WO2016035706A1 - 画像化装置及び方法 - Google Patents
画像化装置及び方法 Download PDFInfo
- Publication number
- WO2016035706A1 WO2016035706A1 PCT/JP2015/074427 JP2015074427W WO2016035706A1 WO 2016035706 A1 WO2016035706 A1 WO 2016035706A1 JP 2015074427 W JP2015074427 W JP 2015074427W WO 2016035706 A1 WO2016035706 A1 WO 2016035706A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- event
- pixel
- detector
- compton
- compton scattering
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01T—MEASUREMENT OF NUCLEAR OR X-RADIATION
- G01T1/00—Measuring X-radiation, gamma radiation, corpuscular radiation, or cosmic radiation
- G01T1/29—Measurement performed on radiation beams, e.g. position or section of the beam; Measurement of spatial distribution of radiation
- G01T1/2914—Measurement of spatial distribution of radiation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01T—MEASUREMENT OF NUCLEAR OR X-RADIATION
- G01T1/00—Measuring X-radiation, gamma radiation, corpuscular radiation, or cosmic radiation
- G01T1/16—Measuring radiation intensity
- G01T1/161—Applications in the field of nuclear medicine, e.g. in vivo counting
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01T—MEASUREMENT OF NUCLEAR OR X-RADIATION
- G01T1/00—Measuring X-radiation, gamma radiation, corpuscular radiation, or cosmic radiation
- G01T1/16—Measuring radiation intensity
- G01T1/161—Applications in the field of nuclear medicine, e.g. in vivo counting
- G01T1/164—Scintigraphy
- G01T1/1641—Static instruments for imaging the distribution of radioactivity in one or two dimensions using one or several scintillating elements; Radio-isotope cameras
- G01T1/1647—Processing of scintigraphic data
Definitions
- the present invention relates to an imaging apparatus and method.
- a method using a Compton camera has been proposed as a method for imaging the spatial distribution of a gamma ray source (see, for example, Patent Document 1 below).
- the Compton camera has a wider imaging field of view and a wider detection energy band than the existing nuclear medicine technologies positron emission tomography (PET) and single photon emission computed tomography (SPECT).
- PET positron emission tomography
- SPECT single photon emission computed tomography
- the Compton camera is not only an improvement of the existing technology, but also enables simultaneous multi-molecule imaging that simultaneously images a wide variety of radionuclides, which has been difficult in the past. This is a technique capable of noninvasively visualizing the behavior of a plurality of radiodiagnostic drugs and biofunctional molecules, and can acquire information on life functions that are more advanced than conventional techniques. Therefore, the Compton camera is expected to contribute to early diagnosis in life science research and clinical medicine.
- a typical Compton camera includes two radiation detectors capable of measuring an interaction position and energy as a front-stage detector and a rear-stage detector.
- An imaging device using a Compton camera pays attention to a gamma ray detection event in which a gamma ray from a gamma ray source is Compton scattered by a front stage detector and then photoelectrically absorbed by a back stage detector.
- the interaction information that is, the detection position and detection energy of the gamma ray at each detector
- the spatial distribution of the gamma ray source is imaged.
- a three-dimensional image showing the spatial distribution of the gamma ray source is called a distribution image.
- the pixel value ⁇ j of the pixel j of the distribution image is obtained according to the following equation (1) (see, for example, Non-Patent Document 1 below).
- the distribution of the estimated gamma ray source is brought closer to the true distribution by updating the pixel value by iterative calculation.
- ⁇ j (l) and ⁇ j (l + 1) indicate pixel values of the pixel j obtained by the 1st and (l + 1) th iterative calculations, respectively.
- s j is a detection sensitivity parameter for the pixel j, and is composed of the geometric efficiency of the previous detector viewed from the pixel j. Division by the detection sensitivity parameter s j means correction of the pixel value in consideration of detection sensitivity. The significance of other parameters in equation (1) will be described in detail later.
- the detection sensitivity is determined depending only on the pixel j.
- the response function changes greatly for each event, so it is considered desirable to optimize the correction of detection sensitivity for each event.
- the correction of the pixel value with the appropriate detection sensitivity contributes to the generation of an appropriate distribution image.
- an object of the present invention is to provide an imaging apparatus and method that contribute to the optimization of a distribution image showing the spatial distribution of a gamma ray source.
- An imaging apparatus includes a post-stage detector, a pre-stage detector disposed between a gamma ray source and the post-stage detector, and gamma rays that are Compton scattered by the pre-stage detector.
- An event detection means for detecting an event photoelectrically absorbed by the detector, and a distribution image showing a spatial distribution of the gamma ray source based on the measurement data of the interaction between each detector and the gamma ray in a plurality of events.
- Computing means for generating an image in an image space containing the image, and the computing means includes a probability parameter indicating a probability that Compton-scattered gamma rays have arrived from the image space in each event. Individually set for each of a plurality of pixels in the distribution image based on the measurement data of the event, and using the set probability parameter And generating the distribution image.
- the probability that the gamma ray has arrived from within the image space is Since it can be said that it is relatively low, the measurement data of the event should not be strongly reflected in the distribution image. And the state of the said intersection changes for every event. Therefore, a probability parameter indicating the probability may be set for each event. This reduces artifacts. Further, if an appropriate probability parameter is set for each pixel using the interaction measurement data, further reduction of artifacts is expected and generation of an appropriate distribution image is expected.
- An imaging apparatus includes a post-stage detector, a pre-stage detector disposed between a gamma ray source and the post-stage detector, and gamma rays Compton scattered by the pre-stage detector.
- An event detection means for detecting an event photoelectrically absorbed by the detector, and an operation for generating a distribution image showing a spatial distribution of the gamma ray source based on measurement data of interaction between each detector and gamma rays in a plurality of events.
- An imaging device comprising: means for calculating a detection sensitivity parameter indicating Compton scattering detection sensitivity at each event based on the measurement data of each event for each of the plurality of pixels in the distribution image. The distribution image is generated individually using the set detection sensitivity parameter.
- An imaging method is a method used for a Compton camera including a post-stage detector and a pre-stage detector disposed between a gamma ray source and the post-stage detector. Based on the measurement data of the interaction between each detector and the gamma ray at each event in which the gamma rays that are Compton scattered by the detector are photoelectrically absorbed by the subsequent detector, a distribution image showing the spatial distribution of the gamma ray source is obtained.
- a probability parameter indicating a probability that Compton-scattered gamma rays have arrived from the image space in each event is set based on the measurement data of each event.
- Each of the plurality of pixels is individually set, and the distribution image is generated using the set probability parameter.
- An imaging method is a method used in a Compton camera including a rear-stage detector and a front-stage detector disposed between a gamma-ray source and the rear-stage detector. Generates a distribution image showing the spatial distribution of the gamma ray source based on the measurement data of the interaction between each gamma ray and each detector at each event in which the gamma rays scattered by the Compton scattering at the detector are photoelectrically absorbed by the latter detector.
