based on the cumulative distribution P(t) of the distribution represented by. As an example, the processing circuit 150 uses a specific function 150c to determine that the cumulative distribution P(t) of the first light emission probability model p i,doi t;m (θ m ;t) reaches a predetermined threshold p at a time, and identifies this time as the second timing Δt p m,i .

[0140] Also, as a second method for identifying a second timing at which the light emission detection probability is equal to or greater than a predetermined threshold, the processing circuit 150 uses a specific function 150c to determine that the time at which the distribution represented by the second light emission probability model p i,doi <00!​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​

[0143] In other words, the processing circuit 150 detects the event timing t detected by the first detector l using the measurement function 150d. l,i d The first timing Δt is measured and corrected based on the correction function 150e. p l,i Based on this, detection timing t l,i d Correct the value, and correct the detection time t l,i corrected To obtain.

[0144] Furthermore, the processing circuit 150, in conjunction with this, uses the correction function 150e to correct the detection time of the first emission probability model p i,doi t;l;corrected (θ m Calculate t). Here, the corrected first luminescence probability model p i,doi t;l;corrected (θ l ;t) and the first luminescence probability model p before correction i,doi t;l;corrected (θ l The relationship with t) is given, for example, by equation (31) below.

[0145]

number

[0146] For the sake of simplicity, in the processing from step S500 onward, the corrected first luminescence probability model p i,doi t;l;corrected (θ l ;t) is simply the first luminescence probability model p i,doi t;l (θ l ;t) is written, but in the subsequent process of calculating the probability distribution model for coincidence, the first luminescence probability model p i,doi t;l (θ l ;t) is the corrected first luminescence probability model p i,doi t;l;corrected (θ l It means ;t).

[0147] In step S460, the same processing as in step S450 is performed, and the processing circuit 150 uses the correction function 150e to determine the second timing Δt estimated in step S400. p m,i Based on this, the detection time correction is performed for the event detected by detector m that is associated with the i-th simultaneous counting event. That is, the processing circuit 150 corrects the detection timing t of the event detected by the first detector l using the measurement function 150d. m,i d Measure the second timing Δt based on the correction function 150e. p m,i Based on this, detection timing t m,i d Correct the value, and correct the detection time t m,i corrected To obtain.

[0148] Furthermore, the processing circuit 150, in conjunction with this, uses the correction function 150e to calculate a second luminescence probability model p with the detection time corrected. i,doi t;m;corrected (θ m ; t) is calculated. In the processing from step S500 onward, the corrected second emission probability model p i,doi t;m;corrected (θ m ;t) is simply the first luminescence probability model p i,doi t;m (θ m ;t) is written, but in the process of calculating the probability distribution model for coincidence from step S500 onwards, the second luminescence probability model p i,doi t;m (θ m ;t) is the corrected second luminescence probability model p i,doi t;m;corrected (θ m It means ;t).

[0149] Next, in step S500, the processing circuit 150 uses the calculation function 150b to calculate the first luminescence probability model p calculated in step S300. t;l doi(t1) and the second luminescence probability model p calculated t;m doi Based on (t2), a probability distribution model p for coincidence is determined for each identical count event, based on the first and second events. i、(doi) ct;l,m (θ l , θ m Identify t).

[0150] In Figure 10, curve 53 represents the first emission probability model p defined for each coincidence count event in the scintillator 51 of the first detector l. i,doi t;l (θ l ,t l ) with a delay time t l The graph shows the result plotted as a function of . Curve 54 also shows the second emission probability model p in the scintillator 52 of the second detector m. i,doi t;m (θ m ,t m ) with a delay time t m The plot shown is a function of . The processing circuit 150 calculates the detection time t at the detector l using the calculation function 150b. l The event at and the detection time t at detector m. m The product of events in this context is expressed as a function of the detection time difference, and a probability distribution model p defined for each simultaneous counting event is used. i,doi ct;l,m (θ l , θ m We identify t). Curve 56 shows such a probability distribution model p i,doi ct;l,m (θ l , θ m This shows an example of ,t).

[0151] Here, the detection time t at the detector l during the time interval dt. l The event at and the detection time t at detector m. m The number of intersection events in this case is given by the following equation (32).

[0152]

number

[0153] Here, θ i l This shows the parameter set obtained by detector l for the i-th simultaneous counting event. Furthermore, the parameter θ can be expressed in a more specific form as follows: equation (33).

[0154]

number

[0155] Here, N i s;l , N i s;m This is the number of scintillation photons in detectors l and m related to the i-th simultaneous counting event, and N i c;l , N i c;m This indicates the number of Cherenkov photons in detectors l and m related to the i-th simultaneous counting event.

[0156] Furthermore, the detection time t at detector l l And, the detection time t at detector m m The difference in detection time between the two is given by the following equation (34).

[0157]

number

[0158] Therefore, p i,doi ct;l,m (θ l , θ m ,t) is given by the following equation (35).

