JP2008190901A - Positron emission tomography system - Google Patents

Abstract

An object of the present invention is to provide a positron emission tomography apparatus with high quantification by a counting rate correction method not using a phantom.
A positron emission tomography apparatus includes a plurality of detectors that output γ-ray detection signals in response to detection of γ-rays, and detection time information and γ-ray detection based on the γ-ray detection signals. A plurality of unit boards 20 for generating detected γ-ray information including a detector ID for identifying the detector and a detected γ-ray energy value, and simultaneous measurement processing in the simultaneous measuring unit 12b based on the detected γ-ray information, and image reconstruction The data processing device 12A that generates a tomographic image in the unit 12h is provided. In the simultaneous measurement process, the data processing device 12A includes a single event measurement unit 12d that acquires a count rate of an input single event, and a correction count calculation unit 12e that calculates a correction coefficient based on the acquired count rate of the single event. And tomographic images are corrected based on the calculated correction count.
[Selection] Figure 2

Description

  The present invention relates to a nuclear medicine diagnostic apparatus, and more particularly to a positron emission computed tomography apparatus (hereinafter referred to as a PET apparatus) which is a kind of nuclear medicine diagnostic apparatus using a radiation detector.

The inspection technique using radiation can inspect the inside of a subject nondestructively. In particular, radiation inspection techniques for the human body include X-ray CT, PET, SPECT single photon emission computed tomography (hereinafter referred to as “SPECT”), and the like.
Each of these techniques is a technique for measuring a physical quantity to be examined as an integral value in the radiation flight direction, and calculating and imaging the physical quantity of each voxel in the subject by back projecting the integral value. With these technologies, it is necessary to process enormous amounts of data, and with the rapid development of computer technology in recent years, the processing speed has been increased and high-detail images have been provided.

PET and SPECT are techniques capable of detecting functions and metabolism at a molecular biology level that cannot be detected by X-ray CT or the like, and can provide a functional image of the body. PET is a technique in which a radiopharmaceutical labeled with positron emitting nuclides such as ¹⁸ F, ¹⁵ O, and ¹¹ C is administered, and its distribution is measured and imaged. Examples of drugs include fluoro-deoxy-glucose (2- [F-18] fluoro-2-deoxy-D-glucose, ¹⁸ FDG), which indicates that drugs are highly accumulated in tumor tissues due to glucose metabolism. It is used to identify the tumor site.
Radionuclides taken into the body decay and release positrons (β +). When the emitted positron is combined with an electron and annihilated, it emits a pair of annihilation γ-rays (annihilation γ-ray pairs) each having energy of 511 keV. Since this annihilation γ-ray pair is emitted in almost opposite directions (180 ° ± 0.6 °), the annihilation γ-ray pair is simultaneously detected by a detection element arranged so as to surround the subject, and the radiation direction thereof Projection data (sinogram data) can be obtained by accumulating data. Projection data is back-projected using a filtered back projection method (see Non-Patent Document 1) and the like, and a radiation position (an accumulation position of radionuclides) can be identified and imaged.

In the radiation imaging apparatus such as the conventional PET apparatus described above, a scintillator is used as a γ-ray detector in order to obtain an image. The scintillator performs a process of once converting incident γ-rays into visible light and then converting it back to an electric signal by a photomultiplier tube (photomultiplier). The scintillator has a problem that the number of photons generated at the time of visible light conversion is small and the two-step conversion process is required as described above, so that the energy resolution is low and a high-precision diagnosis cannot always be performed. Was. The decrease in energy resolution is a cause that cannot be quantitatively evaluated especially when 3D imaging is performed by a PET apparatus. This is because, since the energy resolution is low, the energy threshold of γ-rays has to be lowered, and a lot of internal scattering, which is noise that increases during 3D imaging, is detected.
Further, since the detection position is specified by amplifying signals of a plurality of scintillators with one high electron multiplier and performing the centroid calculation, there is a problem that an error occurs in the detection position.

  Therefore, in recent years, the use of a semiconductor radiation detector as a detector for a radiation imaging apparatus has attracted attention. The semiconductor radiation detector directly converts incident γ-rays into electrical signals, and has a feature of high energy resolution due to the large number of electron-hole pairs generated. Further, since the detection of γ-rays is performed independently by each semiconductor detector, it has a feature that the position resolution is good. Furthermore, the semiconductor detector can be miniaturized and one resolution can be improved.

When a scintillator is used as the γ-ray detector, optical signals from a plurality of scintillator crystals are processed by a photomultiplier tube capable of position discrimination. Therefore, the maximum count rate that can be measured per channel (one scintillator crystal) is determined from the maximum count rate that can be processed by the photomultiplier tube. When the number of scintillator crystals processed by one photomultiplier tube increases, there is a problem that the maximum count rate per channel decreases.
On the other hand, in semiconductor γ-ray detectors, signals from individual detectors can be read out by individual circuit systems, and the maximum coefficient rate per channel (one semiconductor γ-ray detector) is improved. Data loss due to system dead time is reduced, and imaging in a high count rate state becomes possible.
IEEE Transactions on Nuclear Science NS-21 Volume 21 P.21

By the way, in a PET apparatus, when performing a diagnostic process etc. from the output image, the quantitative property of the image is very important. That is, it is necessary to obtain a count rate that provides an image density distribution proportional to the concentration of the radionuclide distributed in the body of the subject.
In a PET apparatus using a scintillator, when the counting rate increases, data loss increases due to dead time in a signal processing system such as a photomultiplier tube as described above, and the quantitativeness cannot be maintained. On the other hand, when a semiconductor γ-ray detector is used, there is little data loss and excellent quantitativeness.

In general, dead time correction is performed to ensure quantitativeness. This dead time correction is a method in which a correction table corresponding to the radioactivity concentration of a subject is prepared in advance, and measurement data is executed from the correction table during imaging.
Specifically, the correction amount (correction table) for the single event count rate (single event rate) before simultaneous measurement is obtained in advance by a phantom test, etc., the single event rate is measured during imaging, and the correction table is referenced. The correction process is performed.

In the correction method, since a correction table is created by a phantom test, time has elapsed since the phantom test during actual imaging, and there is a problem that the correction amount becomes inaccurate due to a change in characteristics of the electronic circuit over time. .
In addition, the phantom used to create the correction table simulates a standard body shape, and there is a problem that the correction becomes inaccurate for a subject that is greatly different from the standard body shape.

  An object of the present invention is to provide a positron emission tomography apparatus with high quantitativeness by a counting rate correction method that does not use a phantom.

  A plurality of γ-ray detectors that output γ-ray detection signals in response to detection of γ-rays, detection time information, detector IDs for identifying γ-ray detectors, and detected γ-ray energy values based on the γ-ray detection signals Data for generating a tomographic image in the image generation unit based on the result of the simultaneous measurement process and a plurality of data generation units that generate the detected γ-ray information and the simultaneous measurement unit based on the detected γ-ray information In the positron emission tomography apparatus comprising the processing device, the data processing device acquires the single event rate acquisition means for acquiring the count rate of the input single event in the simultaneous measurement processing, and the single event rate acquisition means Correction coefficient calculation means for calculating a correction coefficient based on the calculated single event count rate, and a tomographic image based on the calculated correction coefficient It is characterized by correcting.

  According to the present invention, it is possible to provide a positron emission tomography apparatus with high quantitativeness by a counting rate correction method that does not use a phantom.

<< First Embodiment >>
Next, a PET apparatus which is a preferred embodiment of the present invention will be described with reference to the drawings as appropriate.
In the following, description of elements applied to this embodiment such as arrangement (layout) of each element such as nuclear medicine diagnosis apparatus and analog ASIC of this embodiment on the substrate, unitization of the substrate, and the counting rate correction processing The method will be explained.
The analog ASIC means an application specific integrated circuit (ASIC) that is an application specific IC that processes an analog signal, and is a kind of LSI (Large Scale Integrated Circuit).

(PET device)
First, the PET apparatus of this embodiment will be described.
As shown in FIG. 1, the PET apparatus 1A includes a camera 11, a data processing apparatus 12A, an operation console 13, and the like. The subject P (see FIG. 2) is placed on the bed 14 and photographed by the camera 11.
The operation console 13 includes a display device 13a that displays a tomographic image of the PET apparatus 1A, a state check result of the PET apparatus 1A, and the like, and an input operation unit 13b such as a keyboard and a mouse.

