JP2011027601A - Radiation image processor, radiation image processing method, and radiation image processing program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a processor for processing images by quickly generating two-dimensional image data from a lot of photon information detected with a gamma camera without need of a lot of fixed length frame memories and without reducing resolution and fixed quantity property. <P>SOLUTION: The radiation image processor includes: a means for obtaining XY two-dimensional position information of photons radiated from a specimen and obtaining an energy value of each photon as Z values; a means for obtaining detection conditions of each photon as detection condition information; a classifying means for classifying each detection position information and each Z value in each detection condition based on the corresponding detection condition information; and a converting means for converting each detection position information into conversion position information of pixel numbers smaller than a XY two-dimensional space in addresses. The Z value counts up photon output frequencies in each classification due to the classifying means when it is within a previously defined effective range, measures the photon output frequencies in each conversion position information, and generates two-dimensional image data in each classification due to the classifying means. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a radiation image processing apparatus using a gamma camera or the like typified by a PET or SPECT apparatus in nuclear medicine.

  A gamma camera is a device for visualizing the distribution of a radiopharmaceutical administered into a subject by detecting gamma rays and observing the distribution of cancer cells. A radiopharmaceutical is a compound labeled with a radioisotope that emits gamma rays.

  In Patent Document 1, which is a patent application by the applicant, image data representing an image by a plurality of granular patterns, such as an image taken by a photomultiplier tube, is converted to image data based on a scintillator arrangement with high accuracy. Technology etc. are disclosed.

  For example, such an image conversion technique is incorporated in a micro SPECT device for small animals disclosed in Patent Document 2 to visualize blood circulation, and is applied to various animal experiments. Furthermore, this image conversion technology is not limited to SPECT, but has also been applied to PET devices for small animals (positron emission tomography), and administration of clinical drugs synthesized with radioisotopes to experimental animals such as mice and rats. It is also used in the field of drug discovery to visualize the metabolic state and examine the effects of drugs.

JP 2007-121259 A JP 2004-233149 A

  SPECT / PET, a technology for deploying nuclear medicine diagnostic technology developed for humans in the bio field, has come to be collectively called molecular imaging, and the need for medium animals has also increased. There are three middle animal subjects. (1) The first is a large experimental animal, which is used for verification of drug effects closer to humans using primates (cynomolgus monkeys, marmosets, etc.) closer to humans than mice and rats. To do. (2) Secondly, in the field of veterinary medicine, the healing effect of pets is regarded as important in a stressed society, and lifestyle-related diseases are increasing in dogs and cats, etc., and they are used for diagnosing diseases like humans. thing. (3) Thirdly, for the purpose of human brain research, molecular imaging is gaining attention in fMRI and optical topography as well as visualization of brain function, but it is difficult to introduce clinical PET / SPECT in terms of cost and installation space. There is a demand for an inexpensive small PET / SPECT device dedicated to the brain.

  In CT / MRI, high resolution is important. On the other hand, since PET / SPECT performs quantitative calculations such as circulating blood volume, metabolic volume, and blood flow volume based on the captured image, the image is highly quantitative, that is, the accuracy of the value of each pixel (radiation count).・ High reliability is required. Therefore, the calibration accuracy of the basic gamma camera is important.

  In the gamma camera, the position (X, Y) of the gamma ray photon on the detector surface and the approach energy Z are measured, and photon event data (RAW data) is transferred to the PC in time series. For example, conventionally, a three-dimensional frame memory (photon counter) of {X: 256 × Y: 256 × Z: 32} or {X: 512 × Y: 512 × Z: 128} is prepared on a computer. Then, every time one photon is received, the corresponding (X, Y, Z) address value is counted up one by one. These are realized by computer software, and the frame memory is a part of main memory mounted on the computer.

  However, for medium animals, the size of the subject itself is large, and in order to maintain at least the same level of accuracy as that of small animals, it is 8 times in the XY direction in the two-dimensional space range as compared to small animals, and the energy resolution is also high. Eight times as much as {X: 4096 × Y: 4096 × Z: 1024, 32 Gbytes} three-dimensional frame memory is required, which is difficult to realize with a normal computer. Conventionally, images were captured in frame units and correction processing was performed, but for middle animals, images may be taken while moving the bed that supports the subject or rotating the gamma camera around the subject. Since the number of frames to be corrected increases for each imaging condition, a problem in work efficiency also occurs.

  Therefore, the present invention quickly processes a large amount of photon information detected by a gamma camera without requiring a large amount of fixed-length frame memory and without reducing resolution and quantification. It is an object of the present invention to realize a radiation image processing apparatus that can generate image data and perform image processing.

  In order to solve the above-mentioned problem, the invention according to claim 1 is a detection means for detecting photons emitted from a subject, and detection position information in an XY two-dimensional space range of each photon detected by the detection means ( X, Y) and a photon information acquisition means for acquiring the energy value of each photon as a z value, and detection condition information acquisition for acquiring a detection condition when each photon is detected as detection condition information And the detection position information (X, Y) and z value acquired by the photon information acquisition means and the detection condition information acquired by the detection condition information acquisition means for each photon. Storage control means for storing in the storage means, classification means for classifying each of the detected position information (X, Y) and z value for each same detection condition based on the corresponding detection condition information, The detected position information (X, Y) is converted position information (x, y) (x <X and y <in a two-dimensional image defined by the number of pixels smaller than the number of pixels defined in the XY two-dimensional space range. Conversion means for converting the address to Y), determination means for determining whether the z value of each photon is within a predetermined effective range based on the conversion position information (x, y), and As a result, when the z value of the photon is within a pre-defined effective range, the conversion position information (x, y) corresponding to each photon for each photon for each classification by the classification unit. ) Counts up the photon output frequency at the position on the two-dimensional image indicated by any of the z values, and measures the photon output frequency for each converted position information (x, y); Based on the photon output frequency corresponding to the converted position information (x, y) , For each classification by the classifying means, a radiographic image processing apparatus characterized by having a generating means for generating a two-dimensional image data representing the two-dimensional image.

  In order to solve the above-mentioned problem, the invention according to claim 2 is directed to calibration photon detection means for detecting photons emitted by a subject for apparatus calibration, and XY two included in each photon detected by the calibration photon detection means. Calibration photon detection position information (X, Y) in a dimensional space range is acquired, and calibration photon information acquisition means for acquiring the energy value of each photon as a z value, and each of the detected photon detection means For photons, the photon output frequency in the XY two-dimensional space range indicated by the corresponding calibration photon detection position information (X, Y) is counted up regardless of the z value, and corresponds to the XY two-dimensional space range. To create a two-dimensional fixed-length array, and based on the photon output frequency of the two-dimensional fixed-length array corresponding to the XY two-dimensional space range, XY address conversion table creating means for creating an XY address conversion table when the conversion means performs address conversion, and referring to the XY address conversion table, the calibration photon detection position information (X, Y) in the two-dimensional image The z-value as the photon output frequency at the position on the two-dimensional image indicated by the converted position information (x, y) corresponding to each photon for each photon converted to the converted position information (x, y) Counting up separately, three-dimensional fixed length array creating means for creating a three-dimensional fixed length array based on the converted position information (x, y) and z value, and based on the three-dimensional fixed length array, the two-dimensional Z address conversion table creating means for creating a z address conversion table for defining an effective range of z values at each position (x, y) on the image, wherein the conversion means is the XY Based on the XY address conversion table created by the address conversion table creating means, each calibration photon detection position information (X, Y) is converted into address conversion position information (x, y) in the two-dimensional image, The determination means determines whether the z value of each photon is within a predetermined effective range based on the value in the conversion position information (x, y) of the z address conversion table created by the z address conversion table. The radiological image processing apparatus according to claim 1, wherein it is determined whether or not.

  In order to solve the above-mentioned problem, in the invention according to claim 3, the detection means is constituted by a plurality of detectors, and the photon information acquisition means and the detection condition information acquisition means correspond to the plurality of detectors, A plurality of sets of detection position information (X, Y) and z values, and detection condition information are obtained, and the storage control means detects each detection position information (X, Y) and z value of each of the detectors, and detection Condition information is associated and stored in the storage means, and the classification means is based on the plurality of sets of detection position information (X, Y), z values, and detection condition information acquired corresponding to each detector. And each set is classified according to the same detection condition, and the conversion means corresponds to each detection position information (X, Y) corresponding to the detector from which the detection position information (X, Y) is acquired. The address conversion is performed using each XY address conversion table, and the determination means Is the z address conversion table corresponding to the detector from which the z value is obtained, and the z value of each photon is valid for each position (x, y) on the two-dimensional image. Based on each z address conversion table for defining a range, it is determined whether or not the z value of each photon is within a predetermined effective range, and the measurement means, as the photon output frequency, The photon output frequencies at positions on a plurality of two-dimensional images corresponding to the plurality of detectors are independently measured, and the generation unit classifies each of the plurality of detectors by the classification unit. The radiographic image processing apparatus according to claim 1, wherein a plurality of the two-dimensional image data are generated every time.

  In order to solve the above-mentioned problem, the invention according to claim 4 is characterized in that the calibration photon detection means is constituted by a plurality of detectors, and the calibration photon information acquisition means corresponds to each of the detectors. The calibration photon detection position information (X, Y) and z value of the XY address conversion table creating means create a plurality of the XY address conversion tables corresponding to each detector, and the three-dimensional fixed length The array creating means creates a plurality of three-dimensional fixed-length arrays corresponding to the detectors, and the z address conversion table creating means creates a plurality of z address conversion tables corresponding to the detectors. The conversion means, based on the XY address conversion table created corresponding to each detector, the detection position information (X, Y) acquired for each detector to each detector. Do Address conversion to conversion position information (x, y) in the two-dimensional image, the determination means, the conversion position corresponding to each detector of the z address conversion table created corresponding to each detector The radiation according to claim 2 or 3, wherein it is determined whether or not the z value of each photon is within a predetermined effective range based on a value in the information (x, y). An image processing apparatus.