- the detection sensitivity parameter indicating the detection sensitivity of Compton scattering at each event is individually set and set for each of the plurality of pixels in the distribution image based on the measurement data at each event.
- the distribution image is generated using the detection sensitivity parameter.
- an imaging apparatus and method that contribute to the optimization of a distribution image showing the spatial distribution of a gamma ray source.
- FIG. 4 is a diagram showing how measurement data of a plurality of events are obtained. These are the relationship diagrams between each detector and the image space.
- A) And (b) is a figure which shows a mode that an image space is divided
- FIG. 6 is a diagram illustrating an example of a state of intersection between a Compton conical surface and an image space of two events.
- (A) And (b) is a figure which shows the example of the intersection state of the Compton conical surface area
- (A) And (b) is a figure which shows the example of the intersection state of the Compton conical surface area
- (A) And (b) is a comparison figure of a reference
- FIG. 11 is a flowchart illustrating a distribution image generation operation based on measurement data of each event. These are the schematic top views of the cancer bearing mouse
- reference numerals 11 and 12 denote a front-stage detector and a rear-stage detector that form the Compton camera 10, respectively.
- Each of the detectors 11 and 12 is a radiation detector capable of measuring an interaction position and energy, and is formed of a semiconductor or a light emitting substance (scintillator).
- Reference numeral 201 denotes a gamma ray source that emits gamma rays.
- Reference numeral 200 denotes an imaging target including the gamma ray source 201.
- the detectors 11 and 12 are typically arranged in parallel and spaced from each other. When viewed from the imaging target 200 that includes the gamma ray source 201, the front detector 11 and the rear detector 12 are arranged in this order. That is, the front detector 11 is disposed between the imaging target 200 including the gamma ray source 201 and the rear detector 12. In FIG. 1B, only one gamma ray source 201 is shown in the imaging target 200, but many gamma ray sources 201 can be scattered in the imaging target 200.
- a gamma ray from the gamma ray source 201 is incident on the front-stage detector 11 and Compton scattered at a position R 1 in the front-stage detector 11, and the scattered gamma ray is incident on the rear-stage detector 12 and is positioned in the rear-stage detector 12. and it is photoelectrically absorbed by R 2.
- Compton scattering and photoelectric absorption are a kind of interaction between gamma rays and a detector. Further, it is assumed that all energy of gamma rays after Compton scattering is absorbed at a position R 2 in the post-stage detector 12. At this time, the energy detected by the detectors 11 and 12 is represented by E 1 and E 2, respectively.
- the energy E 1 is the energy given to the electron at the position R 1 by the gamma rays from the gamma ray source 201 (that is, the energy lost by the Compton scattering in the energy of the gamma rays from the gamma ray source 201).
- the energy E 2 is the total energy of gamma rays after Compton scattering, and is the photoelectric absorption energy at the position R 2 . Then, the following equation (A1) is established according to the kinematics of Compton scattering.
- ⁇ C is the Compton scattering angle
- me is the static mass of electrons
- c is the speed of light in vacuum
- E 0 is the initial energy of gamma rays.
- the Compton scattering angle ⁇ C is an angle formed by the flight direction (traveling direction) of gamma rays before Compton scattering and the flight direction (traveling direction) of gamma rays after Compton scattering.
- the initial energy E 0 of gamma rays is the initial energy of gamma rays emitted from the gamma ray source 201 and may be known to the gamma ray source distribution imaging apparatus 1 described later.
- a parameter is known to the imaging apparatus 1 means that the imaging apparatus 1 recognizes the value of the parameter in advance.
- an antibody labeled with a predetermined gamma-ray emitting nucleus is administered to a living body as the imaging target 200.
- the value of the energy of gamma rays emitted from the gamma ray emission nucleus as the gamma ray source 201 may be recognized in advance by the gamma ray source distribution imaging apparatus 1.
- the gamma ray source 201 may be a positron emitting nucleus.
- the gamma rays accompanying the annihilation of positrons emitted from the nuclei, which are positron emitting nuclei, by positron decay become the gamma rays from the gamma ray source 201.
- the imaging device 1 creates a gamma-ray energy spectrum during imaging by the Compton camera 10. It is possible to recognize the initial energy E 0 of the gamma ray from its energy spectrum.
- the conical surface 202 has the Compton scattering position R 1 as the apex, the Compton scattering angle ⁇ C as the half apex angle (half apex angle; open angle), and on the straight line of the flight path of gamma rays after Compton scattering. (That is, on a straight line passing through the positions R 1 and R 2 ) is a conical surface having a central axis (hereinafter also referred to as a Compton conical surface).
- a position R 1 be a conical surface (i.e.
- An event in which a set of “R 1 , R 2 , E 1 and E 2 ” is detected is called an event. That is, in one event, one set of “R 1 , R 2 , E 1 and E 2 ” is detected by one-time Compton scattering and total energy absorption of gamma rays after Compton scattering. Two Compton conical surfaces are required. In one event, we can only know that there is a gamma ray source somewhere on the drawn Compton cone, but we measure a lot of Compton scattering phenomena, and multiple Compton cones for multiple events are displayed in the image space. Drawing above suggests that there are gamma ray sources at the intersection of many Compton cones.
- FIG. 2 is a perspective view showing the configuration of the Compton camera 10.
- the Compton conical surface 211C is set for the gamma ray 211 detected in one event
- the Compton conical surface 212C is set for the gamma ray 212 detected in another event.
- a similar Compton conical surface is additionally set. Since the position where more Compton conical surfaces overlap corresponds to the position where many gamma rays are generated, for example, by setting the pixel value of each pixel in the image space according to the multiplicity of overlap of the Compton conical surfaces, The state of the source distribution can be imaged.
- Each of the detectors 11 and 12 is a radiation detector capable of measuring the interaction position between itself and the gamma ray and the energy of the gamma ray.
- movement of the detectors 11 and 12 are mainly demonstrated for the example of the structure in which each of the detectors 11 and 12 is formed using the semiconductor flat plate (planar type detector) which consists of high purity germanium.
- the one surface and the other surface of the semiconductor flat plate are referred to as a front surface and a rear surface, respectively.
- the normal direction of each of the front surface and the rear surface coincides with the thickness direction of the semiconductor flat plate.
- the front surface is located closer to the imaging target 200 including the gamma ray source 201 than the rear surface.
- the front and rear surfaces of the front detector 11 refer to the front and rear surfaces of the semiconductor flat plate forming the front detector 11, and the front and rear surfaces of the rear detector 12 are the front and rear surfaces of the semiconductor flat plate forming the rear detector 12. Point to.
- a plurality of strip electrodes arranged in a stripe shape are arranged on each of the front surface and the rear surface.
- the X axis, Y axis, and Z axis that are orthogonal to each other are defined.
- a plurality of strip electrodes extending in the Y-axis direction are arranged along the X-axis direction on the front surface
- a plurality of strip electrodes extending in the X-axis direction are arranged on the rear surface on the Y-axis. It is arranged along the direction.