As shown in FIG. 2, a semiconductor γ-ray detector (γ-ray detector, hereinafter simply referred to as a detector) 21 is provided inside the camera 11 to detect γ-rays emitted from the subject P (see FIG. 2). A plurality of detector units 2 each including a plurality of unit substrates 20 (see FIG. 4 for details) are arranged in the circumferential direction, and the detector 21 is released from the body of the subject P. Γ rays are detected.
The detector unit 2 has an integrated circuit (ASIC) for measuring the detection energy and detection time of γ-rays on the unit substrate 20 and measures the detection energy and detection time of the detected γ-rays. The address of the detector 21 that detected the γ-ray is detected, and these are output to the data processing device 12A as detected γ-ray information.

  The detector unit 2 has a detector ID corresponding to the detected energy data (detected energy value information) of the detected γ-rays, detection time data (detection time information), and the address of the detector 21 (see FIG. 4). Packet data (detected γ-ray information) including (detector address information) is output to the data processing device 12A via the auxiliary data collection unit 4. The auxiliary data collection unit 4 transfers the packet data from the detector unit 2 to the data sorting unit 12a of the data processing device 12A.

As shown in FIG. 2, the data processing device 12A has a storage device (not shown) and takes in packet data. The data processing device 12A includes a data sorting unit 12a, a simultaneous measurement unit 12b, a sinogram data storage unit 12c, a single event measurement unit 12d, a correction count calculation unit 12e, a sinogram data correction unit 12f, a transmission data storage unit 12g, and an image reconstruction unit. 12h.
Details of the data processing device 12A will be described later.

Incidentally, the subject P is administered with a radiopharmaceutical, for example, fluorodioxyglucose (FDG) containing ¹⁸ F of a half-life of 110/110, and from the body of the subject P, the positron released from the FDG At the time of annihilation, a pair of 511 keV gamma rays (annihilation gamma rays) is emitted in a direction of approximately 180 ° with respect to each other.
At this time, each detector 21 (see FIG. 4) of the camera 11 surrounds the bed 14. From the detector unit 2 to the data processor 12A via the auxiliary data collection unit 4, the detected energy value information and the detection time obtained based on the γ-ray detection signal when the detector 21 interacts with the γ-ray. Information and a detector ID are output for each detector 21 included in the detector unit 2.

  As shown in FIG. 3A, the detector unit 2 has a configuration in which 60 to 70 pieces are detachably disposed in the opening 11b of the camera 11 in the circumferential direction, so that maintenance inspection is easy. Has been. The detector unit 2 is mounted via a unit support member 2A. Further, as shown in FIG. 3B, the detector unit 2 is mounted on the camera 11 with one end supported by the unit support member 2A. The unit support member 2A has a hollow disk shape (doughnut shape), and includes a large number of openings 11b in which the detector unit 2 is mounted in the circumferential direction of the camera 11 (the number of detector units 2 to be mounted). Yes. Thus, in order to support the detector unit 2 at one end, a flange portion serving as a stopper is provided on the front side of the housing 30 of the detector unit 2 in the body axis direction.

  Incidentally, when trying to arrange the detector units 2 as closely as possible in the circumferential direction, the flange portion on the inner side in the circumferential direction becomes an obstacle. In view of this, the flange portion, which is a hindrance, may be eliminated from the housing 30 to leave the outer circumferential flange portion. Another unit support member 2A may be installed in the body axis direction, and both ends of the detector unit 2 may be held by both unit support members 2A.

When the detector unit 2 is attached to the camera 11, the lid 11a is removed to expose the unit support member 2A, and the detector unit 2 is inserted and attached from the opening 11b until the flange portion abuts. ing. By inserting and mounting, the respective connectors of the power supply and signal wiring (not shown) of the camera 11 and the corresponding connectors of the detector unit 2 are connected, and the connection of the signal and power supply between the camera 11 and the detector unit 2 is performed. Is made.
The configurations of the detector 21, the unit substrate 20, and the detector unit 2 will be described in detail later.
Hereinafter, the characteristic part of this embodiment will be described.

(Semiconductor radiation detector)
First, the detector 21 applied to the present embodiment has, for example, a laminated structure as shown in FIG. 4 of Japanese Patent Laid-Open No. 2005-106805, and has both sides of a semiconductor radiation detection element made of a plate-like semiconductor material. It has a structure in which what is covered with thin plate (film-like) electrodes (anode, cathode) are stacked. The semiconductor material is obtained by slicing a single crystal of CdTe (cadmium telluride), for example. The anode and cathode are made of any material such as Pt (platinum), Au (gold), and In (indium).
In the following description, it is assumed that the semiconductor material S is a slice of CdTe single crystal. The detected radiation is assumed to be 511 keV γ rays by ¹⁸ F used in the PET apparatus 1A.

(Unit board)
Next, the detailed structure of the unit board | substrate 20 installed in the detector unit 2 is demonstrated using FIG. FIG. 4A is a front view showing the unit substrate, and FIG. 4B is a side view of FIG. 4A.
The unit board 20 includes a detector board area 20a in which a plurality of detectors 21 are installed on both sides of the unit board body 20c, a capacitor 22, a resistor 23, and an analog ASIC (data generation unit) 24 on both sides of the unit board body 20c. And an ASIC substrate area 20b in which an analog / digital converter (hereinafter referred to as ADC) 25 is installed and a substrate integrated FPGA (Field Programmable Gate Array) (data generation unit) 26 is installed on one side.

(Detector board area)
As shown in FIG. 4A, the detector substrate region 20a has a lateral direction in FIG. 4A corresponding to the circumferential direction around the body axis of the subject P on one side of the unit substrate body 20c. For example, 16 detectors 21 are arranged in one row, and further, four rows are arranged in the vertical direction in FIG. 4A corresponding to the radial direction with respect to the body axis of the subject P. A total of 64 detectors 21 (4 × 4) are arranged in a grid pattern. Further, as shown in FIG. 4B, detectors 21 are similarly installed on the other surface of the detector substrate region 20a, and 128 detectors in total on both sides are provided in one detector substrate region 20a. A vessel 21 is provided.
Here, as the number of detectors 21 increases, it becomes easier to detect γ-rays, and the positional accuracy at the time of γ-ray detection can be increased, so that the detector 21 is as dense as possible on the detector substrate region 20a. It is arranged.

  Incidentally, in FIG. 4A, detection is performed when γ-rays emitted from the subject P on the bed 14 travel from the lower side to the upper side (the direction of the arrow X, that is, the radial direction of the camera 11). It is preferable to arrange the detectors 21 in the left-right direction in the instrument substrate region 20a densely because the number of γ rays that pass through (the number of γ rays that pass through the gap between the detectors 21) can be reduced. Thereby, the detection efficiency of γ-rays is increased, and the spatial resolution in the body axis direction of the obtained image can be increased.

As shown in FIG. 4B, the detector substrate region 20a of the present embodiment has the detectors 21 installed on both sides of the unit substrate body 20c. By mounting on both sides, the unit substrate body 20c can be shared. For this reason, the number of unit substrate bodies 20c can be halved, and the detectors 21 can be arranged more densely in the body axis direction of the subject P. In addition, as described above, the number of unit substrate bodies 20c can be reduced to half.
In the above description, the 16 horizontal detectors 21 are arranged in the camera 11 in the circumferential direction around the body axis of the subject. However, the present invention is not limited to this. For example, 16 horizontal detectors 21 may be arranged in the body axis direction of the subject P in the camera 11.

  A predetermined potential difference (voltage) is applied between the anode and the cathode of each detector 21 by, for example, a high voltage power source (not shown) for collecting charges. This voltage is supplied from the ASIC substrate region 20b side to the detector substrate region 20a side. Further, the γ-ray detection signal output when each detector 21 detects γ-rays is supplied to the ASIC substrate region 20b side. For this reason, in the unit substrate main body 20c of the detector substrate region 20a, an in-board wiring (for detector applied voltage and signal exchange) (not shown) is provided. This intra-substrate wiring has a multilayer structure.

(ASIC substrate area)
Next, the ASIC substrate region 20b on which the ASIC is mounted will be described with reference to FIG. As shown in FIG. 4A, in the ASIC board region 20b, two analog ASICs 24 are installed on both sides of the unit board body 20c, and one board integrated FPGA 26 is installed on one side. That is, one ASIC board region 20 b has a total of four analog ASICs 24 and one board integrated FPGA 26.
In addition, the ASIC substrate region 20b has 32 ADCs 25 in total, 16 on each side of the unit substrate body 20c. Further, capacitors 22 and resistors 23 corresponding to the number of detectors 21 are installed on both surfaces of one unit substrate body 20c. Further, in order to electrically connect the capacitor 22, the resistor 23, the analog ASIC 24, the ADC 25, and the board integrated FPGA 26, the unit board main body 20c in the ASIC board area 20b is similar to the detector board area 20a described above. In-substrate wiring (not shown) is provided. This intra-substrate wiring also has a laminated structure.