  In order to solve the above-mentioned problem, the invention according to claim 5 is characterized in that the subject is supported on a table, and a camera included in the detection means is supported rotatably around the table. And when the stage moves with the center of the rotation arc of the camera as the movement axis, the detection condition is at least a rotation angle of the camera when each photon is detected, and an axial position of the stage. The radiation image processing apparatus according to any one of claims 1 to 4, further comprising:

  In order to solve the above-mentioned problem, the invention according to claim 6 is that, when the camera repeatedly rotates around the base, the detection condition is a rotation angle of the camera when each photon is detected. The radiation image processing apparatus according to claim 5, comprising: at least one of an axial position of the table and a number of repetitions in the repeated rotation.

  In order to solve the above-described problem, in the invention according to claim 7, the detection condition is a detection time when each photon is detected. The radiation image processing apparatus according to the item.

  In order to solve the above-mentioned problem, the invention according to claim 8 is characterized in that the XY address conversion table creating means uses a predetermined method based on the photon output frequency in the XY two-dimensional space range indicated by the calibration photon detection position information. The extracted peak points are formed by connecting the center of gravity point determining means for determining the center of gravity of each quadrangle composed of the four adjacent peak points, and the four adjacent center of gravity points by a straight line. A center-of-gravity point rectangle generating means for generating a center-of-gravity point rectangle, two center points of any one of the center-of-gravity point rectangles, two adjacent center-of-gravity points, and a peak point located at the center. A quadrant quadrature generating unit that generates a quadrant quadrilateral by connecting four straight lines, and the address information determination unit is configured to output the image located at a vertex of the quadrant quadrangle in a predetermined direction. The address information included in the quadrant quadrangle is determined as the address information of the pixels included in the quadrant quadrangle, and the predetermined directions for all the quadrant quadrilaterals are the same direction. It is a radiographic image processing apparatus as described in any one of Claim 4 thru | or 7.

  In order to solve the above problem, the invention according to claim 9 is a radiographic image processing method using a detector that detects photons emitted from a subject and a radiographic image processing apparatus, wherein the image processing apparatus includes the detection A photon information acquisition step for acquiring detected position information (X, Y) in an XY two-dimensional space range of each photon detected by the detector, and a z value indicating an energy value of each photon; and Detection condition information acquisition step for acquiring detection conditions at the time of detection as detection condition information; for each photon, the detected position information (X, Y) and z value acquired in the photon information acquisition step; The detection condition information acquired in the detection condition information acquisition step is associated with the step of storing in the storage means in association with each of the detection position information (X, Y) and the z value. Based on the detection condition information, the step of classifying for each same detection condition, and each detection position information (X, Y) is defined by the number of pixels smaller than the number of pixels determined in the XY two-dimensional space range. Conversion position information (x, y) (x <X and y <Y) in a two-dimensional image, and the z value of each photon is based on the conversion position information (x, y). A determination step for determining whether or not the photon is within a predetermined effective range; and, as a result of the determination, when the z value of the photon is within a pre-defined effective range, For each of the converted position information, the photon output frequency at the position on the two-dimensional image indicated by the converted position information (x, y) corresponding to each photon is counted up regardless of the z value. Step to measure photon output frequency and before Generating two-dimensional image data indicating the two-dimensional image for each classification based on the photon output frequency corresponding to the converted position information (x, y), and a radiation image processing method, It is.

  In order to solve the above-mentioned problem, the invention according to claim 10 is a calibration photon detection step for detecting photons emitted by a subject for apparatus calibration, and XY twos included in each photon detected by the calibration photon detection step. Calibration photon information acquisition step for acquiring calibration photon detection position information (X, Y) in the dimensional space range and z value indicating the energy value of each photon, and each photon detected by the calibration photon detection step The photon output frequency in the XY two-dimensional space range indicated by the corresponding calibration photon detection position information (X, Y) is counted up regardless of the z value to correspond to the XY two-dimensional space range. A two-dimensional fixed length array is created and based on the photon output frequency of the two-dimensional fixed length array corresponding to the XY two-dimensional space range. An XY address conversion table creating step for creating an XY address conversion table for address conversion in the conversion step, and referring to the XY address conversion table, the calibration photon detection position information (X, Y) is Photon output frequency at the position on the two-dimensional image indicated by the converted position information (x, y) corresponding to each photon, for each photon converted into the converted position information (x, y) in the two-dimensional image As a three-dimensional fixed length array creating step for creating a three-dimensional fixed length array based on the converted position information (x, y) and the z value, and counting up for each z value based on the three-dimensional fixed length array A z-address conversion table creating step for creating a z-address conversion table for defining an effective range of z values at each position (x, y) on the two-dimensional image. In the conversion step, based on the XY address conversion table created in the XY address conversion table creation step, each pre-calibrated photon recording detection position information (X, Y) is converted into conversion position information ( x, y), and the determination step includes the z value of each photon based on the value in the conversion position information (x, y) of the z address conversion table created by the z address conversion table. 10. The radiation image processing method according to claim 9, wherein it is determined whether or not is within a predetermined effective range.

  In order to solve the above-mentioned problem, the invention according to claim 11 is directed to a computer, wherein the detected position information (X, Y) in the XY two-dimensional space range of each photon emitted from the subject and the energy value of each photon The photon information acquisition step of acquiring a z-value indicating from the detector, the detection condition information acquisition step of acquiring detection conditions when the photons are detected as detection condition information, and the photons, A step of storing the detected position information (X, Y) and z value acquired in the photon information acquisition step and the detection condition information acquired in the detection condition information acquisition step in association with each other in a storage unit. And classifying each detected position information and z value for each same detection condition based on the corresponding detection condition information, and each detected position information (X, Y), A conversion step of converting an address into conversion position information (x, y) (x <X and y <Y) in a two-dimensional image defined by a number of pixels smaller than the number of pixels determined in the Y two-dimensional space range; A determination step for determining whether the z value of each photon is within a predetermined effective range based on the converted position information (x, y), and as a result of the determination, the z value of the photon is determined in advance When within the defined effective range, for each photon, for each photon, the photon output frequency at the position on the two-dimensional image indicated by the converted position information (x, y) corresponding to each photon. Counting up regardless of the z value, measuring the photon output frequency for each conversion position information (x, y), and the photon output frequency corresponding to the conversion position information (x, y) On the basis of the two-dimensional image for each classification. A radiographic image processing program for causing to function as the step of generating the original image data.

  In order to solve the above-described problem, the invention according to claim 12 is the calibration photon detection position information (X, Y) in the XY two-dimensional space range of each photon emitted from the subject for apparatus calibration. A calibration photon information acquisition step of acquiring from the detector a z value indicating an energy value of each photon, and the corresponding calibration photon detection position information for each photon emitted from the subject for device calibration Counting up the photon output frequency in the XY two-dimensional space range indicated by (X, Y) regardless of the z value to create a two-dimensional fixed length array corresponding to the XY two-dimensional space range, XY address conversion table for address conversion in the conversion step based on the photon output frequency of the two-dimensional fixed length array corresponding to the XY two-dimensional space range XY address conversion table creation step for creating a map, and referring to the XY address conversion table, the calibration photon detection position information (X, Y) is converted into the conversion position information (x, y) in the two-dimensional image. For each photon, the photon output frequency at the position on the two-dimensional image indicated by the conversion position information (x, y) corresponding to each photon is counted up for each z value, and the conversion position information ( a three-dimensional fixed length array creating step for creating a three-dimensional fixed length array based on the x, y) and z values, and at each position (x, y) on the two-dimensional image based on the three-dimensional fixed length array. a z-address conversion table creating step for creating a z-address conversion table for defining an effective range of z-values, wherein the conversion step is based on the XY address conversion table creating step. Based on the generated XY address conversion table, each of the calibration photon detection position information (X, Y) is address-converted into conversion position information (x, y) in the two-dimensional image. Based on a value in the conversion position information (x, y) of the z address conversion table created by the z address conversion table, it is determined whether or not the z value of each photon is within a predetermined effective range. The radiation image processing program according to claim 11, wherein the radiation image processing program is configured to function as described above.

  According to the present invention, when a gamma camera device such as a SPECT device is applied to a middle animal or the like, each photon has a large capacity fixed for a large amount of photon event data having a large XY address space and energy resolution. Generate 2D image data by quickly processing a large amount of photon event data detected by a gamma camera device without requiring a long frame memory and without reducing resolution and quantification Can do.

It is the block diagram which showed roughly one Embodiment of the gamma camera. It is the schematic of the SPECT apparatus 1 in which the gamma camera is mounted. It is principal part explanatory drawing of the SPECT apparatus 1. FIG. (A) is a principal part top view of the SPECT apparatus 1, (B) is a principal part side view seen from the C direction of FIG. 3 (A). It is a block diagram which shows the structure of an image processing apparatus. 4 is a flowchart of an XY address conversion table and z address conversion table creation process executed by a control unit 21. 3 is a flowchart of two-dimensional image data generation processing executed by a control unit 21. It is explanatory drawing of an address 4 division process.

  Preferred embodiments of the present invention will be described with reference to the accompanying drawings. This embodiment is an example in which the radiation image processing apparatus of the present invention is configured by a gamma camera apparatus and an image processing apparatus.