- the strip electrodes adjacent to each other are electrically insulated through spacing. The same applies to the post-stage detector 12.
- each detector 11 and 12 on the front side uppermost n + layer of semiconductor plates uppermost rear side are in the p + layer, a region sandwiched between the n + layer and p + layer For example, a high-purity p-type semiconductor is used.
- each detector is a detector using germanium, each detector is cooled to a liquid nitrogen temperature ( ⁇ 196 ° C.), for example. For this reason, in each semiconductor flat plate of the detectors 11 and 12, supply of conduction carriers due to thermal excitation of electrons from the valence band is suppressed.
- the detectors 11 and 12 may be formed of a semiconductor that does not require cooling for measurement.
- a reverse bias voltage is applied to each strip electrode on both sides so as to generate an electric field in the direction of the thickness of the semiconductor flat plate.
- a region called a depletion layer in which supply of conductive carriers from the impurity level is suppressed is formed in the semiconductor, and the semiconductor is in a high resistance state.
- the gamma rays enter the depletion layer, the gamma rays interact with electrons in the semiconductor, and these electrons receive energy from the gamma rays.
- a large number of carrier charges are generated by excitation of electrons from the valence band along the path of the electrons that have received energy, and the generated carrier charges are extracted by the reverse bias voltage.
- the current due to the carrier charge has a current value reflecting the energy of the electrons interacting with the gamma rays, and the charge amount obtained by integrating the current is proportional to the energy. For this reason, when Compton scattering occurs in the detector 11 as the above interaction, the detector 11 detects a current corresponding to the energy E 1 from which the gamma rays have been lost due to Compton scattering.
- the energy E 1 is specified by the amount of charge corresponding to the integral value of the current detected by the detector 11.
- gamma rays after Compton scattering are absorbed by the detector 12, a photoelectric effect occurs as the above interaction, and a current corresponding to the total energy of gamma rays after Compton scattering is detected by the detector 12.
- the energy E 2 is specified by the amount of charge corresponding to the integral value of the current detected by the detector 12.
- an interaction position is detected by specifying a combination of front and rear strip electrodes where a current signal due to the interaction between the gamma ray and the detector is strongly detected.
- the interaction position detected by the upstream detector 11 is the Compton scattering position R 1
- the interaction position detected by the downstream detector 12 is the photoelectric absorption position R 2 .
- the block (x A , y A ) when a current signal is detected from the x A th strip electrode on the front surface and a current signal is detected from the y A th strip electrode on the rear surface, the block (x It can be detected that an interaction with the gamma ray has occurred at ( A , y A ) (the same applies to the detector 12).
- the position of the block (x A , y A ) In the X axis direction, the position of the block (x A , y A ) is the same as the position of the x A th strip electrode on the front surface, and in the Y axis direction, the position of the block (x A , y A ) is The position is the same as the position of the y A th strip electrode.
- x A is an arbitrary natural number equal to or less than the number of strip electrodes provided on the front surface
- y A is an arbitrary natural number equal to or less than the number of strip electrodes provided on the rear surface.
- the number of strip electrodes arranged on each surface of the front detector 11 is arbitrary (the same applies to the rear detector 12).
- the detection resolution of the interaction position in the X-axis direction depends on the number of strip electrodes arranged on the front surface
- the detection resolution of the interaction position in the Y-axis direction depends on the number of strip electrodes arranged on the rear surface.
- the Z-axis component of the interaction position detected in each detector may be constant (center position of each detector in the Z-axis direction). However, in each of the detectors 11 and 12, the interaction position in the Z-axis direction (Z-axis component of the interaction position) is based on the time difference between the current signal from the front strip electrode and the current signal from the rear strip electrode. May be derived with high accuracy.
- the semiconductor forming the detectors 11 and 12 may be a semiconductor other than germanium as long as it exhibits sensitivity to gamma rays.
- silicon, cadmium telluride, cadmium zinc telluride, or diamond may be used as the semiconductor forming the detectors 11 and 12.
- each of the detectors 11 and 12 may be any radiation detector that can measure the interaction position between itself and the gamma rays and the energy of the gamma rays. You may have such a structure.
- each of the detectors 11 and 12 may be formed by a scintillation detector or a TPC (time projection chamber) using gas or liquid.
- FIG. 3 is a schematic overall configuration diagram of the gamma ray source distribution imaging apparatus 1 according to the present embodiment.
- the imaging apparatus 1 includes respective parts referred to by reference numerals 20, 30 and 41 to 44. However, it can be considered that the Compton camera 10 is not included in the components of the imaging apparatus 1.
- the upstream detector 11 When an interaction with a gamma ray occurs inside the upstream detector 11, the upstream detector 11 outputs a signal 11 OUT corresponding to the content of the interaction.
- the downstream detector 12 interacts with gamma rays inside itself, the downstream detector 12 outputs a signal 12 OUT corresponding to the content of the interaction.
- the interaction position (for example, position R 1 ) with the gamma ray in the detector 11 by the signal 11 OUT and the interaction energy (for example, energy E 1) given to the detector 11 from the gamma ray by the interaction in the detector 11. ) Is identified.
- the interaction position (for example, position R 2 ) with the gamma ray in the detector 12 by the signal 12 OUT and the interaction energy (for example, energy E 2) given from the gamma ray to the detector 12 by the interaction in the detector 12. ) Is identified.
- the event measurement data acquisition unit 20 includes a simultaneous event extraction unit 21 and a total energy absorption determination unit 22, and event measurement data that is interaction data between the detectors 11 and 12 and gamma rays in each event is used. Generate and get.
- FIG. 4 is a flowchart for acquiring event measurement data by the acquisition unit 20.
- the simultaneous event extraction unit 21 outputs corresponding to events that occur simultaneously in the detectors 11 and 12 out of the output signal 11 OUT of the upstream detector 11 and the output signal 12 OUT of the downstream detector 12.
- a set of signals is extracted as a simultaneous event detection signal.
- the simultaneous event extracting unit 21 each time from the previous stage detector 11 signal 11 OUT is output, stored in the storage unit 44 a signal 11 OUT after having granted a time stamp indicating its output time to the signal 11 OUT with previously, each time from the subsequent detector 12 signal 12 OUT is outputted, allowed to store the signal 12 OUT after having granted a time stamp indicating its output time to the signal 12 OUT in the storage unit 44.
- the simultaneous event extracting unit 21 reads the arbitrary signal 11 OUT stored in the storage unit 44, and stores the signal 12 OUT whose time difference from the read signal 11 OUT is within a predetermined allowable time difference. Search from within 44 (step S11).
- step S11 When the signal 12 OUT whose time difference from the read signal 11 OUT is within a predetermined allowable time difference exists in the storage unit 44 (Y in step S11), the simultaneous detection of the read signal 11 OUT and the searched signal 12 OUT is detected. It extracts as a signal (step S12).
- Step S15 When the signal 12 OUT whose time difference from the read signal 11 OUT is within the predetermined allowable time difference does not exist in the storage unit 44 (N in Step S11), the process proceeds to Step S15.