The arrangement of these circuit elements 22, 23, 24, 25, and 26 and the wiring in the board are such that the signal supplied from the detector board region 20 a is obtained from the capacitor 22, the resistor 23, the analog ASIC 24, the ADC 25, They are supplied in order.
The ASIC board region 20b has a board connector C2 for electrical connection with the unit integrated FPGA side described later.
Incidentally, the detector substrate region 20a and the ASIC substrate region 20b are configured on the surface of one common unit substrate body 20c, as shown in FIG. 4B.

Next, the configuration and function of each circuit element in the ASIC substrate area will be described with reference to FIGS.
FIG. 5 is a block diagram of the detector unit showing the schematic configuration of each of the analog ASIC and the board integrated FPGA and the connection relationship between the analog ASIC and the board integrated FPGA.
(Analog ASIC)
FIG. 6 is a block diagram schematically showing the functional configuration of the analog ASIC.
As shown in FIG. 6, one analog ASIC 24 includes an analog having a charge-sensitive preamplifier (hereinafter referred to as a preamplifier) 24a, and a fast system 24A and a slow system 24B connected thereto. For example, 32 sets of signal processing circuits 33 are provided. One analog ASIC 24 is obtained by converting 32 sets of analog signal processing circuits 33 into an LSI.

  The analog signal processing circuit 33 is provided for each detector 21, and one analog signal processing circuit 33 is connected to one detector 21. Here, the first system 24A includes a timing pick-off circuit 24b that outputs a timing signal for specifying the detection time of the γ-ray based on the γ-ray detection signal output from the preamplifier 24a. The timing pick-off circuit 24b generates a timing signal using a technique such as CFD or leading edge trigger. The slow system 24B has a predetermined time constant for the purpose of obtaining the detected energy (crest value) of the γ-ray based on the γ-ray detection signal, and a waveform shaping circuit 24d that shapes the signal waveform into a Gaussian shape. A peak hold circuit 24e is further connected to the waveform shaping circuit 24d. For example, a CR-RC filter or the like is used for the waveform shaping circuit 24d.

  The signal output from the detector 21 and passed through the capacitor 22 is amplified by the preamplifier 24a, output as a γ-ray detection signal, further amplified by the waveform shaping circuit 24d, and input to the peak hold circuit 24e. The peak hold circuit 24e holds the maximum value of the γ-ray detection signal whose waveform has been shaped as described later, that is, the peak value proportional to the detected energy value of the γ-ray.

Although the capacitor 22 and the resistor 23 can be provided inside the analog ASIC 24, the present embodiment is implemented in order to obtain an appropriate capacitor capacity and an appropriate resistance value and to reduce the size of the analog ASIC 24. In form, the capacitor 22 and resistor 23 are located outside the analog ASIC 24.
Incidentally, it is said that the capacitor 22 and the resistor 23 are provided outside, so that there is less variation in individual capacitor capacity and resistance value.

The output of the slow system 24B of the analog ASIC 24 is supplied to the ADC 25 (see FIG. 6). Further, the output of the first system 24A of the analog ASIC 24 and the output of the ADC 25 are supplied to the board integrated FPGA 26.
Here, each analog ASIC 24 and the board integrated FPGA 26 are connected by 32 wires that transmit the signals of the fast system 24A of 32 channels one by one (see FIG. 5). Further, the analog ASIC 24 and each ADC 25, and each ADC 25 and the board integrated FPGA 26 are respectively connected by one wiring that collectively transmits the signals of the slow system 24B for the eight channels of the detector 21. (See FIG. 5).
Further, from the board integrated FPGA 26, one ADC control signal line for controlling one ADC 25 corresponding to the analog signal processing circuit 33 for eight channels, and one analog signal processing of each analog ASIC 24. A peak hold control signal line for 8 channels for individually controlling the peak hold circuit 24e of the circuit 33 is provided, and is connected to the ADC 25 and the analog signal processing circuit 33.
The above-described peak hold control signal lines may be wired from the board integrated FPGA 26 to each analog signal processing circuit 33 one by one without consolidating the eight channels into one.

(Board integrated FPGA)
Next, the board integrated FPGA 26 will be described with reference to FIG.
As shown in FIG. 5, the board integrated FPGA 26 includes 16 sets of detection signal processing units 34 including eight timing detection units 35 and one detector control unit 36, and one data transfer unit 37. These are LSIs.

All of the substrate integrated FPGAs 26 provided in the PET apparatus 1A operate in synchronization with receiving a clock signal from a clock generation circuit (crystal oscillator) of 500 MHz (not shown), for example. The clock signal input to each board integrated FPGA 26 is input to each timing detection unit 35 in all detection signal processing units 34.
The timing detector 35 is provided for each detector 21 and receives a timing signal from the timing pick-off circuit 24b of the corresponding analog signal processing circuit 33. The timing detector 35 determines the detection time of γ rays based on the clock signal when the timing signal is input. Since the timing signal is based on the signal of the first system 24A of the analog ASIC 24, a time close to the true detection time can be set as the detection time (detection time information).

The detector control unit 36 has an address calculation function, an ADC / peak hold control function, a data correction function, and a packet data generation function.
(Address calculation function)
The address calculation function receives detection time information corresponding to the timing signal from which the γ-ray is detected from the timing detection unit 35, and identifies the detector ID. That is, the detector control unit 36 stores a detector ID for each timing detection unit 35 to be connected. When detection time information is input from a certain timing detection unit 35, the detector control unit 36 corresponds to the timing detection unit 35. The address calculation function is performed by specifying the detector ID. This is possible because the timing detector 35 is provided for each detector 21 and is connected to the detector controller 36.

(ADC / peak hold control function)
Furthermore, after the time information is input, the detector control unit 36 outputs a peak hold control signal to the analog signal processing circuit 33 for eight channels including the specified detector ID, and the detector ID and The ADC control signal is output to the ADC 25.

  Upon receiving the peak hold control signal, the peak hold circuit 24e of the analog signal processing circuit 33 performs peak hold processing on the signal input from the waveform shaping circuit 24d. Then, after a predetermined time, the reset signal is received from the detector control unit 36, and the peak hold process is canceled. The ADC 25 converts a peak value (voltage value) output from the peak hold circuit 24e of the analog signal processing circuit 33 corresponding to the detector ID input from the detector control unit 36 into a digital signal, and detects the detector control unit. To 36.

(Data correction function)
The detector control unit 36 corrects the detection time information input from the timing detection unit 35 by, for example, a method described in Japanese Patent Application Laid-Open No. 2005-249806 based on the peak value acquired via the ADC 25. . Further, the gain and offset correction of the detector 21 and the analog ASIC 24 are performed on the peak value input from the ADC 25 based on calibration data collected in advance.

(Packet data generation function)
The detector controller 36 uses the corrected peak value as detected energy value information, adds the corrected detection time information and detector ID to the detected energy value information, and adds packet data (detected γ-ray information) as digital information. ) And output to the data transfer unit 37.
Since the energy of γ rays emitted from the γ-ray source for transmission imaging is different from 511 keV of the γ-ray pair for emission imaging, the γ-ray source for transmission imaging is used when data processing is performed in the board integrated FPGA 26. The predetermined energy detection γ-ray information peculiar to is determined as transmission data, and a determination flag indicating transmission data is added to generate packet data.

(Data transfer part)
The data transfer unit 37 is an input / output integration function that combines a plurality of input systems into one output system, and a buffer memory function that temporarily stores input packet data and outputs it according to the processing speed of the subsequent components And have. When the data transfer unit 37 receives packet data from each detector control unit 36, the data transfer unit 37 buffers the packet data as necessary, and periodically outputs 15 pieces of packet data output from each detection signal processing unit 34, for example. Is transmitted to the unit integrated FPGA 31 provided in one case 30 of the detector unit 2 (see FIG. 7) containing the unit substrate 20.

(Unit integrated FPGA)
The unit integrated FPGA 31 includes a buffer function that once receives packet data from the digital ASICs of all unit substrates 20 accommodated in the detector unit 2, a sort processing function that arranges the packet data in time order based on the detection time information, And a function of performing scattered radiation processing.

A detailed description of the sort processing function and the scattered radiation processing function will be given later.
The unit integrated FPGA 31 transmits the packet data to the auxiliary data collection unit 4 via the information transmission wiring connected to the connector 38. For example, four auxiliary data collection units 4 (see FIG. 2) are arranged in the circumferential direction, and packet data from the detector units 2 that are divided into four groups in the circumferential direction are collected into one auxiliary data collection unit 4. Are collected and transmitted to the data processing device 12A.