  First, photons of gamma rays emitted from a subject such as a middle animal to which a radiopharmaceutical is administered are imaged (detected) with a detector equipped with a gamma camera. Then, the energy value (z value, resolution: 1024) of the detected photon and the positional information (X, Y) in the XY two-dimensional space range in the detector of the detected photon {resolution X: 4096 × Y: 4096 } Is detected, and the detected photon information is classified for each detection condition such as the rotation angle of the gamma camera. Then, the position information (X, Y) is addressed to the converted position information (x, y) in the two-dimensional image defined by a smaller number of pixels (for example, about 100 × 100, the number of pixels varies depending on the scintillator attached to the camera). Convert. Thereafter, based on the converted position information (x, y) and the energy value (z value) of each photon, only when the z value is valid, the same converted position information (x, y) regardless of the z value. The photon output frequency in y) is measured, and two-dimensional image data having the number of pixels {about 100 × 100} is generated for each classification based on the converted position information (x, y) and the photon output frequency.

<Configuration and Function of Gamma Camera Device and Image Processing Device>
FIG. 1 is a configuration diagram schematically showing an embodiment of a large-sized gamma camera for a medium animal.

  A pinhole collimator, a 88 × 88 two-dimensionally arranged scintillator, and a 32 × 32 two-dimensionally arranged photomultiplier tube are provided on the gamma ray incident surface side of the gamma camera head.

  FIG. 2 is a schematic diagram of a SPECT apparatus 1 on which a gamma camera is mounted.

  The SPECT apparatus 1 is an example of a gamma camera apparatus on which a gamma camera is mounted, and visualizes a tissue or the like of a subject with a radiopharmaceutical in the subject.

  The SPECT apparatus 1 includes a bed 101, a gantry 102, and detectors 103a and 103b.

  The bed 101 is a table for placing (supporting) a subject.

  Each of the detectors 103 a and 103 b includes a gamma camera and is installed in the ring-shaped gantry 102. In the following description, either or both of the detectors 103a and 103b may be collectively referred to as the detector 103.

  The gantry 102 is divided into a fixed portion and a movable portion in an axial direction passing through the center point of the gantry, and the movable portion is movable in a circumferential direction so as to slide on the fixed portion, and is attached to the movable portion. The detectors 103a and 103b are supported so as to be freely rotatable around the axis (see arrow R (n)). In general, in order to obtain a cross-sectional image by image reconstruction at the subsequent stage, it is required in the specification that the gantry 102 continuously rotates over a plurality of periods exceeding 360 degrees around the axis. However, in this embodiment, this function is replaced by repeated rotation. For example, after rotating from 0 degrees to 360 degrees on the forward path, it is rotated 0 degrees from 360 degrees on the forward path and returns to the position before the rotation. Further, this series of repeated rotation operations can be repeatedly executed.

  The bed 101 has a mechanism in which only a thin plate surface penetrating the gantry 102 is movable in the axial direction, and is divided into two lower portions of the bed 101 itself with the subject supported on the plate surface. It is movable along the axis indicated by the broken line in the figure so as to slide on the support member (see arrow B (n)).

  Although the bed 101 is configured to transmit gamma ray photons, transmission of gamma ray photons is somewhat blocked immediately below the bed 101, so the movable portion (thin plate surface) of the bed 101 holds the subject. Designed with a minimum width necessary for support. In particular, one of the movable parts of the bed 101 (for example, the side on which the head of the subject is placed) supports the subject with a pillow having a narrow width. Absorption of gamma rays generated in the process of being emitted from the organ radiation source to the outside of the body occurs not only in the bed 101 but also in the tissue of the subject, and the degree of absorption varies depending on the anatomical structure of the subject and the imaging angle. . Therefore, in the state immediately before the start of actual imaging (after the device calibration is completed and before the drug is administered to the subject), X-rays and other radiations are irradiated from outside the body in the same manner as CT, with the subject being placed on the bed 101, and the angle direction Generally, a process called absorption correction is performed on a captured image by actually measuring the radiation transmittance depending on the image. That is, the bed 101 is also handled as one of the tissues of the subject.

  FIG. 3 is an explanatory diagram of a main part of the SPECT apparatus 1. FIG. 3A is a plan view of the main part of the SPECT apparatus 1, and FIG. 3B is a side view of the main part viewed from the direction C of FIG. 3A.

  A repetitive parameter indicating the number of repetitive rotations of the detector 103 when the detection condition is set nth, where n is a time-series serial number assigned when the detection condition is set and n = 1 is a detection condition at the time of starting the apparatus. Let R (n). Here, the intention of performing repeated rotation is as follows. In order to obtain a single cross-sectional image, the acquired information only for the outward path is sufficient and it is not necessary to repeat the rotation. However, if it is only that, the total number of photons detected is too small and a highly sensitive image can be obtained. Therefore, more photons are detected by rotating about 2 to 8 cycles as necessary. However, in applications such as acquiring time-series images, the time resolution is reduced, so there are cases where repeated rotation is not performed, that is, R (n) = 1.

  Similarly, the rotation angle of the detector 103 when the nth detection condition is set is A (n). Further, a position (axial position) in the direction of the axis of the bed 101 when the nth detection condition is set is defined as B (n).

  As described above, the detector 103 is configured to be rotatable with respect to the axis, and the bed 101 is movable along the axis. Therefore, the detector 103 can detect photons emitted in all directions from the subject.

  When a photon is detected by the SPECT apparatus 1, RAW data of the photon and detection condition data are taken into an image processing apparatus described later together with detection time information. Photon RAW data is an example of detection position information in the present invention, and detection condition data is an example of detection condition information in the present invention.

  Photon RAW data includes detection time information, position information (X, Y) in an XY two-dimensional space range, and an energy value (z value). The detection time information is the number of seconds that have elapsed since the start time of the SPECT device 1. The position information (X, Y) is an XY coordinate value in a two-dimensional space range of {X: 4096 × Y: 4096} having a relatively large size for a middle animal or the like. The z value is an energy value within a range of maximum energy (for example, 1024).

  The detection condition is, for example, the relative installation position and the installation of the detector 103 with respect to the subject, the setting of which is changed by driving the movable part of the gantry 102 and the movable part of the bed 101 in time series according to the apparatus incorporation program. Is an angle. The relative installation position and installation angle of the detector 103 with respect to the subject are specifically the relative installation position and installation angle of the gamma camera provided in the detector 103 with respect to the subject. It is a value measured by a sensor. This value can be obtained as an actually measured value, or the installation position and the installation angle can be estimated from the stepping motor driving amount instructed to the movable part by the apparatus incorporation program. The repetitive parameter R (n) of the detector 103 is calculated based on the above-described apparatus built-in program and added as a detection condition.

  In this embodiment, each time a setting is changed, detection condition data is output, the initial state is set to n = 1, the value output at the nth time is indicated by (n), and the repetition parameter R ( n), an axial position B (n) of the bed 101 with respect to the subject of the detector 103, and a rotation angle A (n) of the rotation angle of the detector 103. Here, the detector 103 is repeatedly rotated around the bed 101, but the present invention can also be applied to a case where the detector 103 is rotated only in one direction. In this case, for example, a circulation parameter S (n) is set as a parameter indicating the number of rotations around the bed 101.

<Configuration and function of image processing apparatus>
FIG. 4 is a block diagram showing the configuration of the image processing apparatus according to the present invention.

  The image processing apparatus 2 includes a control unit 21, an external device connection unit 22, a storage unit 23, a display unit 24, and an input unit 25, and each component is connected to each other via a bus 26. ing.

  The control unit 21 includes a CPU and work RAM (not shown), a ROM for storing various control programs including the radiographic image processing program of the present invention, data, and the like, an oscillation circuit, and the like. Control information for controlling the member is generated, and the control information is output to the component to be controlled via the bus 26 to control the operation of each component.

  The external device connection unit 22 transmits various control signals to the SPECT device 1 based on the control of the control unit 21, and takes in the raw data and detection condition data transmitted from the SPECT device 1 into the device. The external device connection unit 23 sends a control signal to the SPECT apparatus 1 by, for example, a serial method, a USB method, IEEE1394, or other appropriate method.

  The storage unit 23 stores, for each detector 103, the RAW data and the detection condition data captured by the external device connection unit 22 inside the apparatus. In addition, an XY address conversion table and a z address conversion table are stored. The method for creating the XY address conversion table will be described in detail later.

  The display unit 24 displays an image or the like based on two-dimensional image data of a {two-dimensional fixed-length array of {about 100 × 100} obtained by address conversion.

  The input unit 25 instructs the display unit 24 to correct the address of the XY address conversion table, for example, or performs various input instructions to the SPECT device 1 via the image processing device 2.

  In addition, the control unit 21 executes the radiation image processing program stored in the ROM or the like, thereby cooperating with other components to detect the detection means, the photon information acquisition means, the detection condition information acquisition means, and the storage of the present invention. Means, classification means, conversion means, determination means, measurement means, generation means, calibration photon detection means, calibration photon information acquisition means, XY address conversion table acquisition means, three-dimensional fixed length array generation means, z address conversion table acquisition It is designed to function as a means.

  In the present embodiment, the display unit 24 is built in the image processing apparatus 2, but a color monitor or the like externally connected via the external device connection unit 22 may be used as the display device. In this case, the color monitor or the like is configured to send a display control instruction signal to the display device via a video card incorporated in the image processing apparatus 2, a VGA cable, a DVI cable, a BNC cable, or the like.

<Creation of XY address conversion table and z address conversion table>
First, based on RAW data, position information (X, Y) of XY two-dimensional space range {fixed to 4096 × 4096 due to detector specifications} is smaller in number of pixels (x, y) {scintillator mounted on detector XY address conversion table for converting the address into a two-dimensional image defined by a variable value} set based on the pixel pattern and a z address conversion table for extracting only z values within the effective range Create At this time, as shown in FIG. 2, when two detectors 103a and 103b are installed, it is necessary to create two types of XY address conversion tables and z address conversion tables corresponding to each detector.