[0159]

number

[0160] Furthermore, the parameter θ can be expressed in a more concrete form as follows: (36)

[0161]

number

[0162] Returning to Figure 2, in step S30, the processing circuit 150 calculates a probability distribution model p relating to the detection time difference of LOR defined by the first detector l and the second detector m using the calculation function 150b. i、doi ct;l,m (θ i l , θ i m The t) is calculated for the i-th simultaneous counting event. The processing circuit 150 performs this calculation for all simultaneous counting events and adds them together to obtain the final probability distribution model p relating to the detection time difference of LOR defined by the first detector l and the second detector m. doi ct;l,m (θ i l , θ i m It is possible to calculate (t) and then reconstruct the data based on that.

[0163] In step S40, the processing circuit 150 uses the calculation function 150b to determine whether the processing in step S30 has been completed for all simultaneous counting events. If the processing circuit 150 determines, using the calculation function 150b, that the processing in step S30 has not been completed for all simultaneous counting events (step S40 No), the process proceeds to step S50, where processing is performed for the (i+1)th data point. On the other hand, if the processing circuit 150 determines, using the calculation function 150b, that the processing in step S30 has been completed for all simultaneous counting events (step S40 Yes), the process proceeds to step S60.

[0164] In step S60, the processing circuit 150 uses the calculation function 150b to calculate the probability distribution model p for each simultaneous counting event calculated in step S30.i、doi ct;l,m (θ i l , θ i m Adding t) gives the probability distribution model p for coincidence in the first detector l and the second detector m. tot ct;l、m Generates.

[0165] Specifically, the probability distribution model p relating to coincidence in the first detector l and the second detector m. tot ct;l、m This is expressed by the following equation (37).

[0166]

number

[0167] Next, in step S70, the processing circuit 150 uses the reconstruction function 150f to create a probability distribution model p relating to coincidence, which is defined based on the first and second photon number information identified in step S60. tot ct;l、m By using this as a TOF kernel, the data corresponding to the LOR between the first detector l and the second detector m is reconstructed.

[0168] Figures 11 to 14 show the verification results of the method according to the embodiment. In this verification, the first timing Δt p l,i We varied the value of the threshold p used to calculate the peak of the radioactive material distribution in the reconstructed image to see if it became sharper.

[0169] Figure 11 shows the first luminescence probability model p for three events, events 1, 2, and 3. i,doi t;l (θ l ;t) is shown. For these three events, the threshold p on the right side of equation (28) is changed to determine the first timing Δt for each event. pThe following was calculated. Line 60 corresponds to p=0.1, line 61 corresponds to p=0.4, and line 62 corresponds to p=0.8. Also, arrow 63 represents the first timing Δt for the first event when p=0.4. p_1 Arrow 64 indicates the first timing Δt for the second event when p=0.4. p_2 Arrow 65 indicates the first timing Δt for the third event when p=0.4. p_3 This indicates that.

[0170] Figure 12 shows the corrected coincidence spectra for p=0.4 and p=0.1. Graph 71 shows the corrected coincidence spectrum when p=0.4, and Graph 70 shows the corrected coincidence spectrum when p=0.1. Furthermore, comparing the counts around a delay time of 0, the counts are high but the width is also large at p=0.1, while at p=0.4, both the counts are high and the width is small. Therefore, when the threshold is set to around p=0.4, the counts in the coincidence spectrum become high and the width becomes small.

[0171] Furthermore, we verified whether the peak of the radioactive material distribution in the reconstructed image became sharper when correction was performed according to the embodiment and when it was not. First, as shown in Figure 13, a fitting curve 81 was generated for the measured data 80 and the spectrum of the ToF kernel was identified. Next, reconstruction was performed using the ML-EM method based on the obtained spectrum of the ToF kernel. Figure 14 shows the reconstruction results. Graph 90 shows the reconstruction results when correction was performed according to the embodiment, and graph 91 shows the reconstruction results when correction was not performed according to the embodiment. Comparing graph 90 and graph 91, it was confirmed that the count number near the peak was approximately doubled, and the peak of the radioactive material distribution became sharper.

[0172] As described above, in this embodiment, the delay time from when gamma rays enter the scintillator until emission occurs is estimated based on the emission model, and the ToF kernel is corrected based on the estimation result. This makes it possible to sharpen the ToF spectrum and improve the image quality of the reconstructed PET image.

[0173] With respect to the above embodiments, the following additional notes are disclosed as aspects of the invention and selective features.

[0174] (Note 1) A nuclear medicine diagnostic device provided in one aspect of the present invention comprises an acquisition unit, a calculation unit, a specification unit, a measurement unit, and a correction unit. The acquisition unit acquires first photon count information detected by a first detector. The calculation unit calculates a first emission probability model corresponding to the first detector based on the first photon count information. The specification unit identifies a first timing at which the detection probability is greater than or equal to a predetermined threshold, based on the first emission probability model. The measurement unit measures the detection timing of an event detected by the first detector. The correction unit corrects the detection timing based on the first timing.

[0175] (Note 2) The first photon number information may include the scintillation photon number.

[0176] (Note 3) The first photon number information may further include the Cherenkov photon number.