- the second signal whose time difference from the first signal is within a predetermined allowable time difference is associated with the time stamp indicating a time whose time difference from the time indicated by the time stamp corresponding to the first signal is within the predetermined allowable time difference. The second signal is indicated.
- the simultaneous event extraction unit 21 extracts the simultaneous event detection signal in real time from the output signals 11 OUT and 12 OUT of the detectors 11 and 12 without the storage of the signals 11 OUT and 12 OUT by the storage unit 44.
- step S13 the total energy absorption determination unit 22 performs a total energy absorption determination process on the extracted simultaneous event detection signal. In the total energy absorption determination process, it is determined whether the signal 12 OUT in the simultaneous event detection signal indicates total energy absorption. Specifically, the total energy absorption determination unit 22 determines the simultaneous event determination formula “E 0 ⁇ E 0 ⁇ E for the interaction energies E 1 and E 2 indicated by the signals 11 OUT and 12 OUT in the simultaneous event detection signal.
- ⁇ E 0 is introduced into the simultaneous event determination formula.
- ⁇ E 0 has a positive predetermined value. For example, when E 0 is 511 keV (kiloelectron volts), ⁇ E 0 is set to 5 keV.
- the total energy absorption determination unit 22 determines that the signal 12 OUT in the time event detection signal does not indicate total energy absorption (N in step S13), and discards the simultaneous event detection signal. Then, the process proceeds to step S15. This is because the Compton scattering angle cannot be correctly estimated from the measurement data if the simultaneous event determination formula does not hold even if Compton scattering occurs.
- step S14 the event measurement data acquisition unit 20 recognizes that the extracted simultaneous event detection signal represents a detection signal in a valid event, and event measurement data based on the signals 11 OUT and 12 OUT in the valid event. (Interaction measurement data) is acquired, and the acquired event measurement data is stored in the storage unit 44. After step S14, the process proceeds to step S15.
- One event measurement data includes information identifying interaction positions R 1 and R 2 and energies E 1 and E 2 in one event.
- the acquisition unit 20 detects an event using the extraction unit 21 and the determination unit 22 and extracts event measurement data from the output signals 11 OUT and 12 OUT of the detectors 11 and 12.
- step S15 the event measurement data acquisition unit 20 determines whether or not a predetermined measurement data acquisition end condition is met. If the measurement data acquisition end condition is satisfied, the event measurement data acquisition process ends. If the measurement data acquisition end condition is not satisfied, the process returns to step S11. For example, the measurement data acquisition end condition is satisfied when the number of event measurement data acquisition reaches a predetermined number (for example, several 10000). Alternatively, for example, when the process related to the extraction of the simultaneous event detection signal is completed for all the signals 11 OUT stored in the storage unit 44, the measurement data acquisition end condition is satisfied.
- a predetermined number for example, several 10000.
- the main arithmetic unit 30 includes a highly parallel calculation accelerator such as a CPU (Central Processing Unit) and a SIMD (single instruction multiple data) arithmetic unit.
- the main calculation unit 30 sets a predetermined image space in the space including the imaging target 200, and based on event measurement data generated for a plurality of events, a distribution image (gamma ray source distribution image) indicating the density distribution of the gamma ray source As an image in the image space. A method for generating the distribution image will be described later.
- the main calculation unit 30 may be formed using a dedicated calculation device for generating a reconstructed image including a distribution image.
- the dedicated arithmetic device is a device having a hardware configuration optimum for image reconstruction, which is constructed using, for example, ASIC (application specific integrated circuit) or FPGA (field-programmable gate array).
- the overall control unit 41 is realized by using a CPU (Central Processing Unit) or the like, and comprehensively controls the operation of each part in the imaging apparatus 1.
- the display unit 42 includes a liquid crystal display panel and the like, and displays an arbitrary image including a distribution image under the control of the overall control unit 41.
- the operation unit 43 includes a pointing device, a keyboard, and the like, and accepts arbitrary instructions and operations from the user of the imaging apparatus 1.
- the storage unit 44 includes a ROM (Read Only Memory) and a RAM (Random Access Memory).
- the storage unit 44 stores arbitrary programs as well as programs to be operated by the main arithmetic unit 30, the overall control unit 41, and the like.
- the storage unit 44 may include a secondary storage device such as an HDD (Hard disk drive) or a flash memory.
- FIG. 5 shows how event measurement data for a plurality of events are acquired one after another.
- the event measurement data in the i-th event (that is, the i-th event) is also referred to as i-th event measurement data.
- i is an arbitrary natural number.
- the detected energies E 1 and E 2 at the detectors 11 and 12 in the measurement data of the i-th event are represented by E i1 and E i2, respectively.
- the interaction position at the detector 11 ie, the detected Compton scattering position
- R 1 and the interaction position (ie, the detected photoelectric absorption position) R 2 at the detector 12 in the measurement data of the i-th event In particular, R i1 and R i2 respectively.
- the main calculation unit 30 calculates the Compton scattering angle ⁇ C and the Compton conical surface using the above formula (A1) based on the event measurement data (E i1 , R i1 , E i2 , R i2 ) in the event. Can be derived and set.
- FIG. 6 shows an image space IS set and defined by the main arithmetic unit 30.
- the image space IS may be set in a dedicated memory device that can be included in the storage unit 44 or the like.
- the image space IS is a three-dimensional virtual space set in association with the real space, and includes at least the position where the imaging target 200 exists.
- the image space IS has a finite size.
- the shape of the image space IS may be arbitrary, but here, the image space IS is a rectangular parallelepiped space having two planes parallel to the XY plane, the YZ plane, and the ZX plane.
- the XY plane is a plane parallel to the X axis and the Y axis
- the YZ plane is a plane parallel to the Y axis and the Z axis
- the ZX plane is a plane parallel to the Z axis and the X axis.
- one of the planes in the image space IS parallel to the XY plane is adjacent to the front surface of the detector 11, and the center of the detector 11, the center of the detector 12, and the center of the image space IS are aligned. Deploy.
- the main calculation unit 30 can generate a distribution image representing the density distribution state of the gamma ray source as a three-dimensional image in the image space IS.
- FIG. 7A shows a state where the image space IS is equally divided into a plurality of parts in each of the X, Y, and Z axis directions.
- the image space IS is formed from element areas equally divided into a plurality of parts in each of the X, Y, and Z axis directions, and each element area functions as a pixel of a distribution image (see FIG. 7B).
- Each pixel forming the distribution image is a three-dimensional pixel having a size in each of the X, Y, and Z axis directions, and can be called a voxel.
- each pixel forming the distribution image has a cubic shape.
- a certain focused pixel is expressed as a pixel j or a pixel k using an arbitrary integer j or k indicating a pixel number.
- the position of the pixel j is represented by the symbol r j (see FIG. 8).
- FIG. 8 shows the relationship between the Compton conical surface of the i-th event, the pixel j, the Compton scattering position R i1, and the like.