  First, for calibration, a phantom that emits uniform gamma rays with respect to the field of view is placed on the bed 101 as a subject, and photons are detected from the subject by the detector 103.

  At this time, the gantry 102 is stationary, and the bed 101 is not moved. That is, the detector 103 performs imaging in a fixed and stationary state with respect to the subject. The repetition parameter R (n), the axial position B (n), and the rotation angle A (n) are constant. The raw data and detection condition data of each font detected by the detector 103 are taken into the storage unit 23 in the image processing apparatus 2 via the external device connection unit 23. Here, the time 1 is taken in time series. Normally, hundreds of millions of photons are detected, but in order to simplify the explanation, here, a case where five photons are detected by the detector 103a from time 2 to time 6 will be described as an example. As shown in FIGS. 2 and 3, when two detectors 103a and 103b are installed, the following RAW data is collected for each detector, and the following procedure is performed for each RAW data. Two types of XY address conversion tables and z address conversion tables are created independently.

<Example 1 of data in calibration (still imaging) mode>
Time 1: B (1), R (1), A (1)
Time 2: Detector 103a, X coordinate 1, Y coordinate 1, Z value 1
Time 3: Detector 103a, X coordinate 2, Y coordinate 2, Z value 2
Time 4: Detector 103a, X coordinate 3, Y coordinate 3, Z value 3
Time 5: Detector 103a, X coordinate 4, Y coordinate 4, Z value 4
Time 6: Detector 103a, X coordinate 5, Y coordinate 5, Z value 5

  The image processing device 2 receives detection condition data (B (1), R (1), A (1)) from the SPECT device 1 at time 1.

  RAW data based on photons detected by the detector 103a is received from the SPECT apparatus 1 at time 2, time 3, time 4. Strictly speaking, each time is the time when the image processing apparatus 2 receives the RAW data from the detector 103, but the time required for data exchange between the image processing apparatus 2 and the SPECT apparatus 1 is smaller than 1 second in the measurement time unit. Considering that the time is slight, the time may be used as the photon detection time by the detector 103.

  When the image processing apparatus 2 receives the RAW data and the detection condition data, the image processing apparatus 2 classifies and stores the RAW data in association with each same detection condition in the storage unit 23. In the case of <Static imaging mode data example 1> above, the detection conditions are all the same for the detectors 103a, B (1), R (1), and A (1), and thus all received RAW data is the same. It can be stored (classified) in the storage unit 23 as a frame. In addition, in this mode, classification by time is not performed, so that it is represented by the first photon detection time. An example of frame storage will be shown below.

<Frame storage example of data example 1 in calibration (still imaging) mode>
Directory information
Total number of frames: 1
Frame 1: Time 1, B (1), R (1), A (1), Data pointer 1
(Final): Time 6, B (1), R (1), A (1), data pointer 6
[RAW data]
Data pointer 1, X coordinate 1, Y coordinate 1, Z value 1
Data pointer 2, X coordinate 2, Y coordinate 2, Z value 2
Data pointer 3, X coordinate 3, Y coordinate 3, Z value 3
Data pointer 4, X coordinate 4, Y coordinate 4, Z value 4
Data pointer 5, X coordinate 5, Y coordinate 5, Z value 5

  Frame 1 uses RAW data from data pointer 1 to data pointer 5 immediately preceding (final) data pointer 6. Then, based on the X and Y coordinates of the RAW data of each photon, the photon appearance frequency in the XY two-dimensional space range of the number of pixels {4096 × 4096} is measured (counted up) regardless of the Z value, Create an XY address conversion table.

  In the data example 1 described above, the case where five photons are detected has been described, but actually, hundreds of millions of photons are detected. Accordingly, the photons having (X coordinate 1, Y coordinate 1) as detected position information include not only the photon of the data pointer 1 but also a large number of photons. In this way, for each coordinate (X, Y) of {4096 × 4096}, the frequency at which photons are output (detected) is measured (counted up) as the photon output frequency.

  Then, the photon output frequency is obtained for each coordinate (X, Y) of {4096 × 4096}, and a two-dimensional fixed length array is created. Then, based on the created two-dimensional fixed length array, after various processes such as differential filter processing, ridge line extraction processing, boundary line extraction processing, etc., an XY two-dimensional space range defined by the number of pixels (4096 × 4096) Then, an XY address conversion table to the coordinates (x, y) of the two-dimensional image defined by the number of pixels (for example, 100 × 100) smaller than the number of pixels (4096 × 4096) determined in this way is created. Note that specific methods for creating the XY address conversion table (various processing such as differential filter processing, ridge line extraction processing, and boundary line extraction processing) are described in detail in a patent application filed by the present applicant (Japanese Patent Laid-Open No. 2007-121259). Since it is disclosed, description thereof is omitted.

  Thereafter, referring to the XY address conversion table based on the RAW data of each photon, the photon output frequency is counted up for each z (value) for each coordinate (x, y) after conversion, and (x, y) and z Create a three-dimensional fixed-length array showing the frequency of photon output for each value.

  Then, based on the three-dimensional fixed length array, a z address conversion table for defining an effective range of z values for each coordinate (x, y) is created. A specific method for creating the z-address conversion table is disclosed in detail in a patent application filed by the applicant of the present application (Japanese Patent Laid-Open No. 2007-121259), and thus the description thereof is omitted.

  The above is the creation processing of the XY address conversion table and the z address conversion table executed by the control unit 21 of the image processing apparatus 2. FIG. 5 shows a flowchart of an XY address conversion table and z address conversion table creation process executed by the control unit 21. The processing is performed by the operator after the RAW data of each font detected by the detector 103 and the detection condition data are taken into the storage unit 23 in the image processing apparatus 2 via the external device connection unit 23. This is started when an instruction to start execution is given via the input unit 25.

  Since the control unit 21 only collects each RAW data fetched into the storage unit 23 into a single frame, the classification process in step S1 is not normally performed.

  Next, the control unit 21 measures (counts up) the photon appearance frequency in the XY two-dimensional space range of the number of pixels {4096 × 4096} based on the X coordinate and the Y coordinate of the RAW data of each photon to obtain an XY address. A conversion table is created (step S2). The created XY address conversion table is stored in the storage unit 23 by the control unit 21.

  Then, the control unit 21 refers to the XY address conversion table based on the RAW data of each photon, and performs a three-dimensional operation in which the photon output frequency is counted up for each z value for each coordinate (x, y) after conversion. A fixed-length array is created, and a z-address conversion table for defining an effective range of z-values for each coordinate (x, y) is created based on the created three-dimensional fixed-length array (step S3). The generated z address conversion table is stored in the storage unit 23 by the control unit 21. Then, the process ends.

<Generation of two-dimensional image data based on RAW data>
Subsequently, using the XY address conversion table and the z address conversion table created by the above-described method, based on the RAW data of photons emitted from a subject such as an actual middle animal, a smaller number of pixels {100 A procedure for generating a two-dimensional image defined by about × 100} will be described.

  First, a subject such as a middle animal into which a medicine labeled with a radioisotope that emits gamma rays in advance has been injected is placed on the bed 101, and photons are detected from the subject by the detector 103.

  The image processing apparatus 2 classifies the photon RAW data received from the SPECT apparatus 1 into a plurality of frames for each detection condition, and generates a plurality of two-dimensional image data based on each RAW data classified into each frame. .

  Here, a method of classifying the photon RAW data received from the SPECT apparatus 1 into a plurality of frames for each detection condition will be described.

<Bed movement mode>
In the bed movement mode, imaging is performed while moving the bed 101 on the axis while the gantry 102 is stationary. Therefore, the rotation angle A (n) and the repetition parameter R (n) are constant regardless of the value of n, and only the axial position B (n) varies. The raw data and detection condition data of each font detected by the detector 103a are taken into the storage unit 23 in the image processing apparatus 2 through the external device connection unit 23 in time series in order from time 1. The following is a case where the detector 103a is used. As shown in FIGS. 2 and 3, when two detectors 103a and 103b are installed, the following RAW data is provided for each detector. And two types of XY address conversion tables and z address conversion tables are created independently for each RAW data by the following procedure.

Time 1: B (1), R (1), A (1)
Time 2: Detector 103a, X coordinate 1, Y coordinate 1, Z value 1
Time 3: Detector 103a, X coordinate 2, Y coordinate 2, Z value 2
Time 4: Detector 103a, X coordinate 3, Y coordinate 3, Z value 3
Time 5: Detector 103a, X coordinate 4, Y coordinate 4, Z value 4
Time 6: B (2), R (1), A (1)
Time 7: Detector 103a, X coordinate 5, Y coordinate 5, Z value 5
Time 8: Detector 103a, X coordinate 6, Y coordinate 6, Z value 6
Time 9: Detector 103a, X coordinate 7, Y coordinate 7, Z value 7
Time 10: Detector 103a, X coordinate 8, Y coordinate 8, Z value 8

  In this example, the image processing apparatus 2 receives detection condition data (B (1), R (1), A (1)) from the SPECT apparatus 1 at time 1, and at times 2 to 5 RAW data is received at time 6, the detection condition data (B (2), R (1), A (1)) is received again at time 6, and RAW data is received at time 7-10. is doing. It may be more appropriate that the detection condition data (B (2), R (2), A (2)) is received at time 6, but R (2) = R (1). And as described above to emphasize that A (2) = A (1).

  That is, the RAW data received at times 2 to 5 is data captured under the detection conditions (B (1), R (1), A (1)), and at times 7 to 10. The received RAW data is data captured under detection conditions (B (2), R (1), A (1)).