(Detector unit; unitized by storing the unit board)
Next, unitization by storing the unit substrate 20 in the housing 30 will be described.
7A is a cross-sectional view of the housing in a plane perpendicular to the body axis of the subject P, and FIG. 7B is a YY cross-sectional view in FIG.
As shown in FIG. 7, the detector unit 2 performs, for example, 15 unit boards 20, the 15 unit boards 20, an integrated board 32 on which the unit integrated FPGA 31 is mounted, and exchanges signals with the outside. A housing 30 is provided for housing and holding a signal connector, a power connector for receiving power supply from the outside, and the like.

As shown in FIG. 7, for example, the unit substrate 20 is removed from the top plate 30 a of the housing 30, and 15 sheets are stacked in the depth direction (the body axis direction of the subject P) and arranged in the housing 30 from the top to the bottom. It can be inserted toward That is, 15 unit substrates 20 are accommodated in one housing 30, and one integrated substrate 32 is accommodated in the back thereof. In order to store the unit substrate 20 and the integrated substrate 32 in the housing 30 as described above, as shown in FIG. 7A, a pair of left and right guide members 39 extending in the longitudinal direction on the paper surface are provided on the subject P. 16 rows are attached to the inner surface of the side plate 30c of the housing 30 with appropriate separation in the body axis direction. The guide member 39 has a guide groove (not shown).
In the present embodiment, as shown in FIG. 7A, the 16 detectors 21 in the horizontal direction on the paper surface are arranged in the circumferential direction of the camera 11, but the 16 detectors in the horizontal direction are detected. The container 21 may be stored in the housing 30 so as to be arranged in the body axis direction of the subject P.

  Further, as shown in FIG. 7A, four position holding members 41b extending in the depth direction on the paper surface (left and right direction on the paper surface in FIG. 7B) are provided on the unit substrate body 20c. The four position holding holes 41a and four position holding holes (not shown) provided in the integrated substrate 32 are inserted, and both ends of the position holding member 41b are provided at corresponding positions on the front and back side surfaces of the housing 30. It is held by a holding hole (not shown). The position holding member 41b is, for example, male threaded on one end side, here the back side, and fastened and fixed to the side surface on the back side of the housing 30 with a nut 41c. Thus, the four position holding members 41b prevent the unit substrate 20 and the integrated substrate 32 from being displaced in the vertical direction on the paper surface of FIG. 7 in the guide groove of the guide member 39 described above.

The upper part and the lower part of the casing 30 are when the casing 30 is taken out from the camera 11, and when the casing 30 is provided in the camera 11, as shown in FIG. May be reversed, and the top and bottom may be rotated 90 ° to the left or right, or may be slanted. Therefore, after the unit substrate 20 and the integrated substrate 32 are inserted into the guide member 39 of the housing 30, they are fixed by the position holding member 41 b so as not to slide and move within the guide groove of the guide member 39.
After the predetermined number of unit boards 20 are arranged in the housing 30, the integrated board 32 and each unit board 20 are connected via the board connector C2 with cables 42 including signal lines and high-voltage power supply lines. In addition, the applied voltage for the detector 21 can be supplied from the integrated substrate 32 side to the unit substrate 20, and the detected γ-ray information can be transmitted from the unit substrate 20 to the integrated substrate 32.

  Thereafter, the top plate 30a is detachably attached to the upper end of the housing 30 with screws or the like. At this time, an opening is provided in the top plate 30a so that the connector receiving portion of the integrated board 32 is exposed on the top surface of the top plate 30a, and a cable including a communication line and a high-voltage power supply line on the camera 11 side by the connector 38. 43.

  Since the detector 21 using CdTe as a semiconductor material used in the present embodiment generates charges in response to light, the casing 30 is made of a light-shielding material such as aluminum or an aluminum alloy. In addition, when the camera 11 is assembled as shown in FIG. 3 and the lid 11a is attached, a gap through which light enters the detector substrate region 20a together with the structure on the camera 11 side is eliminated.

  Further, packet data (all packet data for all detectors 21 of the unit substrate 20) output from the data transfer unit 37 of all the unit substrates 20 in the detection unit 2 is supplemented from the unit integrated FPGA 31 provided in the detection unit 2. The data is sent to the data processing device 12A via the data collection unit 4.

(Power supply)
Next, a high voltage power supply device that supplies a voltage for collecting charges will be described. Although not shown in FIG. 7 and FIG. 3B, a high voltage power supply device that supplies a voltage for collecting charges to the detector 21 is provided on the back side of the unit integrated FPGA 31 in a casing made of a conductive metal material. Install it in the box. This high-voltage power supply device is supplied with low-voltage power from the camera 11 side, boosts the voltage to 500 V by a DC-DC converter that boosts a voltage (not shown), and supplies it to each detector 21.
By the way, 64 detectors 21 are provided on one unit substrate and 128 detectors on both sides. Then, 15 unit substrates 20 are stored in one housing 30. Therefore, a voltage is supplied to 128 × 15 = 1920 detectors 21 from the high-voltage power supply device.

In this embodiment, the high voltage power supply device attached to the back side of the detector unit 2 is connected to an external low voltage (5 to 15 V) DC power supply by a voltage power supply line (not shown). The high voltage side terminal of the high voltage power supply device is connected to the connector C2 by the high voltage power supply line of the cable 42 via the connector 38 provided on the top plate 30a.
The high-voltage power supply is connected to each detector 21 provided on the unit substrate 20 via a substrate connector C2 of each unit substrate 20 and a high-voltage power supply line (not shown) in the unit substrate body 20c. The connector C2 includes a connector for a high-voltage power supply line in addition to the connector that transmits the output signal of the detector 21.

(Scattering processing function)
Next, the scattering processing function in the unit integrated FPGA 31 will be described with reference to FIG.
The unit integrated FPGA 31 has a function of performing scattered radiation processing, and includes a first data sorting unit 80, a first scattered radiation processing unit 81, a second data sorting unit 82, a second scattered radiation processing unit 83, and a scattered radiation data sorting unit. 84. The first data sorting unit 80, the first scattered radiation processing unit 81, the second data sorting unit 82, the second scattered radiation processing unit 83, and the scattered radiation data sorting unit 84 are connected in this order. A plurality of board integrated FPGAs 26 are connected to the first data sorting unit 80. In the first and second detector units 2 adjacent to each other in the circumferential direction of the camera 11 having the detector unit 2, the first scattered radiation processing unit 81 of the first detector unit 2 is the second detector unit. Connected to the second second data sorting unit 82. The second scattered radiation processing unit 83 of the second detector unit 2 is connected to the scattered radiation data sorting unit 84 of the first detector unit 2. In other words, the second detector unit 2 that inputs the output of the first scattered radiation processing unit 81 of the first detector unit 2 and the second scattered radiation processing unit 83 of the first detector unit 2 are input. The third detector unit 2 is arranged so as to sandwich the first detector unit 2 in the circumferential direction.

First, an overview of the function of the first data sorting unit 80 will be described. Packet data is input to the first data sort unit 80 from the plurality of board integrated FPGAs 26. The first data sorting unit 80 includes, for example, 15 entrance side changeover switches, 15 data buffers, an exit side changeover switch, and one unit sort circuit.
In the present embodiment, the case where there are 15 board integrated FPGAs 26 in the detector unit 2 has been described (see FIG. 5), but the number of board integrated FPGAs 26 may be other than 15. In this case, the numbers of the entrance-side changeover switch, the data buffer, and the exit-side changeover switch are all the same as the number of the board integrated FPGA 26.
Each data buffer includes two buffers, a first buffer and a second buffer.

  Each entrance side changeover switch is connected to one input terminal of each of the first buffer and the second buffer of one data buffer. Each outlet side changeover switch is connected to the output terminals of the first buffer and the second buffer of one data buffer, respectively. One board integrated FPGA 26 is connected to one of the first buffer and the second buffer of one data buffer by switching operation of the corresponding entrance side changeover switch. The unit sort circuit is connected to one of the first buffer and the second buffer of one data buffer by the switching operation of the corresponding exit side changeover switch. The output side of the unit sort circuit is connected to the first scattering processing unit 81.

  The data buffer is a buffer storage device having a function of storing packet data in the order of input, a function of outputting the stored packet data in the order of input, and a function of erasing the stored packet data collectively. The inlet side switch and the outlet side switch corresponding to one data buffer are connected to the unit sort circuit and one of the first buffer and the second buffer connected to the board integrated FPGA 26 for each time frame. It works to switch the person who is. That is, connection control is performed so that one or both of the first buffer and the second buffer are not connected at the same time with the board integrated FPGA 26 and the unit sort circuit in a certain time frame.