- the position r j of the pixel j represents the center position of the pixel j (the center coordinate of the pixel j in the image space IS).
- a pixel refers to a pixel in the image space IS (and thus a pixel in the distribution image).
- the main arithmetic unit 30 can generate a distribution image of the gamma ray source by an image reconstruction method based on a list-mode maximum-likelihood expectation-maximization (LM-ML-EM) method.
- LM-ML-EM list-mode maximum-likelihood expectation-maximization
- lambda j represented by lambda j (l) or lambda j (l + 1) indicates the pixel value of the pixel j in the distribution image (the j th pixel distribution image).
- the pixel value of the pixel j represents an expected value of the number of radionuclides that have collapsed at the pixel j within the imaging time of the Compton camera 10. Since the number of decays of radionuclide per unit time follows a Poisson distribution, ⁇ j is an expected value of this Poisson distribution.
- the radionuclide here is a gamma ray emitting nucleus or positron emitting nucleus as the gamma ray source 201.
- the pixel value is updated by iterative calculation to bring the gamma ray source distribution closer to the estimated true distribution.
- ⁇ j (l) indicates the pixel value of the pixel j obtained in the first iteration (the above expected value for pixel j), and ⁇ j (l + 1) is the (l + 1) th iteration.
- the pixel value of the pixel j obtained by (the above expected value for the pixel j) is shown. Accordingly, the symbol “l” associated with ⁇ j indicates the number of iterations (thus an integer).
- t ij is called a system response function, and represents the probability that the gamma ray emitted from the pixel j is measured as the i-th event.
- the system response function t ij is set for pixels that intersect the Compton conical surface estimated from the event measurement data.
- Y i represents the number of gamma rays detected by the Compton camera 10 in the i-th event.
- Y i 1”.
- the actual number of detected gamma rays (corresponding to “ ⁇ k t ik ⁇ k ” of the denominator) obtained by calculation (corresponding to the number of detected gamma rays) (Corresponding to “Y i ”).
- the pixel value is updated by iterative calculation based on this ratio.
- the pixel value is updated in the iterative calculation so that the calculated gamma ray detection number (denominator) approaches the actually measured gamma ray detection number (numerator).
- the pixel value is updated so that the calculated gamma ray detection number becomes smaller than before the update, and the calculated gamma ray detection number becomes the actual gamma ray detection number. If it is smaller than the number, the pixel value is updated so that the number of detected gamma rays is larger than before the update.
- the detection sensitivity parameter s j is a detection sensitivity parameter for pixel j, and division by s j means correction of the pixel value in consideration of detection sensitivity.
- the detection sensitivity parameter s j mainly includes the geometric efficiency of the detector 11 viewed from the pixel j, the interaction probability between the gamma ray and the detectors 11 and 12, and the time from when the gamma ray is emitted until it is detected. Consists of factors such as attenuation. Guests can set the detection sensitivity parameter s j through a numerical calculation using these factors may be obtained detection sensitivity parameter s j by Monte Carlo simulation.
- the geometric efficiency of the detector 11 viewed from the pixel j represents the size (solid angle) of the detector 11 viewed from the pixel j.
- the probability that the gamma ray emitted from the pixel j hits the detector 11 depends on the geometric efficiency of the detector 11 viewed from the pixel j. The closer the distance between the pixel j and the detector 11 is, the larger the size of the detector 11 viewed from the pixel j is, and the probability that the gamma rays emitted from the pixel j hit the detector 11 increases.
- the geometric efficiency of the detector 11 viewed from the pixel j is approximately proportional to the inverse square of the distance between the pixel j and the detector 11.
- Formula (B1a) and Formula (B1b) are a part of Formula (B1).
- i and k indicate an event number and a pixel number, respectively.
- Expression (B1a) represents the total sum of “t ik ⁇ k (l) ” for all pixels in the i-th event.
- i and j represent an event number and a pixel number, respectively.
- Expression (B1b) represents the sum of all the fractional events having expression (B1a) as the denominator and “Y i t ij ” as the numerator. All events refer to all of the events used for obtaining the pixel value of each pixel of the distribution image.
- the arrival direction of the gamma ray toward the upstream detector 11 is limited to the generatrix direction of the Compton conical surface.
- the Compton scattering angle is determined by the energy at Compton scattering, it has uncertainty derived from a physical phenomenon during gamma ray detection.
- the distribution function of the Compton scattering angle (the distribution function of the error distribution of the Compton scattering angle) showing this uncertainty is called a scattering angle distribution function.
- the Compton conical surface also has an uncertainty according to the scattering angle distribution function, and it is estimated that the gamma ray source exists stochastically in the spread derived from the scattering angle distribution function.
- a function f O ( ⁇ ) represented by the following formula (C1) is defined as a scattering angle distribution function.
- FIG. 9 shows a graph of the scattering angle distribution function f O ( ⁇ ).
- Z represents the atomic number of a semiconductor material that constitutes the detectors 11 and 12 and interacts with gamma rays (hereinafter referred to as detector material), and n l represents the detection thereof. This represents the number of electrons in the l-th electron shell in the organic material.
- the summation operation symbol “ ⁇ ” in equation (C1) indicates the summation for all electron shells in the detector material.
- ⁇ e represents a standard deviation obtained from the energy detection resolution when the detector 11 detects the energy E 1 .
- ⁇ d, l represents the standard deviation of the detected energy due to Doppler broadening due to the electron momentum of the l th electron shell in the detector material.
- parameters other than ⁇ C , ⁇ , ⁇ e and ⁇ d, l have predetermined values known to the imaging apparatus 1.
- ⁇ C represents the Compton scattering angle obtained by the above formula (A1).
- ⁇ represents the Compton scattering angle when it is assumed that the gamma rays emitted from the target position NP are Compton scattered at the interaction position R 1 in the detector 11. That is, the angle formed by the straight line connecting the interaction positions R 1 and R 2 of the detectors 11 and 12 (the central axis of the Compton conical surface) and the straight line connecting the target position NP and the position R 1 corresponds to the angle ⁇ . To do.
- the angle ( ⁇ C ) is the Compton scattering angle ⁇ obtained when the gamma rays emitted from the target position NP are Compton scattered at the interaction position R 1 in the detector 11 and the Compton obtained by the above formula (A1).
- the error angle ⁇ with respect to the scattering angle ⁇ C is represented.
- the intensity of the scattering angle distribution function f O ( ⁇ ) (that is, the value of the function f O ( ⁇ )) is maximum when the error angle ⁇ is zero, and is the absolute value of the error angle ⁇ . As it increases, it decreases toward zero.
- a back projection operation is performed to project the conical gamma ray arrival direction estimated from the gamma ray measurement data onto the image space IS.
- the conical gamma ray arrival direction estimated from the gamma ray measurement data includes an error based on the above-described uncertainty, it is necessary to project a cone having a thickness due to the error in the back projection operation. Since the cone of thickness to be considered in the backprojection is due uncertainty Compton scattering angle may be determined the thickness from the scattering angle distribution function f O ( ⁇ ).