  Therefore, the image processing apparatus 2 classifies each received data in the storage unit 23 in association with each same detection condition as follows. That is, the RAW data received at times 2 to 5 is stored in (classified) as frame 1 and the RAW data received at times 7 to 10 is stored as frame 2. An example of frame storage will be shown below.

<Example of frame storage in bed movement mode>
Directory information
Total number of frames: 2
Frame 1: Time 1, B (1), R (1), A (1), Data pointer 1
Frame 2: Time 6, B (2), R (1), A (1), Data pointer 5
(Final): Time 10, B (2), R (1), A (1), data pointer 9
[RAW data]
Data pointer 1, X coordinate 1, Y coordinate 1, Z value 1
Data pointer 2, X coordinate 2, Y coordinate 2, Z value 2
Data pointer 3, X coordinate 3, Y coordinate 3, Z value 3
Data pointer 4, X coordinate 4, Y coordinate 4, Z value 4
Data pointer 5, X coordinate 5, Y coordinate 5, Z value 5
Data pointer 6, X coordinate 6, Y coordinate 6, Z value 6
Data pointer 7, X coordinate 7, Y coordinate 7, Z value 7
Data pointer 8, X coordinate 8, Y coordinate 8, Z value 8

  Frame 1 uses RAW data from data pointer 1 to data pointer 4 immediately preceding data pointer 5 of frame 2. The frame 2 uses RAW data from the data pointer 5 to the data pointer 8 immediately before the (final) data pointer 9.

<Intermittent rotation mode>
In the intermittent rotation mode, the gantry 102 is repeatedly rotated by a specified period at an angle of a predetermined interval. The angle at the predetermined interval is, for example, 0 degrees and 90 degrees, and still imaging is performed at each rotation angle. The bed 101 is moved by a predetermined interval after a series of repetitive rotation operations is completed and the same processing is continued. In the following description, the bed 101 is assumed to be at the initial position B (1). To do. Accordingly, the rotation angle A (n) and the repetition parameter R (n) change, and the axial position B (n) is constant. The following is a case where the detector 103a is used. As shown in FIGS. 2 and 3, when two detectors 103a and 103b are installed, the following RAW data is provided for each detector. And two types of XY address conversion tables and z address conversion tables are created independently for each RAW data by the following procedure.

Time 1: B (1), R (1), A (1)
Time 2: Detector 103a, X coordinate 1, Y coordinate 1, Z value 1
Time 3: Detector 103a, X coordinate 2, Y coordinate 2, Z value 2
Time 4: B (1), R (1), A (2)
Time 5: Detector 103a, X coordinate 3, Y coordinate 3, Z value 3
Time 6: Detector 103a, X coordinate 4, Y coordinate 4, Z value 4
Time 7: B (1), R (2), A (2)
Time 8: Detector 103a, X coordinate 5, Y coordinate 5, Z value 5
Time 9: Detector 103a, X coordinate 6, Y coordinate 6, Z value 6
Time 10: B (1), R (2), A (1)
Time 11: Detector 103a, X coordinate 7, Y coordinate 7, Z value 7
Time 12: Detector 103a, X coordinate 8, Y coordinate 8, Z value 8

  In this example, the image processing apparatus 2 receives the detection condition data (B (1), R (1), A (1)) from the SPECT apparatus 1 at time 1, and at times 2 and 3. RAW data is received at time 4, detection condition data (B (1), R (1), A (2)) is received at time 4, and RAW data is received at time 5 and 6. , The detection condition data (B (1), R (2), A (2)) is received at time 7, RAW data is received at time 8 and 9, and at time 10, The detection condition data (B (1), R (2), A (1)) is received, and RAW data is received at time 11 and 12. At time 4, it may be more appropriate to receive detection condition data (B (2), R (2), A (2)), but B (2) = B (1 ) And R (2) = R (1), so as to emphasize that (B (1), R (1), A (2)). Similarly, at time 7, it may be more appropriate that the detection condition data (B (3), R (3), A (3)) is received, but B (3) = B ( 1) and A (3) = A (2), so use R (2) unused instead of R (3), and use (B (1), R (2), A (2)) It was described. Similarly, at time 10, it may be more appropriate to receive detection condition data (B (4), R (4), A (4)), but B (4) = B ( Since 1), R (4) = R (2), and A (4) = A (1), they were described as (B (1), R (2), A (1)).

  The raw data received at times 2 and 3 are data captured under the detection conditions (B (1), R (1), A (1)). For example, the rotation angle A (1) is “0 degree”, and the repetition parameter R (1) is “1” (outward in the first cycle).

  The RAW data received at time 5 and 6 is data captured under the detection conditions (B (1), R (1), A (2)). For example, the rotation angle A (2) is “90 degrees”.

  The RAW data received at times 8 and 9 is data captured under the detection conditions (B (1), R (2), A (2)). For example, the repetition parameter R (2) is “2” (return path in the first cycle).

  The raw data received at times 11 and 12 are data captured under the detection conditions (B (1), R (2), A (1)).

  Therefore, the image processing apparatus 2 classifies each received data in the storage unit 23 in association with each same detection condition as follows. That is, RAW data received at times 2 and 3 is frame 1, RAW data received at times 5 and 6 is frame 2, and RAW data received at times 7 and 8 is frame 3. RAW data received at times 11 and 12 is stored (classified) in the storage unit 23 as a frame 4. An example of frame storage will be shown below.

<Frame storage example in intermittent rotation mode>
Directory information
Total number of frames: 4
Frame 1: Time 1, B (1), R (1), A (1), Data pointer 1
Frame 2: Time 4, B (1), R (1), A (2), Data pointer 3
Frame 3: Time 7, B (1), R (2), A (2), data pointer 5
Frame 4: Time 10, B (1), R (2), A (1), Data pointer 7
(Final): Time 12, B (1), R (2), A (1), data pointer 9
[RAW data]
Data pointer 1, X coordinate 1, Y coordinate 1, Z value 1
Data pointer 2, X coordinate 2, Y coordinate 2, Z value 2
Data pointer 3, X coordinate 3, Y coordinate 3, Z value 3
Data pointer 4, X coordinate 4, Y coordinate 4, Z value 4
Data pointer 5, X coordinate 5, Y coordinate 5, Z value 5
Data pointer 6, X coordinate 6, Y coordinate 6, Z value 6
Data pointer 7, X coordinate 7, Y coordinate 7, Z value 7
Data pointer 8, X coordinate 8, Y coordinate 8, Z value 8

  Frame 1 uses RAW data from data pointer 1 to data pointer 2 immediately preceding data pointer 3 of frame 2. The frame 2 uses RAW data from the data pointer 3 to the data pointer 4 immediately before the data pointer 5 of the frame 3. The frame 3 uses RAW data from the data pointer 5 to the data pointer 6 immediately before the data pointer 7 of the frame 4. The frame 4 uses RAW data from the data pointer 7 to the data pointer 8 immediately before the (final) data pointer 9.

  As described above, as described in the “bed movement mode” and the “intermittent rotation mode”, the image processing apparatus 2 classifies the RAW data detected from the SPECT apparatus 1 according to the same detection condition and associates them with each frame. It is stored in the storage unit 23. Further, as shown in FIGS. 2 and 3, when two detectors 103a and 103b are installed, two sets of RAW data classified for each detector are stored.

<Generation of two-dimensional image data>
Next, generation of two-dimensional image data will be described.

  First, the X coordinate and the Y coordinate (position coordinates in the XY two-dimensional space range of the number of pixels {4096 × 4096}) of the RAW data classified into an arbitrary frame correspond to the detector that has received the RAW data. Using the XY address conversion table, address conversion is performed to conversion position information in a two-dimensional image defined by the number of pixels {about 100 × 100}. The coordinates after conversion of (X coordinate, Y coordinate) are represented as (x coordinate, y coordinate).

  In the above-described example, there are several pieces of RAW data classified into an arbitrary frame for the sake of simplicity, but in reality, hundreds of millions of RAW data are classified. Therefore, for example, there are many RAW data whose conversion position information after address conversion is (x coordinate 1, y coordinate 1). Thus, for each coordinate (x, y) of {about 100 × 100}, the frequency at which photons are output (detected) is measured (counted up) as the photon output frequency.

  At this time, referring to the z address conversion table corresponding to the detector that has received the RAW data using the conversion position information (x, y), if the z value of the RAW data is outside the valid range, the photon Do not measure (count up) as output frequency. This is because the photon has undergone some scattering, and the acquired position coordinates (X, Y) are assumed to be unreliable data with errors. In this way, two-dimensional image data is generated for each frame. Furthermore, as shown in FIGS. 2 and 3, when two detectors 103a and 103b are installed, two sets of two-dimensional image data are generated for each detector.

  The above is the two-dimensional image data generation process executed by the control unit 21 of the image processing apparatus 2.

  FIG. 6 shows a flowchart of the two-dimensional image data generation process executed by the control unit 21. The processing is performed by the operator after the RAW data of each font detected by the detector 103 and the detection condition data are taken into the storage unit 23 in the image processing apparatus 2 via the external device connection unit 23. This is started when an instruction to start execution is given via the input unit 25.

  The control unit 21 classifies each RAW data captured in the storage unit 23 in association with each same detection condition (step S11).

  Next, the control unit 21 refers to the XY address conversion table and the z address conversion table corresponding to the detector that has received the RAW data stored in the storage unit 23 based on the RAW data of each photon, and performs address conversion. Is performed (step S12).

  Then, the control unit 21 generates two-dimensional image data for each frame based on the photon output frequency at each conversion position coordinate (x, y) (step S13). The control unit 21 displays an image based on the generated two-dimensional image data on the display unit 24 and ends the process.