When the unit sort circuit reads stored packet data from the first buffer (or the second buffer), the unit sort circuit sequentially outputs the packet data from the plurality of first buffers (or the plurality of second buffers) in order of detection time. Has the function of Specifically, the unit sort circuit is the most out of the packet data of each of the fifteen first buffers (or fifteen second buffers) that have not yet been read and are located on the most output side. Packet data including early detection time information is retrieved and read. The unit sort circuit adds the read packet data to the end of the packet data string to be output. This packet data string includes a plurality of packet data arranged in order of detection time. As described above, the unit sort circuit repeatedly extracts the packet data including the earliest detection time information and adds it to the end of the packet data string. In the first buffer (or second buffer), a flag indicating that the packet has been extracted is set in the extracted packet data.
The unit sort circuit collects a predetermined number of packet data read in the order of detection times as described above, and outputs the packet data to the first scattering processing unit 81.

The first scattered radiation processing unit 81 has a function of performing scattered radiation processing on the input packet data and collecting a plurality of packet data resulting from a single incident γ-ray of predetermined energy. An example of the scattered radiation processing is described in Japanese Patent Application Laid-Open No. 2003-255048.
The output of the first scattered radiation processing unit 81 is input to the second data sorting unit 82 in the detector unit 2 and the second data sorting unit 82 in the adjacent detector unit 2. A plurality of scattered γ-rays caused by one γ-ray are not necessarily detected by the detectors 21 in the same detector unit 2 but are also detected by the detectors 21 in the adjacent detector units 2. There is. For this reason, after the scattered radiation processing is performed by the first scattered radiation processing section 81, the packet data is also transferred to the second data sorting section 82 in the adjacent detector unit 2.

  The first scattered radiation processing unit 81 of the first detector unit 2 is a partial detector (preferably, the first detector unit 2 located in a region close to the second detector unit 2 in the first detector unit 2. All the detectors 2 are divided into two equal parts in the circumferential direction, and all detectors (1/1 of the first detector unit 2 are located in the area close to the second detector unit 2). 2)), each packet data resulting from the γ-ray detection signal of 21 is output to the second data sorting unit 82 of the second detector unit 2.

  The second data sorting unit 82 of the second detector unit 2 includes a first scattered radiation processing unit 81 of the second detector unit 2 and a second detector unit adjacent to the detector unit 2 in the circumferential direction. In a state in which the respective data packets input from the second first scattered radiation processing unit 81 are combined, the packet data are arranged in order of detection time and output to the second scattered radiation processing unit 83.

  As described above, each second data sorting unit 82 outputs the detector unit 2 to which the second data sorting unit 82 belongs and the packet data collected from the adjacent detector units 2 in the order of detection times. Here, the two packet data strings input to the second data sorting unit 82 are already arranged in the order of detection times. For this reason, it is possible to arrange the packet data in the order of the detection time only by selecting the data packet sequence having the earliest detection time and merging the data packet into one column.

  The second scattered radiation processing unit 83 has the same configuration as the first scattered radiation processing unit 81 and operates on the same principle. The second scattered radiation processing unit 83 performs scattered radiation processing using the input data packet, and among the plurality of data packets caused by one γ-ray, the first γ-ray of the detector 21 on which the γ-ray is incident. It has a function of grasping a data packet based on the detection signal. The output of the second scattered radiation processing unit 83 is input to the scattered radiation data sorting unit 84 in the detector unit 2 and the scattered radiation data sorting unit 84 in the adjacent detector unit 2. That is, if the packet data after the scattered radiation processing is that of the detector unit (for example, the second detector unit) 2 to which the second scattered radiation processing unit 83 belongs, If the scattered radiation data sorting unit 84 in the unit (for example, the second detector unit) 2 is the packet data of the adjacent detector unit (for example, the first detector unit) 2, the packet data is stored. The data is output to the scattered radiation data sorting unit 84 in the adjacent detector unit (for example, the first detector unit) 2.

  The scattered radiation data sorting unit 84 has the same configuration and function as the second data sorting unit 82 and arranges the packet data output to the auxiliary data collection unit 4 in the order of detection time data.

Next, an outline of functions of the first scattered radiation processing unit 81 or the second scattered radiation processing unit 83 will be described.
In the first scattered radiation processing unit 81 or the second scattered radiation processing unit 83, the transmission data flag is added to the packet data to which the transmission data flag is added out of the packet data without being subjected to the scattering process. Process only unacknowledged packet data. Hereinafter, packet data in the scattering process is packet data to which no transmission data flag is added.
The detection energy of γ rays of the packet data arranged in the order of the detection time data is checked. When the detected energy matches 511 keV within a predetermined error set in advance, it is assumed that one γ ray of the γ ray pair emitted from the body of the subject P is detected as it is, and is valid for the packet data. A determination flag indicating that the detected γ ray information is present is added.
Then, when exceeding a predetermined error and exceeding 511 keV, a determination flag indicating that the detected γ-ray information is invalid is added to the packet data as noise.

Next, for packet data of detected energy of γ rays exceeding a predetermined error and smaller than 511 keV, within a predetermined time window length, for example, less than 6 nanoseconds, and within a predetermined distance, for example, less than 4 cm When the detectors 21 are within the distance of each other, the sum of the detected energy of the γ rays of the two packet data is checked, and when the sum is within a predetermined value range, the two packet data are It is determined that the scattered radiation for the incident γ-ray of 511 keV has been detected.
The range where the sum is a predetermined value is, for example, a case where it exceeds 450 keV and is less than 511 keV.

Incidentally, 6 nanoseconds set as the predetermined time window length is determined from the time resolution of the CdTe detector 21 of the present embodiment. The distance between the detectors 21 of less than 4 cm is obtained from the result of evaluating the probability of detection by another detector 21 after Compton scattering in the detector 21 by simulation calculation.
The value of the time window length for the scattering determination and the distance between the detectors 21 are the types of elements used in the detector 21, the amount of radioactive material used for the inspection, and the γ-rays emitted from the radioactive material. It is better to optimize according to energy.

If it is determined that the scattered radiation has been detected, a determination flag is added to indicate that one of the two packet data is valid detected γ-ray information, and the sum of the detected energy of the γ-ray is used. Replace with detected γ-ray energy. A determination flag indicating invalid detection γ ray information is added to the other packet data.
In the second scattered radiation processing unit 83, packet data to which a determination flag that is already valid γ-ray information or a determination flag that is invalid γ-ray information in the first scattered radiation processing unit 81 is added. Is not determined for the scattering process described above.
The packet data that has been subjected to the scattering processing by the first scattering processing unit 81 and the second scattering processing unit 83 in this way is sent to the scattered radiation data sorting unit 84 and is assumed to be invalid γ-ray information in the scattered radiation data sorting unit 84. The packet data to which the determination flag is added is rejected, and only the packet data to which the determination flag that is valid gamma ray information is added and the packet data to which such a determination flag is not added are sent to the auxiliary data collection unit 4. Output.
Incidentally, the packet data that has not been determined to be effective γ-ray information by the scattered radiation data sorting unit 84 includes transmission data described later. The packet data related to the transmission data is not rejected here.

  The auxiliary data collection unit 4 includes packet data including a plurality of detector units 2, which are added with a determination flag that is valid γ-ray information output from 10 detector units 2 in the example of FIG. 2. Are sorted in the order of detection time and input to the data sorting unit 12a of the data processing device 12A.

(Function of data processing device)
Next, the function of each component in the data processing device 12A will be described with reference to FIG. As described above, the data processing device 12A includes the data sorting unit 12a, the simultaneous measurement unit 12b, the sinogram data storage unit 12c, the single event measurement unit (single event rate acquisition unit) 12d, the correction coefficient calculation unit (correction coefficient calculation unit) 12e. , A sinogram data correction unit 12f, a transmission data storage unit 12g, and an image reconstruction unit (image generation unit) 12h.
In the data sorting unit 12a, the packet data input from the four auxiliary data collection units 4 are sorted in the order of detection time, and the simultaneous measurement unit 12b, the single event measurement unit 12d, and the transmission data storage unit 121g are flagged with transmission data. Whether or not is added and the detected γ-ray energy value are checked and distributed. Packet data with a transmission data flag not added and a detected γ-ray energy value of more than 450 keV and less than or equal to 511 keV is sent to the simultaneous measurement unit 12b and the single event measurement unit 12d, and the transmission data flag is added. The packet data is sent to the transmission data storage unit 121g.