- a Compton conical surface having a thickness is called a Compton conical surface region.
- the main calculation unit 30 can set the Compton conical surface area by giving the thickness to the Compton conical surface for each event as follows.
- the integrated values when the function f O ( ⁇ ) is integrated at the angle ⁇ are referred to as INT A and INT B, respectively.
- INT A and INT B are referred to as INT A and INT B, respectively.
- a positive angle ⁇ REF at which the ratio “INT B / INT A ” is close to 1 and becomes a predetermined positive value (for example, 95%) less than 1 is obtained.
- FIG. 11 shows three conical surfaces 311 to 313 considered for one event.
- the conical surfaces 311 to 313 are all centered on the straight line of the flight path of the gamma ray after Compton scattering (that is, on the straight line passing through the positions R 1 and R 2 ) with the measured Compton scattering position R 1 as the vertex.
- the conical surface 312 is a Compton conical surface having an angle ⁇ C as a half apex angle (half apex angle; opening angle), while the conical surfaces 311 and 313 each have an angle ( ⁇ C ⁇ REF ) as a half.
- the main calculation unit 30 sets a region between the lower limit conical surface 311 and the upper limit conical surface 313 as the Compton conical surface region.
- the Compton conical surface 312 is included in the region between the lower limit conical surface 311 and the upper conical surface 313, that is, the Compton conical surface region.
- ( ⁇ REF ⁇ 2) corresponds to the above-described thickness expressed as an angle.
- the thickness of the Compton conical surface at a portion where the distance from the position R 1 is d corresponds to the length of an arc having a radius d and a central angle ( ⁇ REF ⁇ 2).
- a first improvement method relating to generation of a distribution image will be described.
- the first improvement method and each improvement method to be described later are also based on the LM-ML-EM method as in the reference method described above.
- a distribution image is generated based on the following formula (D1).
- t ij ⁇ j represents an expected value measured by gamma rays emitted from the pixel j in which the i-th event has an accumulation amount (in other words, an abundance) of radionuclides corresponding to the pixel value ⁇ j .
- v ij and s ij will be described later.
- the significance of the sum ( ⁇ ) in formula (D1) will be supplementarily described.
- the formula (D1a) and the formula (D1b) are a part of the formula (D1).
- i and k indicate an event number and a pixel number, respectively.
- Expression (D1a) represents the sum of “t ik ⁇ k (l) ” for all pixels in the i-th event.
- i and j represent an event number and a pixel number, respectively.
- Equation (D1b) represents the sum of all fractional events having the product of equation (D1a) and s ij as the denominator and “v ij t ij ” as the numerator. All events refer to all of the events used for obtaining the pixel value of each pixel of the distribution image.
- a cone 360 is a cone having a Compton conical surface derived based on the measurement data of the first event as a side surface and the length of the generatrix as d A
- the cone 370 is the measurement data of the second event. Is a cone having a Compton conical surface derived from the above as a side surface and a bus bar length d A. Near the bottom surface of the cone 360, the cone 360 is within the image space IS, while a portion of the cone 370 protrudes from the image space IS near the bottom surface of the cone 370. In the illustration of the cone 370 in FIG.
- a portion where the cone 370 and the image space IS intersect (that is, a portion of the cone 370 located in the image space IS) is indicated by a solid line, and the cone 370 and the image are displayed.
- a portion where the space IS does not intersect is indicated by a broken line (corresponding to a broken line 371).
- a spherical surface having the Compton scattering position R 11 in the first event and having a radius d A (hereinafter referred to as a first spherical surface), and having a Compton scattering position R 21 in the second event and having a radius d A.
- a second spherical surface Is assumed to be a spherical surface (hereinafter referred to as a second spherical surface).
- reference numeral 380 shows a curved surface that is the intersection of the first spherical surface and the image space IS projected onto the XY plane.
- 390 shows a curved surface that is the intersection of the second spherical surface and the image space IS projected on the XY plane.
- FIG. 14A Although the shape of the projection of the intersection of the first spherical surface and the image space IS on the XY plane may not actually be a quadrangle, in FIG. 14A, it is assumed that the shape is a quadrangle for simplicity. is doing. The same applies to FIG. 14B.
- the thickness of the Compton conical surface based on the measurement data of the first event is given a thickness based on the scattering angle distribution function, so that the Compton conical surface region of the first event is set.
- a hatched area 381 is an intersection area between the Compton conical surface area of the first event and the curved surface 380 (that is, an intersection area of the Compton conical surface area of the first event, the first spherical surface, and the image space IS).
- a hatched area 391 is an intersection area between the Compton conical surface area of the second event and the curved surface 390 (that is, an intersection area of the Compton conical surface area of the second event, the second spherical surface, and the image space IS).
- a dot region 392 represents a region that does not intersect the image space IS among the intersecting regions of the second event Compton conical surface region and the second spherical surface.
- the ratio of the area intersecting the image space IS in the intersecting area between the Compton conical surface area and the first spherical surface of the first event is 100%.
- a distance d A from Compton scattering position R 11 in and first event pixels belonging to the region 381 is a 100 and uniform intensity of the scattering angle distribution function in the region 381
- the probability is approximately 1/100, and the system response function t based on such estimation is assumed. 1j is set.
- the ratio of the region intersecting the image space IS in the intersecting region between the Compton conical surface region and the second spherical surface of the second event is 50%. Therefore, the number of pixels belonging to the region 391 is 50, the gamma ray source in the second event is located at a distance d A from the Compton scattering position R 21 , and the intensity of the scattering angle distribution function in the region 391 is uniform. Assuming that the gamma ray source of the second event exists in each pixel belonging to the region 391, the probability that the second event gamma ray source exists is approximately “1/50”, and the system response function t based on such estimation is assumed. 2j is set.
- the regions 391 and 392 are present in the image space IS, it should be estimated that gamma rays have arrived from approximately 100 pixels in each event, and therefore belong to the region 391.
- the probability that a second event gamma source is present in each pixel should be estimated to be approximately "1/100" equally.
- the gamma ray source of the second event is assigned to each pixel belonging to the region 391. It is estimated that there is a probability of “1/50”.
- a cone 410 is a cone having a Compton conical surface derived from the measurement data of the i-th event as a side surface and the length of the generatrix as d A
- the cone 420 is the measurement data of the i-th event.
- the events corresponding to the cones 410 and 420 are common, and since “0 ⁇ d A ⁇ d B ”, the conical surface of the cone 410 is part of the conical surface of the cone 420.
- the cone 410 is within the image space IS, while a portion of the cone 420 protrudes from the image space IS near the bottom surface of the cone 420.
- a portion where the cone 420 intersects the image space IS is indicated by a solid line, and the cone 420 and the image are displayed.
- a portion where the space IS does not intersect is indicated by a broken line (corresponding to the broken line 421).
- a sphere having a Compton scattering position R i1 in the i-th event and having a radius d A (hereinafter referred to as a spherical surface A), and a Compton scattering position R i1 in the i-th event and having a radius d B in the center.