  In the example described above, an example in which a subject such as a middle animal is imaged using only the detector 103a has been described. However, when both the detectors 103a and 103b are used, the image processing apparatus 2 Raw data is detected for each detector 103. Then, two-dimensional image data is generated based on the RAW data detected while referring to the XY address conversion table and the z address conversion table corresponding to each detector 103.

  As described above, the image processing apparatus 2 according to the present embodiment uses the raw information (X, Y) in the XY two-dimensional space range of {4096 × 4096} included in the photons emitted from the subject such as a middle animal to the RAW. The data and the detection condition data are received from the detector 103 of the SPECT apparatus 1, and the RAW data is classified for each detection condition with respect to the RAW data collected for each detector. Then, the address information is converted to a smaller number of pixels {about 100 × 100}. Thereafter, based on the converted position information (x, y) and the z value, only when the z value is valid, the photon output frequency in the corresponding converted position information (x, y) is measured, and the converted position information (x The two-dimensional image data indicating the photon output frequency based on y) is generated for each detector and each classification.

  Thus, for example, when the SPECT apparatus 1 is applied to a middle animal or the like, {4096 × 4096 × for RAW data including {4096 × 4096} XY two-dimensional spatial range and {1024} -stage energy information. A large amount of RAW data detected by the SPECT apparatus 1 can be quickly processed without requiring a large amount of frame memory of a fixed length of 1024} and without reducing resolution and quantification. Can be generated.

<Application example 1>
Uniformity correction processing may be performed on the generated two-dimensional image data.

  In this case, a phantom that emits uniform gamma rays with respect to the visual field is placed on the bed 101 as a subject, and photons from the subject are detected by the detector 103. Then, the image processing apparatus 2 calculates a uniform correction coefficient (average pixel value / each pixel value) based on the detected RAW data, and a uniform correction coefficient table {100 such that the values of all the pixels are the same. X100} is created. The created uniformization correction coefficient table is stored in the storage unit 23. Furthermore, as shown in FIGS. 2 and 3, when two detectors 103a and 103b are installed, two sets of such uniformization correction coefficient tables are created for each detector and stored in the storage unit 23. Keep it.

  Then, in step S13, the control unit 21 corresponds to the pixel (x, y) of the uniformization correction coefficient table for the value (photon frequency) of each pixel (x, y) of each generated two-dimensional image data. The uniformization correction process is executed by multiplying the value (correction coefficient).

<Application example 2>
When the SPECT apparatus 1 is applied to a medium animal or the like, it is preferable to increase the number of pixels of the scintillator of the gamma camera as the subject size increases. However, there is a limit in fine processing to increase the number of pixels of the scintillator. Therefore, when creating the XY address conversion table, the resolution can be improved in a pseudo manner by increasing the number of pixels in a pseudo manner.

  Here, an address four-division process for dividing each coordinate into four divisions when creating the XY address conversion table will be described.

  First, after differentiating the image data based on the two-dimensional fixed length array, a predetermined number of peak points are determined by a predetermined method based on the photon output frequency. For example, a predetermined number of peak pixels are extracted in descending order of photon output frequency, and a peak point is determined based on the extracted peak pixels. A specific technique for peak point extraction based on the photon output frequency is disclosed in detail in a patent application filed by the applicant of the present application (Japanese Patent Laid-Open No. 2007-121259), and will not be described.

  FIG. 7 is an explanatory diagram of the address four-division process.

  The control unit 21 determines the barycentric point of a peak point quadrangle composed of four adjacent peak points (indicated by Δ in FIGS. 7A to 7C). At this time, a virtual barycentric point is determined for each of the peripheral portions of the upper end, lower end, left end, and right end of the image data. As shown in FIG. 7A, the virtual centroid point is connected to a quadrangle (centroid point rectangle) generated by the straight line when connected to the centroid point located at the periphery of the image data by a straight line. A location where a peak point located in the peripheral portion of the gradation image data is included is determined as a virtual center of gravity (shown by ▲ in FIGS. 7A to 7C). Next, the control unit 21 connects four adjacent barycentric points with straight lines, and generates a barycentric quadrangle formed by the straight lines (see FIG. 7B).

  Then, the control unit 21 is located at the center of two midpoints (shown by ☆ in FIG. 7C) between any one of the barycentric quadrangular points and two neighboring barycentric points. The peak points are connected by four straight lines to generate a quadrant quadrangle (see FIG. 7C).

  Subsequently, the control unit 21 assigns and determines the address values of the vertices in a predetermined direction of the quadrant quadrilateral, and based on this, determines the XY address conversion table including the address split areas as shown in FIG. create. In the case of the example shown in FIG. 7D, the address value of the vertex in the upper left direction shown by (2) is assigned as the address information of each pixel included in each quadrant.

  When a quadrant is generated, as shown in FIG. 7C, the operator may operate the input unit 25 so that the positions of the center of gravity and the midpoint that define the quadrant can be corrected. . Note that such interactive correction by the operator may be performed in the same manner as the method disclosed in detail in the patent application (Japanese Patent Laid-Open No. 2007-121259) by the applicant of the present application.

<Application example 3>
When the SPECT apparatus 1 is applied to a middle animal or the like, the z-value energy resolution increases as the subject size increases, and when creating a z-address conversion table for defining the effective range of the z-value, The probability that an error will occur in the valid range of values increases. For this reason, it is preferable that the operator can specify (correct) the effective range of the z value.

  The correction may be performed in any way. For example, the position coordinate of the correction target on the image data is designated on the display unit 24 and a graph of the z value at the designated position coordinate is displayed. While referring to the graph, the input unit 25 is operated to specify (correct) the effective range of the z value. The graph display of the z value and the like are disclosed in detail in a patent application filed by the applicant of the present application (Japanese Patent Laid-Open No. 2007-121259), and thus the description thereof is omitted.

<Modification of frame storage in intermittent rotation mode>
In the case of storing the frame in the “intermittent rotation mode” described above, the RAW data is classified into four frames for each of the four types of detection conditions. However, the RAW data is stored in one frame for each same rotation angle A (n). Also good. In other words, changes in the repetition parameter R (n) are not considered. Then, instead of reducing the total number of frames, the photon frequency of each frame increases and the sensitivity of the image improves. An example of frame storage will be shown below. Although the original RAW data is arranged in time series, in the following example, it is sorted by angle, and the original data pointer 7 and the data pointer 8 are moved up to the data pointer 3 and the data pointer 4.

Directory information
Total number of frames: 2
Frame 1: Time 1, B (1), R (1), A (1), Data pointer 1
Frame 2: Time 4, B (1), R (1), A (2), data pointer 5
(Final): Time 12, B (1), R (2), A (1), data pointer 9
[RAW data]
Data pointer 1, X coordinate 1, Y coordinate 1, Z value 1
Data pointer 2, X coordinate 2, Y coordinate 2, Z value 2
Data pointer 3, X coordinate 7, Y coordinate 7, Z value 7
Data pointer 4, X coordinate 8, Y coordinate 8, Z value 8
Data pointer 5, X coordinate 3, Y coordinate 3, Z value 3
Data pointer 6, X coordinate 4, Y coordinate 4, Z value 4
Data pointer 7, X coordinate 5, Y coordinate 5, Z value 5
Data pointer 8, X coordinate 6, Y coordinate 6, Z value 6

  Frame 1 uses RAW data from data pointer 1 to data pointer 4 immediately preceding data pointer 5 of frame 2. The frame 2 uses RAW data from the data pointer 5 to the data pointer 8 immediately before the (final) data pointer 9.

<Other frame storage examples>
In the above-described embodiment, the frame storage example in the “bed movement mode” and the “intermittent rotation mode” has been described, but other frame storage modifications will be described.

<Continuous rotation mode>
In the continuous rotation mode, the gantry 102 is repeatedly rotated by a designated period continuously (for example, every 0.1 degrees). Moving images are taken while moving the detector 103 without stopping the rotation. The bed 101 is moved by a predetermined interval after a series of repetitive rotation operations, and the same processing is continued. In the following description, the bed 101 is assumed to be at the initial position B (1). To do. Accordingly, the rotation angle A (n) and the repetition parameter R (n) change, and the axial position B (n) is constant.

Time 1: B (1), R (1), A (1)
Time 2: Detector 103a, X coordinate 1, Y coordinate 1, Z value 1
Time 3: B (1), R (1), A (2)
Time 4: Detector 103a, X coordinate 2, Y coordinate 2, Z value 2
Time 5: B (1), R (1), A (3)
Time 6: Detector 103a, X coordinate 3, Y coordinate 3, Z value 3
Time 7: B (1), R (1), A (4)
Time 8: Detector 103a, X coordinate 4, Y coordinate 4, Z value 4
Time 9: B (1), R (2), A (4)
Time 10: Detector 103a, X coordinate 5, Y coordinate 5, Z value 5
Time 11: B (1), R (2), A (3)
Time 12: Detector 103a, X coordinate 6, Y coordinate 6, Z value 6
Time 13: B (1), R (2), A (2)
Time 14: Detector 103a, X coordinate 7, Y coordinate 7, Z value 7
Time 15: B (1), R (2), A (1)
Time 16: Detector 103a, X coordinate 8, Y coordinate 8, Z value 8

  In this example, the image processing apparatus 2 receives detection condition data (B (1), R (1), A (1)) from the SPECT apparatus 1 at time 1, and SPECT at time 2. RAW data (detector 103a, X coordinate 1, Y coordinate 1, Z value 1) from the detector 103a is received from the apparatus 1, and detection condition data (B (1), R (1)) at time 3 , A (2)), and RAW data (detector 103a, X coordinate 2, Y coordinate 2, Z value 2) from the detector 103a is received from the SPECT device 1 at time 4, and thereafter Up to 16 detection condition data and RAW data are received.