The simultaneous measurement unit 12b performs simultaneous measurement processing on the packet data of the emission data. This simultaneous measurement process is a known technique, and two are included within a predetermined time window length T (nanosecond), and both of the detected γ-ray energy values are within a predetermined range (511 keV). For example, based on two pieces of packet data that satisfy 450 keV and less than or equal to 511 keV), a coincidence event is determined and pair data is generated. Then, the obtained pair data is input to the sinogram data storage unit 12c. The pair data includes data of a combination of the two detectors 21.
Note that the imaging area is limited by FOV (Field Of View), and the combinations of the two detectors 21 to be subjected to simultaneous measurement processing are limited.
Incidentally, when three or more pieces of packet data are included within the predetermined time window length T, it is determined as multi-event, and the simultaneous measurement unit 12b determines the data for the packet data determined as multi-event. Is not generated, and the packet data is discarded.

  The sinogram data storage unit 12c stores sinogram data in association with the input pair data.

Single event measurement unit 12d, a specific γ-ray energy emitted by the body of the radionuclide of the subject P, and packet data emission data measured with γ line pair 511keV from this case ^{18 F,} simultaneous measurement It accumulates for each combination of two detectors 21 to be processed. For example, the two packet data output from the detector 21 with the detector ID i and the detector 21 with the detector ID j, which are the targets of the simultaneous measurement process, are stored as the single event count rate Sij. The accumulation of the single event count rate Sij is performed for each movement position in the body axis direction of the subject P of the bed 14 (for each body axis direction scan position).
Even if the emission data is collected simultaneously with the transmission data, the packet data includes the detected energy value information. Therefore, the single event measuring unit 12d reads the detected energy value and counts the single event count rate Sij of the emission data. Can be accumulated.

  The sinogram data correction unit 12f has a detector 21 having a detector ID i corresponding to each point of the sinogram data and a detector 21 having a detector ID j corresponding to each position of the subject P in the body axis direction. In order to perform the correction process for the combination of the correction coefficient, the correction coefficient calculation unit 12e calculates a correction coefficient (1 + αij · T · Sij) to be described later, and the correction coefficient calculated by the correction coefficient calculation unit 12e corresponds to the corresponding data point of the sinogram data. Multiply

Hereinafter, a method for obtaining the correction coefficient (1 + αij · T · Sij) will be described. As described above, when the multi-event is determined in the simultaneous measurement process, the packet data is discarded without being used for the sinogram data.
As described above, the semiconductor γ-ray detector can generate detection γ-ray information for each detector 21, and the time resolution is higher than that when a scintillator and a photomultiplier tube are used. Data loss rate is small during signal processing. However, data loss due to multi-events during simultaneous measurement processing is an essential problem of simultaneous measurement processing, and data loss due to multi-events cannot be reduced except by reducing the time window length T during simultaneous measurement. .
Therefore, in a PET apparatus using a semiconductor γ-ray detector, in order to improve the throughput of the data processing system and reduce data loss, it is only necessary to consider multi-events during simultaneous measurement processing.

If the time resolution of the detector 21 and the subsequent time resolution of the analog ASIC 24, ADC 25, and board integrated FPGA 26 are evaluated and the time window length T is appropriately set, the multi-event during the simultaneous measurement process is the same as the simultaneous measurement process. A true coincidence event in which a gamma ray pair is detected in a detector 21 having a target detector ID i and a detector 21 having a detector ID j (hereinafter referred to as a detector pair ij), and another gamma ray pair Is detected only by one of the detector 21 having the detector ID i and the detector 21 having the detector ID j, and a single event that is not detected by the other is accidentally simultaneously measured. is there. There is a low probability that a plurality of coincidence events and a plurality of single events are accidentally simultaneously measured and become multi-events.
The count rate Mij (cps) at which coincidence events and single events are accidentally measured simultaneously is given by the following equation, where the count rates of coincidence events and single events are Cij (cps) and Sij (cps), respectively.
Mij = αij · Cij · T · Sij (1)
Here, the coefficient αij is a product of the ratio of limiting the simultaneous measurement pair (detector pair ij) at the time of simultaneous measurement and the proportional coefficient of the effective time window length T, and is a constant that does not depend on the subject P. . The correction coefficient calculation unit 12e is stored in advance in a storage device (not shown) so that it can be referred to.

Since the multi-event count rate Mij is equal to the data loss rate, the corrected coincidence event count rate obtained by adding the data loss rate is given by the following equation.
Cij + Mij = Cij · (1 + αij · T · Sij) (2)
From Equation (2), the correction coefficient is (1 + αij · T · Sij).
The correction coefficient calculation unit 12e determines the combination of the detector pair ij corresponding to each point of the sinogram data requested from the sinogram data correction unit 12f, and the corresponding detector pair ij accumulated in the single event measurement unit 12d. The single event count rate Sij is read out and multiplied by the time window length T and constant αij stored in advance in the storage device to calculate the correction coefficient (1 + αij · T · Sij), and the sinogram data correction unit 12f Output.

  The image reconstruction unit 12h back-projects the sinogram data corrected by the sinogram data correction unit 12f using a filtered back projection method or the like to generate a PET image (tomographic image). Similarly, transmission data is back-projected using a filtered back projection method or the like to obtain an emission data attenuation distribution due to absorption in the subject P, and this is converted into a back-projected image based on the emission data for each pixel. Multiply to correct the absorption.

  According to the present embodiment, there is no need to obtain a correction amount in advance by a phantom test or the like, and when the body shape of the subject P is significantly different from the standard body shape used in the phantom test, Even when there is a characteristic variation, the correction is made based on the emission data, so that there is little correction error for the counting loss of the simultaneous measurement.

In addition, since the single event count rate Sij is obtained for each pair of detectors 21 to be subjected to the simultaneous measurement processing and the correction coefficient (1 + αij · T · Sij) is calculated, the radioactivity in the body of the subject P is calculated. When the deviation of the concentration distribution (radioactivity concentration) of the substance is large and the count rate Sij of the single event between the detector pair ij is large, the correction coefficient is corrected for each detector pair ij. The accuracy of sinogram data is improved.
The method for determining the coefficient αij may be calculated analytically or using experimental values.

(Modification of the first embodiment)
In the present embodiment, the single event count rate Sij is acquired for each scan position in the body axis direction of the subject P and for each detector pair ij to be subjected to simultaneous measurement processing. It is not limited.
In the correction coefficient calculation unit 12e, the detector pairs ij are grouped, and the detector pairs ij are grouped for each group in which the value of αij is substantially the same from the geometrical relationship of the detector pairs ij. Then, the coincidence event count rate Cij is read from the sinogram data storage unit 12c, and a correlation formula between the coincidence event count rate Cij and the single event count rate Sij is calculated and stored. When the sinogram data correction unit 12f instructs the calculation of the correction coefficient for the data of each point of the sinogram data (corresponding to the coincidence event count rate Cij), it is determined to which group the detector pair ij belongs, A correction factor (1 + αij · T · Sij) may be calculated by estimating the single event count rate Sij from the correlation equation of the determined detector pair ij group, and may be output to the sinogram data correction unit 12f. .

In this way, the detector pairs ij are grouped, and from the correlation equation between the distribution of the single event count rate Sij and the coincidence event count rate Cij for the detector pair ij in the same group, By obtaining the estimated value of the single event count rate Sij with respect to the count rate Cij, only by the correlation equation between the coincidence event count rate Cij and the single event count rate Sij at a certain body axis position of the subject P, It is also used to calculate correction coefficients for other body axis positions.
As a result, the number of processes for obtaining the single event count rate Sij can be reduced, so that the processing time of the entire image processing can be shortened.

  In addition, since the counting loss at the time of simultaneous measurement processing by multi-events is larger as the counting rate of coincidence events is higher, the detector pair ij at the body axis direction scan position where the counting rate of coincidence events is higher in order to improve the quantitativeness of the image. The correction based on the correlation formula between the distribution of the single event count rate Sij and the coincidence event count rate Cij for the grouping is preferable because it is highly accurate correction for the higher density image portion.

Further, even when grouped detector pair ij, a plurality of groups from the detector pair ij, grouped choose detector pair ji _G representative, detector pair ji _G representative of that particular group Only, the distribution of the count rate Sij _G of the single event and the distribution of the count rate Cij _G of the coincidence event may be obtained, and the correlation equation thereof may be obtained. This reduces the amount of data processing and reduces the burden on the data processing device 12A.

  As described above, according to the present embodiment and the modification thereof, the counting drop at the time of the simultaneous measurement process due to the multi-event is corrected. Therefore, as shown in the curve before correction in FIG. It can be improved that the counting loss increases due to the multi-event and the deviation rate of the coincidence event from the theoretical straight line becomes large.