- a spherical surface (hereinafter referred to as spherical surface B) is assumed.
- reference numeral 440 indicates a curved surface that is the intersection of the spherical surface A and the image space IS projected onto the XY plane.
- reference numeral 450 denotes a curved surface that is the intersection of the spherical surface B and the image space IS projected onto the XY plane.
- the shape of the projected portion of the intersection of the spherical surface A and the image space IS on the XY plane may not actually be a quadrangle, in FIG. 16A, it is assumed that the shape is a quadrangle for simplicity. ing. The same applies to FIG.
- the thickness of the Compton conical surface based on the measurement data of the i-th event is given a thickness based on the scattering angle distribution function, whereby the Compton conical surface region of the i-th event is set.
- a hatched area 441 is a crossing area between the Compton conical surface area of the i-th event and the curved surface 440 (that is, a crossing area of the Compton conical surface area of the i-th event, the spherical surface A, and the image space IS). Represents.
- FIG. 16A a hatched area 441 is a crossing area between the Compton conical surface area of the i-th event and the curved surface 440 (that is, a crossing area of the Compton conical surface area of the i-th event, the spherical surface A, and the image space IS).
- a hatched area 451 is a crossing area between the Compton conical surface area of the i-th event and the curved surface 450 (that is, a crossing area of the Compton conical surface area of the i-th event, the spherical surface B, and the image space IS).
- a dot region 452 represents a region that does not intersect the image space IS among the intersecting regions of the Compton conical surface region of the i-th event and the spherical surface B.
- the ratio of the area intersecting the image space IS in the intersection area between the Compton conical surface area of the i-th event and the spherical surface A is 100%. Therefore, the number of pixels belonging to the region 441 is N A , the gamma ray source in the i-th event is located at a distance d A from the Compton scattering position R i1 , and the intensity of the scattering angle distribution function in the region 441 is uniform.
- the ratio of the area intersecting the image space IS among the intersecting areas of the Compton conical surface area and the spherical surface B of the i-th event is 50%. Therefore, the number of pixels belonging to the region 451 is N B , the gamma ray source at the i-th event is at a position away from the Compton scattering position R i1 by the distance d B , and the intensity of the scattering angle distribution function in the region 451 is uniform.
- the probability parameter v ij (refer to the above formula (D1)) introduced in the first improvement method serves as a regularization factor for suppressing such a phenomenon (overfitting).
- v ij is defined by the following formula (D2).
- R i1 is an interaction position (ie, Compton scattering position) between the gamma ray detected in the i-th event and the pre-stage detector 11 (see FIG. 8).
- r j indicates the position (center position) of the pixel j.
- r represents an arbitrary position (coordinates) in the image space IS.
- V k represents the volume of the pixel k.
- “R ⁇ V k ” represents a position r included in the pixel k (in other words, coordinates r included in the volume of the pixel k).
- v ij is p for all pixels k intersecting a sphere centered on the Compton scattering position R i1 of the i-th event and having a radius between the position (r j ) of pixel j and the distance between Compton scattering positions R i1. Represents the sum of ik .
- the probability parameter v ij indicates the probability that gamma rays that have caused Compton scattering at the i-th event have arrived from within the image space IS.
- the probability parameter is set for each pixel, and v ij is a probability parameter set for the pixel j. More specifically, v ij is gamma rays caused the Compton scattering and radius the distance between the position (r j) and Compton scattering position R i1 of and the pixel j around the Compton scattering position R i1 at the i event When it is assumed that the position has arrived from a position on the spherical surface, the probability that the gamma ray has arrived from within the image space IS is shown.
- p ik is the gamma ray passing through the pixel k among the gamma-ray arrival directions estimated for the i-th event. Represents the percentage of the arrival direction. That is, p ik represents the probability density of the gamma ray arrival direction passing through the pixel k when the gamma ray arrival direction of the i-th event is regarded as a random variable.
- v ij is the sum of p ik for all the pixels k including the position r satisfying “
- ”. ik represents the contribution of pixel k to v ij .
- p ik is based on the uncertainty of the Compton scattering angle and is defined by the equation (D3).
- p ij is defined using variable j instead of variable k in p ik .
- the angle ⁇ i (r) in the formula (D3) is defined by the formula (D3a).
- f O, i represents the scattering angle distribution function (the distribution function of the error distribution of the Compton scattering angle) at the i-th event.
- f O, i is regarded as a function of the angle ⁇ i (r) depending on the position r.
- the position r represents an arbitrary position in the image space IS as described above, and may be considered to correspond to the target position NP in FIG.
- ⁇ i (r) represents the Compton scattering angle of the gamma ray when it is assumed that the i-th event is caused by the gamma ray emitted from the position r.
- R ii and R i2 are the interaction position of the gamma ray detected at the i-th event and the detector 11 (ie, the detected Compton scattering position) and the gamma ray detected at the i-th event and the detector 12 respectively.
- Interaction position ie, detected photoelectric absorption position
- V j represents the volume of pixel j.
- the position r is used as an integration variable, and “f O, i [ ⁇ i (r)] / ⁇
- ⁇ C in the above formula (C1) is calculated using the formula (A1) for the i-th event. Compton scattering angle.
- ⁇ in the above equation (C1) represents ⁇ i (r), and positions R i1 and R i2 Is equivalent to the angle formed by the straight line connecting the two (the central axis of the Compton conical surface) and the straight line connecting the positions R i1 and r.
- f O, i [ ⁇ i (r)] represents the probability density of the error distribution of the Compton scattering angle at the position r, and p ij is obtained by integrating this over the volume V j of the pixel j. .
- simple integration does not take into account the geometric efficiency of the detector 11 as viewed from the pixel j.
- f O, i is a function of angle, it is necessary to consider a solid angle due to a difference in angle. Therefore, in the equation (D3), f O, i [ ⁇ i (r)] is expressed as the reciprocal of the geometric efficiency approximation “
- the value of the function f O, i for pixels that do not belong to the Compton conical surface region is considered to be sufficiently small. Therefore, the right side of the equation (D2) is for all the pixels k that include the position r satisfying “
- the main arithmetic unit 30 uses the probability parameter v ij indicating the probability that the Compton-scattered gamma rays arrived from the image space IS in each event as the measurement data (R i1 , R i2 , E i1 , E i2 ) are set for each pixel of the distribution image by calculation, and a distribution image is generated using the probability parameter v ij set for each pixel (that is, the pixel value of each pixel is obtained). ).
- the pixels (target pixels) for which the probability parameter v ij is to be set based on the measurement data at each event need not be all the pixels of the distribution image (however, they may be all the pixels of the distribution image). .
- the main calculation unit 30 individually calculates the probability parameter v ij for each of a plurality of target pixels in the distribution image by calculation based on measurement data at each event. It can be said that a distribution image is generated using the set probability parameter v ij (that is, a pixel value of each pixel is obtained).