  In the intermittent rotation mode, for example, still images are captured while changing the rotation angle A (n) in units of 10 degrees, whereas in the continuous rotation mode, the rotation angle A (n) is changed in units of 0.1 degrees, for example. While taking a picture dynamically. In the “bed movement mode” and “intermittent rotation mode” described above, the RAW data is classified for each detection condition. However, if the detection condition changes frequently as in the continuous rotation mode, the RAW data is generated based on the detection condition. It is not practical to classify the data into different frames.

  Therefore, in the continuous rotation mode, RAW data having an angle range (for example, 10 degrees) similar to the angular interval of the rotation angle A (n) for each detection condition in the “intermittent rotation mode” is grouped into the same frame. Take the way.

  Here, the RAW data detected at the rotation angle A (1) and the RAW data detected at the rotation angle A (2) are classified into the same frame, and at the rotation angle A (3). The detected RAW data and the RAW data detected at the rotation angle A (4) are classified into the same frame. An example of frame storage will be shown below.

<Frame storage example in continuous rotation mode>
Directory information
Total number of frames: 4
Frame 1: Time 1, B (1), R (1), A (1), Data pointer 1
Frame 2: Time 5, B (1), R (1), A (3), Data pointer 3
Frame 3: Time 9, B (1), R (2), A (4), Data pointer 5
Frame 4: Time 13, B (1), R (2), A (2), Data pointer 7
(Final): Time 16, B (1), R (2), A (1), data pointer 9
[RAW data]
Data pointer 1, X coordinate 1, Y coordinate 1, Z value 1
Data pointer 2, X coordinate 2, Y coordinate 2, Z value 2
Data pointer 3, X coordinate 3, Y coordinate 3, Z value 3
Data pointer 4, X coordinate 4, Y coordinate 4, Z value 4
Data pointer 5, X coordinate 5, Y coordinate 5, Z value 5
Data pointer 6, X coordinate 6, Y coordinate 6, Z value 6
Data pointer 7, X coordinate 7, Y coordinate 7, Z value 7
Data pointer 8, X coordinate 8, Y coordinate 8, Z value 8

  Frame 1 uses RAW data from data pointer 1 to data pointer 2 immediately preceding data pointer 3 of frame 2. The frame 2 uses RAW data from the data pointer 3 to the data pointer 4 immediately before the data pointer 5 of the frame 3. The frame 3 uses RAW data from the data pointer 5 to the data pointer 6 immediately before the data pointer 7 of the frame 4. The frame 4 uses RAW data from the data pointer 7 to the data pointer 8 immediately before the (final) data pointer 9.

  Frames 1 and 2 are RAW data detected in the forward path (repetition parameter R (1)), and frames 3 and 4 are RAW data detected in the return path (repetition parameter R (2)).

<Variation 1 of frame storage in continuous rotation mode>
In the case of frame storage in the “continuous rotation mode” described above, the change in the repetition parameter R (n) is not taken into consideration, and the forward path (repetition parameter R (1)) at the rotation angles A (1) and A (2). Return path (repetition parameter R (2)) at rotation angles A (2) and A (1), forward path (repetition parameter R (1)) and rotation angle at rotation angles A (3) and A (4) You may classify | categorize into two frames of the return path (repetition parameter R (2)) in A (4) and A (3). Then, instead of reducing the total number of frames, the photon frequency of each frame increases and the sensitivity of the image improves. An example of frame storage will be shown below. Although the original RAW data is arranged in time series, in the following example, it is sorted by angle, and the original data pointer 7 and the data pointer 8 are moved up to the data pointer 3 and the data pointer 4.

Directory information
Total number of frames: 2
Frame 1: Time 1, B (1), R (1), A (1), Data pointer 1
Frame 2: Time 5, B (1), R (1), A (3), data pointer 5
(Final): Time 16, B (1), R (2), A (1), data pointer 9
[RAW data]
Data pointer 1, X coordinate 1, Y coordinate 1, Z value 1
Data pointer 2, X coordinate 2, Y coordinate 2, Z value 2
Data pointer 3, X coordinate 7, Y coordinate 7, Z value 7
Data pointer 4, X coordinate 8, Y coordinate 8, Z value 8
Data pointer 5, X coordinate 3, Y coordinate 3, Z value 3
Data pointer 6, X coordinate 4, Y coordinate 4, Z value 4
Data pointer 7, X coordinate 5, Y coordinate 5, Z value 5
Data pointer 8, X coordinate 6, Y coordinate 6, Z value 6

  Frame 1 uses RAW data from data pointer 1 to data pointer 4 immediately preceding data pointer 5 of frame 2. The frame 2 uses RAW data from the data pointer 5 to the data pointer 8 immediately before the (final) data pointer 9.

<Modification 2 of frame storage in continuous rotation mode>
In the “continuous rotation mode”, when the gantry 102 is continuously rotating at a constant speed, time may be set as the detection condition. That is, the image processing apparatus 2 classifies each RAW data at a predetermined time interval T based on the detection time.

  For example, detection is performed when time 1 to time 6 is less than time T, time 7 to time 12 is greater than or equal to time T and less than time 2T, and time 13 to time 16 is greater than or equal to time 2T and less than time 3T. The generated RAW data is classified into RAW data detected at times 2, 4, and 6, RAW data detected at times 8, 10, and 12, and RAW data detected at times 14 and 16. Examples of frame storage will be shown below.

Directory information
Total number of frames: 3
Frame 1: Time 1, B (1), R (1), A (1), Data pointer 1
Frame 2: Time 7, B (1), R (1), A (4), data pointer 4
Frame 3: Time 13, B (1), R (2), A (2), Data pointer 7
(Final): Time 16, B (1), R (2), A (2), data pointer 9
[RAW data]
Data pointer 1, X coordinate 1, Y coordinate 1, Z value 1
Data pointer 2, X coordinate 2, Y coordinate 2, Z value 2
Data pointer 3, X coordinate 3, Y coordinate 3, Z value 3
Data pointer 4, X coordinate 4, Y coordinate 4, Z value 4
Data pointer 5, X coordinate 5, Y coordinate 5, Z value 5
Data pointer 6, X coordinate 6, Y coordinate 6, Z value 6
Data pointer 7, X coordinate 7, Y coordinate 7, Z value 7
Data pointer 8, X coordinate 8, Y coordinate 8, Z value 8

  In the present embodiment, each application example, and each modification example, an example in which a middle animal or the like is used as an object has been described. Since the resolution is also improved, the present invention may be applied to a case where the subject is a human head or a small animal if desired, as long as it can be stored in the gantry.

1 Image processing device
11 Control Unit 12 Storage Unit 13 External Device Connection Unit 14 Display Unit 15 Input Unit 16 Bus

Claims (12)