<< Second Embodiment >>
Next, a second embodiment will be described with reference to FIG. In the PET apparatus 1B of the present embodiment, the data processing apparatus 12B is combined with the camera 11. The data processing device 12B includes a data sorting unit 12a, a simultaneous measurement unit 12b, a sinogram data storage unit 12c, a random event measurement unit (single event rate acquisition unit) 12i, a single event rate calculation unit (single event rate acquisition unit) 12j, and a correction. A coefficient calculation unit (correction coefficient calculation means) 12e, a sinogram data correction unit 12f, a transmission data storage unit 12g, and an image reconstruction unit (image generation unit) 12h are included.
In the first embodiment, the single event count rate is measured by the single event measurement unit 12d in order to calculate the correction coefficient (1 + αij · T · Sij). However, in this embodiment, the data processing unit is used instead. 12B has a random event measurement unit 12i, and measures a random event count rate Rij. The same components as those in the first embodiment are denoted by the same reference numerals, and redundant description is omitted.

The random event measurement unit 12i acquires the random event count rate Rij by a method (delayed coincidence) in which the time window is shifted by at least the time window length T of the simultaneous measurement and is simultaneously measured with the time window length T.
The random event count rate R _1ij is theoretically related to the single event count rate Sij by the following equation.
R _1ij = βij · T · (Sij) ² (3)
Here, βij is the product of the ratio of the simultaneous measurement detector pair ij and the proportional count of the effective time window length, and is a constant that does not depend on the subject P. However, even in the case of this delayed coincidence, if three or more single events are measured, the multi event is rejected from the random event count rate Rij, so the number of multi events to be rejected at the time of delayed coincidence is counted. The rate R _2ij is as follows:
R _2ij = γij · R _1ij · T · Sij (4)
Here, γij is the product of the ratio of the simultaneous measurement detector pair ij and the proportional count of the effective time window length, and is a constant that does not depend on the subject P.

Therefore, the measured random event count rate Rij is expressed as follows.
Rij = R _1ij −R _2ij = βij · T · (Sij) ² −γij · βij · R _1ij · T ² · (Sij) ³ (5)
From the equation (5), the single event count rate Sij can be calculated by solving the cubic equation relating to the unknown number Sij in the equation (5) using the measured random event count rate Rij.

The single event rate calculation unit 12j calculates the single event count rate Sij from the equation (5) based on Rij obtained by the random event measurement unit 12i.
If the single event count rate Sij is obtained, as in the first embodiment, the correction coefficient (1 + αij ··) is applied to the coincidence event count rate Cij corresponding to each point of the sinogram data from the sinogram data correction unit 12f. In response to the T.Sij) calculation instruction, the correction coefficient calculation unit 12e reads the single event count rate Sij from the single event rate calculation unit 12j, calculates the correction coefficient (1 + αij · T · Sij), and calculates the sinogram data correction unit 12f. Can be output.

  Further, as in the modification of the first embodiment, the detector pairs ij are grouped, and a correlation coefficient of data between the coincidence event count rate Cij and the single event count rate Sij in the same group is calculated as a correction coefficient. The correction coefficient calculation unit 12e estimates the single event count rate Sij corresponding to the coincidence event count rate Cij and obtains the correction coefficient (1 + αij · T · Sij). good.

As described above, in the second embodiment, without counting the single event count rate Sij directly, the coincidence event count rate Cij is corrected so as to be proportional to the radiation concentration in the same manner as in the first embodiment. And quantitative images can be reconstructed.
Usually, the data processing devices 12A and 12B of the PET apparatuses 1A and 1B have the simultaneous measurement unit 12b in order to increase the accuracy of the simultaneous measurement processing, and also have the random event measurement unit 12i for random correction. There are many. Therefore, in the case of this embodiment, it is not necessary to newly provide the single event measurement unit 12d, and a simple configuration can be achieved.

(Other modifications of the first and second embodiments)
In the first and second embodiments described above, correction is performed at the sinogram data stage, but the present invention is not limited to this. You may correct | amend for every pixel after reconstruction to a PET image. In the first and second embodiments, the correction is performed for each detector pair ij. For example, the count rate Sij of the single event of the detector pair ij having a high contribution rate is related to each pixel in the PET image. An average value <Sij> may be obtained, and the pixel density may be multiplied by a correction coefficient (1+ <αij> · T · <Sij>).
Here, <αij> relates to each pixel in the PET image, for example, an average value of αij of the detector pair ij having a high contribution rate.
In this case as well, the correction coefficient may be calculated for each scan position of the subject P in the body axis direction, or one scan position for all the scan positions of the subject P in the body axis direction. It is also possible to calculate a correlation formula between the density of the pixel before correction and the correction coefficient in step, and apply the correlation formula to the pixels at all scan positions.

Further, in the first and second embodiments, an average single event count rate <Sij _G > of a plurality of representative detector pairs ij _G is obtained so that the quantitativeness of the PET image is improved as a whole. Then, a correction coefficient (1+ <αij _G > · T · <Sij _G >) may be calculated and multiplied by all pixels or sinogram data.
In order to obtain the average single event count rate <Sij _G > of the representative plurality of detector pairs ij _G, a pair of representative detector units 2 is selected, and the pair of the detector units 2 is selected. An average single event count rate <Sij _G > may be calculated for the included detector 21 to obtain a correction coefficient (1+ <αij _G > · T · <Sij _G >).
In addition, as a unit (population) for calculating the average single event count rate <Sij _G >, for example, a pair of unit substrates 20 may be considered.
By such correction, it is possible to uniformly correct the deterioration of quantitativeness due to the counting loss in the simultaneous measurement process over the entire PET image.

As described above, according to the first embodiment, the second embodiment, and modifications thereof, the following effects can be obtained.
(1) The correction of the coincidence event count rate is performed by measuring the single event count rate input to the simultaneous measurement unit 12b by the single event measurement unit 12d using an analytically obtained correction coefficient calculation formula, or The correction coefficient can be calculated by indirectly measuring by the random event measuring unit 12i, and correction with high accuracy is possible without depending on the subject P, and the quantitativeness of the PET image by the PET apparatuses 1A and 1B is improved. Therefore, more accurate diagnosis is possible.
Even when the counting rate, that is, the radioactive concentration of the subject P is high, a highly quantitative PET image can be obtained.
As a result, the imaging time can be shortened while maintaining high quantitativeness.
(2) The unit of the population of the detectors 21 for which the correction coefficient is to be calculated is small, the unit of the detector 21 being small, the unit of the detector 21 included in the unit substrate 20, and the detector 21 included in the detector unit 2 The unit can be made flexible.
And according to the target unit, the correction coefficient is corrected for each projection data of the sinogram data, uniform correction for the entire sinogram data, correction for each pixel of the reconstructed image, uniform correction for the entire reconstructed image, Can be corrected flexibly.
After the emission data has been acquired, if an input is made from the operation console 13 so that the user can select and switch the correction method, an appropriate correction image according to the purpose can be obtained.
(3) The correction coefficient is calculated based on the single event count rate, and the correction coefficient can be easily obtained. Further, the correction process can be easily performed by multiplying sinogram data by a correction coefficient or multiplying pixel data of a PET image by a correction coefficient.
(4) When the single event count rate is calculated from the random event count rate in the random event measurement unit 12i as in the second embodiment, the PET apparatus originally performs a delayed coincidence process to perform random correction. Since there are many things and the functions can be used, the single event measuring unit 12d is not necessary, and the data processing device 12B can be simplified.

<< Third Embodiment >>
Next, the case of a PET apparatus using the scintillator and photomultiplier tube of the third embodiment will be described with reference to FIGS. 1 and 11. In addition, although description of the detailed structure of the camera in that case is abbreviate | omitted, in the structure of the detector unit 2 shown in FIG. 2, it is the same structure except a detector part fundamentally.
The photomultiplier tube constituting the γ-ray detector has a large dead time and a poor time resolution as compared with the semiconductor detector. Therefore, data loss in the data processing system of the analog ASIC 24 and the substrate integrated FPGA 26 is large. This is for data loss in the data processing system in the unit substrate 20 (see FIG. 5), and is called dead time correction. This can be obtained analytically or can be corrected based on experimentally obtained data.

  The function of each component in the data processing device 12C in this embodiment will be described. The data processing device 12C includes a data sorting unit 12a, a simultaneous measurement unit 12b, a sinogram data storage unit 12c, a single event measurement unit (single event rate acquisition unit) 12d, a correction coefficient calculation unit (correction coefficient calculation unit) 12e, and transmission data storage. A unit 12g, an image reconstruction unit (image generation unit) 12h, a random event measurement unit 12i, a random correction unit 12k, a single event rate correction unit 12m, a correction coefficient calculation unit 12n, and a sinogram data correction unit 12p. The same components as those in the first embodiment and the second embodiment are denoted by the same reference numerals, and redundant description is omitted.