[0177] (Note 4) The identifying unit may identify the first timing based on the cumulative distribution of the distribution shown by the first luminescence probability model.

[0178] (Note 5) The identifying unit may identify the first timing based on the time at which the distribution shown by the first luminescence probability model takes an extreme value.

[0179] (Note 6) The first emission probability model may relate to the probability density of detecting the first photon among a plurality of photons included in the detection event.

[0180] (Note 7) The first emission probability model may relate to the probability density of detecting a predetermined number of photons from among a plurality of photons included in the detection event.

[0181] (Note 8) The calculation unit may generate a distribution function of light emission at each time point based on the first photon number information, and calculate the first light emission probability model based on the generated distribution function.

[0182] (Note 9) The acquisition unit may acquire the first photon count information for each detector.

[0183] (Note 10) The acquisition unit may acquire the first photon count information for each event.

[0184] (Note 11) The calculation unit may calculate the first emission probability model taking into account the uncertainty of the emission position within the scintillator.

[0185] (Note 12) A data processing method provided in one aspect of the present invention involves acquiring first photon count information detected by a first detector, calculating a first emission probability model corresponding to the first detector based on the first photon count information, identifying a first timing at which the detection probability is greater than or equal to a predetermined threshold based on the first emission probability model, measuring the detection timing of an event detected by the first detector, and correcting the detection timing based on the first timing.

[0186] (Note 13) A program provided in one aspect of the present invention acquires first photon count information detected by a first detector, calculates a first emission probability model corresponding to the first detector based on the first photon count information, identifies a first timing at which the detection probability is greater than or equal to a predetermined threshold based on the first emission probability model, measures the detection timing of an event detected by the first detector, and causes a computer to perform a process to correct the detection timing based on the first timing.

[0187] According to at least one embodiment described above, image quality can be improved.

[0188] While several embodiments have been described, these embodiments are presented as examples only and are not intended to limit the scope of the invention. These embodiments can be implemented in a variety of other forms, and various omissions, substitutions, modifications, and combinations of embodiments are possible without departing from the spirit of the invention. These embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims and their equivalents. [Explanation of symbols]

[0189] 150a acquisition function 150b Calculation Function 150c specific functions 150d measurement function 150e Correction function 150f reconfiguration function 150g control function 150h reception function 150i image generation function 150j Display Control Function

Claims

1. An acquisition unit that acquires first photon number information detected by the first detector, A calculation unit that calculates a first emission probability model corresponding to the first detector based on the first photon number information, A specification unit that identifies a first timing in which the detection probability is greater than or equal to a predetermined threshold, based on the first emission probability model, A measuring unit that measures the detection timing of an event detected by the first detector, A correction unit corrects the detection timing based on the first timing, A nuclear medicine diagnostic device equipped with the following features.

2. The nuclear medicine diagnostic apparatus according to claim 1, wherein the first photon number information includes the scintillation photon number.

3. The nuclear medicine diagnostic apparatus according to claim 2, wherein the first photon number information further includes the Cherenkov photon number.

4. The nuclear medicine diagnostic apparatus according to claim 1, wherein the identifying unit identifies the first timing based on the cumulative distribution of the distribution shown by the first luminescence probability model.

5. The nuclear medicine diagnostic apparatus according to claim 1, wherein the identifying unit identifies the first timing based on the time at which the distribution shown by the first luminescence probability model takes an extreme value.

6. The nuclear medicine diagnostic apparatus according to claim 1, wherein the first emission probability model relates to the probability density for detecting the first photon among a plurality of photons included in a detected event.

7. The nuclear medicine diagnostic apparatus according to claim 1, wherein the first emission probability model relates to the probability density for detecting a predetermined number of photons among a plurality of photons included in a detected event.

8. The nuclear medicine diagnostic apparatus according to claim 6, wherein the calculation unit generates a distribution function of light emission at each time based on the first photon number information, and calculates the first light emission probability model based on the generated distribution function.

9. The acquisition unit acquires the first photon number information for each detector, as described in claim 1.

10. The nuclear medicine diagnostic apparatus according to claim 1, wherein the acquisition unit acquires the first photon number information for each event.

11. The nuclear medicine diagnostic apparatus according to claim 1, wherein the calculation unit calculates the first luminescence probability model taking into account the uncertainty of the luminescence position within the scintillator.

12. The first photon number information detected by the first detector is acquired, Based on the first photon number information, a first emission probability model corresponding to the first detector is calculated. Based on the first emission probability model, a first timing is identified in which the detection probability exceeds a predetermined threshold. The detection timing of the event detected by the first detector is measured, A data processing method for correcting the detection timing based on the first timing.

13. The first photon number information detected by the first detector is acquired, Based on the first photon number information, a first emission probability model corresponding to the first detector is calculated. Based on the first emission probability model, a first timing is identified in which the detection probability exceeds a predetermined threshold. The detection timing of the event detected by the first detector is measured, A program that causes a computer to perform a process to correct the detection timing based on the first timing.

Images