Detection means for detecting photons emitted by the subject;
Photon information acquisition means for acquiring detected position information (X, Y) in an XY two-dimensional space range of each photon detected by the detection means, and acquiring the energy value of each photon as a z value;
Detection condition information acquisition means for acquiring detection conditions when each photon is detected as detection condition information;
For each photon, the detected position information (X, Y) and z value acquired by the photon information acquisition means and the detection condition information acquired by the detection condition information acquisition means are associated with each other in the storage means. Storage control means for storing;
Classification means for classifying each detection position information (X, Y) and z value for each same detection condition based on the corresponding detection condition information,
Each of the detected position information (X, Y) is converted position information (x, y) (x <X and in a two-dimensional image defined by the number of pixels smaller than the number of pixels defined in the XY two-dimensional space range a conversion means for converting the address to y <Y);
Determination means for determining whether the z value of each photon is within a predetermined effective range based on the converted position information (x, y);
As a result of the determination, when the z value of the photon is within a pre-defined effective range, for each photon classified by the classifying means, the converted position information (x , y) counting means that counts up the photon output frequency at the position on the two-dimensional image regardless of the z value, and measures the photon output frequency for each converted position information (x, y); ,
Based on the photon output frequency corresponding to the converted position information (x, y), generating means for generating two-dimensional image data indicating the two-dimensional image for each classification by the classification means;
A radiation image processing apparatus comprising: Calibration photon detection means for detecting photons emitted by the subject for device calibration;
Calibration photon information for acquiring calibration photon detection position information (X, Y) in the XY two-dimensional space range of each photon detected by the calibration photon detection means, and acquiring the energy value of each photon as a z value Acquisition means;
For each photon detected by the calibration photon detection means, the photon output frequency in the XY two-dimensional space range indicated by the corresponding calibration photon detection position information (X, Y) is counted up regardless of the z value. Then, a two-dimensional fixed length array corresponding to the XY two-dimensional space range is created, and the conversion means performs address conversion based on the photon output frequency of the two-dimensional fixed length array corresponding to the XY two-dimensional space range. XY address conversion table creating means for creating an XY address conversion table when
With reference to the XY address conversion table, the calibration photon detection position information (X, Y) is converted into the conversion position information (x, y) in the two-dimensional image, and each photon corresponds to each photon. Counting up each z value as the photon output frequency at the position on the two-dimensional image indicated by the converted position information (x, y), and fixing the three-dimensional value based on the converted position information (x, y) and the z value A three-dimensional fixed length array creating means for creating a long array;
Z address conversion table creating means for creating a z address conversion table for defining an effective range of z values at each position (x, y) on the two-dimensional image based on the three-dimensional fixed-length array; Have
The conversion means converts the calibration photon detection position information (X, Y) into the conversion position information (x, y in the two-dimensional image) based on the XY address conversion table created by the XY address conversion table creation means. )
The determination means determines that the z value of each photon is within a predetermined effective range based on the value in the conversion position information (x, y) of the z address conversion table created by the z address conversion table. The radiological image processing apparatus according to claim 1, wherein it is determined whether or not. The detection means includes a plurality of detectors, and the photon information acquisition means and the detection condition information acquisition means correspond to the plurality of detectors, and a plurality of sets of detection position information (X, Y) and z values, and Get detection condition information
The storage control means associates and stores the detection position information (X, Y) and z value of each detector, and detection condition information in the storage means,
The classifying means classifies each set for each same detection condition based on the plurality of sets of detection position information (X, Y) and z value and detection condition information acquired corresponding to each detector. ,
The conversion means converts the detected position information (X, Y) using each XY address conversion table corresponding to the detector from which the detected position information (X, Y) is acquired,
The determination means is a z address conversion table corresponding to the detector from which the z value of each photon is obtained, and z at each position (x, y) on the two-dimensional image. Based on each z address conversion table for defining an effective range of values, it is determined whether the z value of each photon is within a predetermined effective range,
The measurement means measures the photon output frequency at positions on a plurality of two-dimensional images corresponding to the plurality of detectors independently as the photon output frequency,
The radiation image processing apparatus according to claim 1, wherein the generation unit generates a plurality of the two-dimensional image data for each of the plurality of detectors and for each classification by the classification unit. The calibration photon detection means is composed of a plurality of detectors,
The calibration photon information acquisition means acquires a plurality of sets of calibration photon detection position information (X, Y) and z values corresponding to the detectors,
The XY address conversion table creating means creates a plurality of the XY address conversion tables corresponding to the detectors,
The three-dimensional fixed length array creating means creates a plurality of three-dimensional fixed length arrays corresponding to the detectors,
The z address conversion table creating means creates a plurality of z address conversion tables corresponding to the detectors,
The conversion means converts each detection position information (X, Y) acquired for each detector into each detector based on an XY address conversion table created corresponding to each detector. Address conversion to conversion position information (x, y) in the two-dimensional image,
The determination means determines the z of each photon based on a value in the conversion position information (x, y) corresponding to each detector of the z address conversion table created corresponding to each detector. The radiographic image processing apparatus according to claim 2 or 3, wherein it is determined whether or not the value is within a predetermined effective range. The object is supported on a table, and the camera included in the detection unit is supported so as to be rotatable around the table, and the table moves on the center of the rotation arc of the camera. When moving as
5. The detection condition according to claim 1, wherein the detection condition includes at least one of a rotation angle of the camera when each photon is detected and an axial position of the table. The radiation image processing apparatus described.   When the camera repeatedly rotates around the table, the detection conditions include: a rotation angle of the camera when each photon is detected, an axial position of the table, and the number of repetitions in the repeated rotation. The radiographic image processing apparatus according to claim 5, comprising at least one of them.   The radiographic image processing apparatus according to claim 1, wherein the detection condition is a detection time when each photon is detected. The XY address conversion table creating means includes:
Peak points extracted by a predetermined method based on the photon output frequency in the XY two-dimensional space range indicated by the calibration photon detection position information, each of which is determined as a quadratic barycentric point composed of four adjacent peak points. Centroid point determination means to
A center-of-gravity point quadrangle generating unit that connects four adjacent center-of-gravity points with a straight line and generates a center-of-gravity point quadrangle formed by the straight line;
A quadrant quadrangle is generated by connecting two midpoints of any one of the barycentric quadrangles, two adjacent barycentric points, and a peak point located in the center with four straight lines 4 A divided rectangle generating means;
Have
The address information determination means determines the address information of the pixel located at the apex of the quadrant quadrangle in a predetermined direction as the address information of the pixel included in the quadrant quadrilateral, The radiographic image processing apparatus according to claim 2, wherein the predetermined directions related to the divided quadrangles are the same direction. A radiographic image processing method using a detector and a radiographic image processing device for detecting photons emitted from a subject,
The image processing apparatus is
Photon information acquisition step of acquiring detection position information (X, Y) in the XY two-dimensional space range of each photon detected by the detector, and a z value indicating the energy value of each photon;
A detection condition information acquisition step for acquiring detection conditions when each photon is detected as detection condition information;
For each photon, the detected position information (X, Y) and z value acquired in the photon information acquisition step and the detection condition information acquired in the detection condition information acquisition step are stored in association with each other. Memorizing the means;
Classifying each detection position information (X, Y) and z value for each same detection condition based on the corresponding detection condition information;
Each of the detected position information (X, Y) is converted position information (x, y) (x <X and in a two-dimensional image defined by the number of pixels smaller than the number of pixels defined in the XY two-dimensional space range a conversion step for converting the address to y <Y);
A determination step of determining whether the z value of each photon is within a predetermined effective range based on the converted position information (x, y);
As a result of the determination, when the z value of the photon is within a pre-defined effective range, the conversion position information (x, y) corresponding to each photon is obtained for each photon for each classification. Counting up the photon output frequency at a position on the two-dimensional image to indicate regardless of the z value, and measuring the photon output frequency for each of the converted position information;
Generating two-dimensional image data indicating the two-dimensional image for each classification based on the photon output frequency corresponding to the converted position information (x, y);
A radiation image processing method comprising: A calibration photon detection step for detecting photons emitted by a subject for device calibration;
Calibration photon information acquisition for acquiring calibration photon detection position information (X, Y) in the XY two-dimensional space range of each photon detected by the calibration photon detection step and a z value indicating the energy value of each photon Steps,
For each photon detected in the calibration photon detection step, the photon output frequency in the XY two-dimensional space range indicated by the corresponding calibration photon detection position information (X, Y) is counted up regardless of the z value. Then, a two-dimensional fixed length array corresponding to the XY two-dimensional space range is created, and an address is generated in the conversion step based on the photon output frequency of the two-dimensional fixed length array corresponding to the XY two-dimensional space range. An XY address conversion table creation step for creating an XY address conversion table for conversion;
With reference to the XY address conversion table, the calibration photon detection position information (X, Y) is converted into the conversion position information (x, y) in the two-dimensional image, and each photon corresponds to each photon. Counting up each z value as the photon output frequency at the position on the two-dimensional image indicated by the converted position information (x, y), and fixing the three-dimensional value based on the converted position information (x, y) and the z value A three-dimensional fixed-length array creation step for creating a long array;
A z-address conversion table creating step for creating a z-address conversion table for defining an effective range of z-values at each position (x, y) on the two-dimensional image based on the three-dimensional fixed-length array; Have
In the conversion step, based on the XY address conversion table created in the XY address conversion table creation step, each pre-calibration photon recording detection position information (X, Y) is converted into conversion position information (x, Y) in the two-dimensional image. y)
In the determination step, the z value of each photon is within a predetermined effective range based on a value in the conversion position information (x, y) of the z address conversion table created by the z address conversion table. The radiographic image processing method according to claim 9, wherein it is determined whether or not. Computer
The photon information acquisition step of acquiring detection position information (X, Y) in the XY two-dimensional space range of each photon emitted from the subject and a z value indicating the energy value of each photon from the detector;
The detection condition information acquisition step of acquiring detection conditions when each photon is detected as detection condition information;
For each photon, the detected position information (X, Y) and z value acquired in the photon information acquisition step and the detection condition information acquired in the detection condition information acquisition step are stored in association with each other. Memorizing the means;
Classifying each detection position information and z value for each same detection condition based on the corresponding detection condition information;
Each of the detected position information (X, Y) is converted position information (x, y) (x <X and in a two-dimensional image defined by the number of pixels smaller than the number of pixels defined in the XY two-dimensional space range a conversion step for converting the address to y <Y);
A determination step of determining whether the z value of each photon is within a predetermined effective range based on the converted position information (x, y);
As a result of the determination, when the z value of the photon is within a pre-defined effective range, the conversion position information (x, y) corresponding to each photon is obtained for each photon for each classification. Counting up the photon output frequency at the position on the two-dimensional image to indicate regardless of the z value, and measuring the photon output frequency for each converted position information (x, y);
A radiographic image processing program that functions as a step of generating two-dimensional image data indicating the two-dimensional image for each classification based on the photon output frequency corresponding to the converted position information (x, y). . The computer,
Calibration photon detection position information (X, Y) in the XY two-dimensional space range of each photon emitted from the subject for apparatus calibration and a z value indicating the energy value of each photon are acquired from the detector. Calibration photon information acquisition step;
For each photon emitted from the apparatus calibration subject, the photon output frequency in the XY two-dimensional space range indicated by the corresponding calibration photon detection position information (X, Y) is counted regardless of the z value. To create a two-dimensional fixed-length array corresponding to the XY two-dimensional space range, and based on the photon output frequency of the two-dimensional fixed-length array corresponding to the XY two-dimensional space range, in the conversion step An XY address conversion table creation step for creating an XY address conversion table for address conversion;
With reference to the XY address conversion table, the calibration photon detection position information (X, Y) is converted into the conversion position information (x, y) in the two-dimensional image, and each photon corresponds to each photon. Counting up each z value as the photon output frequency at the position on the two-dimensional image indicated by the converted position information (x, y), and fixing the three-dimensional value based on the converted position information (x, y) and the z value A three-dimensional fixed-length array creation step for creating a long array;
A z-address conversion table creating step for creating a z-address conversion table for defining an effective range of z-values at each position (x, y) on the two-dimensional image based on the three-dimensional fixed-length array; Have
In the conversion step, the calibration photon detection position information (X, Y) is converted into the conversion position information (x, y in the two-dimensional image) based on the XY address conversion table created in the XY address conversion table creation step. )
In the determination step, the z value of each photon is within a predetermined effective range based on a value in the conversion position information (x, y) of the z address conversion table created by the z address conversion table. The radiographic image processing program according to claim 11, wherein the program is made to function to determine whether or not.

Images