Hereinafter, a case where the data loss coefficient is obtained analytically will be described as an example.
The simultaneous measurement unit 12b and the random event measurement unit 12i perform simultaneous measurement (Cij) and delayed coincidence measurement (Rij) using a time window length Ts that is larger than that of the semiconductor γ-ray detector. In a γ-ray detector using a scintillator and a photomultiplier tube, the time resolution is poor, and therefore the simultaneous measurement result Cij in the simultaneous measurement unit 12b includes an accidental simultaneous measurement Rij of a single event, so the random correction unit 12k In Cij-Rij, random correction is performed.
The single event measurement unit 12d measures the single event count rate Si of each detector i or each data processing system i.
The single event rate calculation unit 12m calculates the single event count rate Stotal input to the simultaneous measurement unit 12b, which is the sum of the single event count rates Si measured by the single event measurement unit 12d.

When the single event count rate of each detector i or each data processing system i is Si, the data loss rate h _i (Si) due to dead time is obtained, and the total single event data loss rate h is the product of them. It can be calculated as:
_{h = Πh i (Si) ·····} (6)
Furthermore, the simultaneous measurement is a pairing of a single event pair, the square of the data loss rate of the single event is the data loss rate of the coincidence event, and the reciprocal thereof is the dead time correction coefficient H of the simultaneous measurement as in the following equation: It becomes.
H = h- ² (7)
Therefore, the total correction coefficient for both the correction of counting loss and the dead time correction in the simultaneous measurement processing by multi-event is expressed by the following equation.
Cij · H · (1 + α · Ts · Stotal) (8)
Here, the correction coefficient calculation unit 12n calculates the data loss rate h and the dead time correction coefficient H, and the sinogram data correction unit 12p corrects each data of the sinogram data storage unit 12c by the total correction coefficient. Do.

  From Expressions (6) and (7), the dead time correction coefficient H is a function of the single event count rate Si, and high-accuracy correction can be performed by measuring with the single event measurement unit 12d. Instead of this, it is possible to omit measuring the single event count rate Si of each detector i or data processing system i by obtaining and using the average value Si from Stotal. In that case, the in-vivo radioactivity concentration distribution according to the measurement position cannot be considered, but the correction process can be simplified.

  In order to obtain the dead time correction coefficient H experimentally, a curve of the single event count rate input to the simultaneous measurement unit 12b with respect to the radioactivity concentration is obtained, and the data loss of the single event count rate by the ratio Melco with the theoretical straight line. The rate h is obtained. Next, the dead time correction coefficient H is obtained from the equations (6) and (7), and the correction coefficient calculation unit 12n may be referred to as a table function depending on the single event count rate Si.

According to the present embodiment, the data loss due to the dead time of the data processing system in the unit substrate and the data loss due to multi-event during the simultaneous measurement process are analytically separated, and the total correction coefficient is calculated from each correction coefficient. Thus, it is possible to correct a highly accurate PET image independent of the subject P. As a result, a highly quantitative PET image can be obtained.
As shown in FIG. 12, it is possible to approximate a theoretical straight line by a multi-event count rate correction process and a dead time correction process for the coincidence event count rate.
In addition, since the data loss in the data processing system due to the dead time is calculated by integrating from the single event count rate Si of each data processing system i, the dead time can be corrected with higher accuracy.

It is a perspective view which shows the structure of the PET apparatus which concerns on the 1st Embodiment of this invention. It is the figure which showed typically the cross section of the circumferential direction of the camera of the PET apparatus of FIG. 1, and the functional block structure of a data processor. (A) is a partially broken perspective view of a camera showing a configuration for mounting a detector unit to a camera, and (b) is a cross-sectional view showing a detector unit mounted state at the center of the camera. (A) is the front view which showed the unit board | substrate using the semiconductor gamma-ray detector based on 1st Embodiment, (b) is the side view of the same (a). FIG. 2 is a block diagram showing a schematic configuration of a unit integrated FPGA and a connection relationship between an analog ASIC and a unit integrated FPGA. It is the block diagram which showed typically the analog signal processing circuit of analog ASIC. It is a figure which shows the structure of the detector unit which accommodated the several unit board | substrate, (a) is sectional drawing which looked at the detector unit in the circumferential direction, (b) is YY sectional drawing in (a). is there. It is a block block diagram explaining the scattered radiation processing function in integrated FPGA. It is a figure which shows the relationship between the radioactive concentration and the counting rate of coincidence event. It is the figure which showed typically the cross section of the circumferential direction of the camera of the PET apparatus which concerns on the 2nd Embodiment of this invention, and the functional block structure of a data processor. It is the figure which showed typically the functional block structure of the data processor of the PET apparatus as a nuclear medicine diagnostic apparatus which concerns on the 3rd Embodiment of this invention. It is a figure which shows the relationship between the radioactive concentration and the counting rate of coincidence event.

Explanation of symbols

1A, 1B PET device 2 Detector unit 2A Unit support member 4 Auxiliary data collection unit 11 Camera 11a Lid 11b Opening portion 12A, 12B Data processing device 12a Data sorting unit 12b Simultaneous measurement unit 12c Sinogram data storage unit 12d Single event measurement unit ( Single event rate acquisition means)
12e Correction coefficient calculation unit (correction coefficient calculation means)
12f sinogram data correction unit 12g transmission data storage unit 12h image reconstruction unit (image generation unit)
12i Random event measurement unit (single event rate acquisition means)
12j Single event rate calculation unit (single event rate acquisition means)
13 operation console 13a display device 13b input operation unit 14 bed 20 unit substrate 20a detector substrate region 20b ASIC substrate region 20c unit substrate body 21 semiconductor radiation detector (γ-ray detector)
22 Capacitor 23 Resistor 24 Analog ASIC (Data Generation Unit)
24A First system 24B Slow system 24a Preamplifier 24b Timing pickoff circuit 24c Threshold control circuit 24d Waveform shaping circuit 24e Peak hold circuit 25 Analog / digital converter 26 Substrate integrated FPGA (data generation unit)
30 Housing 30a Top plate 30c Side plate 31 Unit integrated FPGA
32 integrated substrate 33 analog signal processing circuit 34 detection signal processing unit 35 timing detection unit 36 detector control unit 37 data transfer unit 38 connector 39 guide member 41a position holding hole 41b position holding member 41c nut 42, 43 cable 80 first data sort Unit 81 First scattered radiation processing unit 82 Second data sorting unit 83 Second scattered radiation processing unit 84 Scattered radiation data sorting unit C2 Board connector

Claims (7)

a plurality of γ-ray detectors that output γ-ray detection signals in response to detection of γ-rays;
A plurality of data generation units for generating detection γ-ray information including detection time information, a detector ID for identifying the γ-ray detector, and a detected γ-ray energy value based on the γ-ray detection signal;
A data processing device that performs simultaneous measurement processing in the simultaneous measurement unit based on the detected γ-ray information, and generates a tomographic image based on the result of the simultaneous measurement processing in the image generation unit;
In a positron emission tomography apparatus comprising:
The data processing device includes:
In the simultaneous measurement process, a single event rate acquisition means for acquiring a count rate of an input single event;
Correction count calculation means for calculating a correction coefficient based on the single event count rate acquired by the single event rate acquisition means;
Have
A positron emission tomography apparatus which corrects the tomographic image based on the calculated correction count.   The data processing apparatus includes a single event measurement unit that measures a count rate of the single event based on the detected γ-ray information input to the simultaneous measurement unit as the single event rate acquisition unit. The positron emission tomography apparatus according to claim 1. The data processing device, as the single event rate acquisition means, a random event measurement unit that measures a count rate of random events based on the detected γ-ray information,
The positron emission tomography apparatus according to claim 1, further comprising: a single event rate calculation unit that calculates the count rate of the single event based on the count rate of the random event.   The correction count calculation means calculates a value obtained by further multiplying a product of the count rate of the single event and the time window length T in the simultaneous measurement processing and adding 1 as the correction coefficient. The positron emission tomography apparatus according to any one of claims 1 to 3.   5. The positron emission according to claim 1, wherein the correction coefficient calculation unit calculates the correction coefficient for each pair of the γ-ray detectors to be combined in the simultaneous measurement process. Type tomography equipment.   The single event rate acquisition means is capable of selecting and switching a population of the γ-ray detectors that acquire the count rate of the single event, and based on the calculated correction count according to the selection. The positron emission tomography apparatus according to claim 5, wherein the tomographic image is corrected.   The positron emission tomography apparatus according to claim 1, wherein a semiconductor γ-ray detector is used as the γ-ray detector.

Images