- the main calculation unit 30 sets the probability parameter v ij for each event and for each target pixel. That is, the main calculation unit 30 pays attention to each of a combination of a plurality of events and a plurality of target pixels, and sets the probability parameter v ij for each combination.
- the main processing unit 30 When attention is paid to the i-th event and one pixel j (when attention is paid to the combination of the i-th event and pixel j), the main processing unit 30 is centered on the detected Compton scattering position R i1 and the position of the focus pixel j (R j ) and a sphere having a radius (eg, d A or d B in FIG. 15) between the Compton scattering position R i1 , the Compton conical region of the i-th event of interest, and the intersection of the image space IS Based on the state, a probability parameter v ij is set for the pixel of interest j of the i-th event that has been noticed.
- the hatched area 441 in FIG. 16A or the hatched area 451 in FIG. 16B corresponds to the intersecting area of the spherical surface, the Compton conical surface area, and the image space IS, and FIGS. These crossing states are different from each other.
- the main calculation unit 30 when focusing on the i-th event and one pixel j, the main calculation unit 30 is centered on the detected Compton scattering position R i1 and the position (r j ) and Compton scattering position of the target pixel j.
- each pixel arranged in the intersection region is extracted, and an index (p ij ) corresponding to the intensity of the scattering angle distribution function (f O, i ) is derived for each extracted pixel (formula (D3 )reference). Then, based on the sum of the indices derived for each extracted pixel (see Expression (D2)), the probability parameter v ij for the pixel of interest j of the i-th event of interest is set.
- v ij is considered to be an integral of the scattering angle uncertainty with the integral range being the range of the total scattering angle
- the ideal value of v ij is 1.
- v ij is smaller than 1 when a part of the Compton conical surface region is outside the image space IS. Therefore, an event in which the Compton conical surface region slightly intersects the image space IS has a reduced contribution to the reconstructed image (distributed image) due to the introduction of v ij , thereby reducing artifacts.
- the intersection state is different if the target pixel is different in one event. In the first improvement method, this is taken into consideration, and the optimum v ij is set for each pixel in each event, so that the artifact reduction effect is enhanced.
- the factor that constitutes the detection sensitivity has the greatest influence on the geometric efficiency of the detector 11 viewed from the individual pixel j.
- the geometric efficiency is approximately the pixel j and the detector 11. It is proportional to the inverse square of the distance between.
- the detection sensitivity (s j ) depends only on the position of the pixel j, and thus depends on the difference in Compton scattering position (interaction position in the detector 11). Differences in geometric efficiency are averaged.
- the distance (511, 521) between the Compton scattering position and the pixel j differs between when the Compton scattering position is the position 510 and when it is the position 520.
- the geometric efficiency Should be different, but assuming that the detection sensitivity (s j ) depends only on the position of pixel j, the difference in geometric efficiency due to the difference in Compton scattering position will be averaged. . The closer to the detector 11, the greater the influence of this averaging (because the difference between the distances 511 and 521 is likely to occur), and the accuracy of sensitivity correction using the detection sensitivity parameter is reduced.
- the detection sensitivity parameter for each pixel is set in consideration of the Compton scattering position R i1 for each event.
- s ij is a detection sensitivity parameter indicating the detection sensitivity of the detector 11 with respect to Compton scattering of the i-th event.
- the detection sensitivity parameter is set for each pixel
- s ij is a detection sensitivity parameter set for the pixel j.
- division by s jj means correction of the pixel value in consideration of detection sensitivity.
- FIGS. 19A and 19B show a comparison between the reference method and the first improved method.
- the detection sensitivity parameter s ij is defined by the following equation (D4).
- ⁇ t (E 0 ) represents the total interaction cross section of the gamma ray having the energy of E 0 and the detector material
- d i, j1 is the detection of the gamma ray from the pixel j at the i-th event.
- the range in the vessel 11 is represented (see FIG. 20). That is, d i, j1 represents the distance that the gamma ray from the pixel j in the i-th event has passed through the detector 11 until Compton scattering occurs.
- exp [ ⁇ t (E 0 ) d i, j1 ] means that the gamma ray from the pixel j does not interact with the detector substance in the detector 11 until the Compton scattering position R i1 in the i th event. It represents the probability of arrival (in other words, the attenuation that occurs in the detector 11 until the gamma ray from the pixel j in the i-th event is Compton scattered).
- the value of ⁇ t (E 0 ) is uniquely determined based on the initial energy E 0 of gamma rays emitted from the gamma ray source 201 and the physical properties of the detector material, and is known to the imaging apparatus 1.
- the range d i, j1 is determined from the position r j of the pixel j, the Compton scattering position R i1 of the i-th event, and the shape of the detector 11.
- the detector 11 can detect the interaction position in the X, Y, and Z axis directions in m, n, and o stages, respectively.
- M, n and o are any integers of 2 or more).
- the detector 11 can be considered as a collection of (m ⁇ n ⁇ o) independent detector elements.
- each detector element has detection sensitivity for all pixels in the image space IS.
- R i1 may be considered to represent the center position of the detector element including the Compton scattering position at the i-th event.
- ⁇ 2 ” represents an approximation of the geometric efficiency of the detector element centered on the position R i1 as viewed from the pixel j.
- the main calculation unit 30 performs calculation based on the detection sensitivity parameter s ij indicating the detection sensitivity of Compton scattering at each event based on the measurement data (including R i1 ) at each event. Is set for each pixel of the distribution image, and the distribution image is generated using the detection sensitivity parameter s ij set for each pixel (that is, the pixel value of each pixel is obtained).
- the pixels (target pixels) for which the detection sensitivity parameter s ij should be set based on the measurement data at each event need not be all pixels of the distribution image (however, they may be all pixels of the distribution image). ). In each event, it is sufficient to include each pixel in the Compton conical surface region as a target pixel for which the detection sensitivity parameter s ij is to be set. That is, in each event, it is sufficient to calculate s ij based on the equation (D4) only for each pixel in the Compton conical surface area. In each event, a constant value may be substituted for s ij for each pixel outside the Compton conical surface area.
- the main calculation unit 30 applies the detection sensitivity parameter s ij to each of a plurality of target pixels in the distribution image by calculation based on measurement data at each event. It can also be said that the distribution image is generated (that is, the pixel value of each pixel is obtained) using the set detection sensitivity parameter s ij individually.
- the main calculation unit 30 sets the detection sensitivity parameter s ij for each event and for each target pixel. That is, the main calculation unit 30 sets the detection sensitivity parameter s ij for each combination by paying attention to each combination of a plurality of events and a plurality of target pixels.
- the main calculation unit 30 sets the detected Compton scattering position R i1 and the position r j of the noticed pixel j to each other. Accordingly, the detection sensitivity parameter (s ij ) is set. More specifically, when attention is paid to the i-th event and one pixel j, the main calculation unit 30 calculates gamma rays before Compton scattering in the detector 11 based on the Compton scattering position R i1 and the position r j of the target pixel j.
- Sensitivity parameter for the target pixel j of the i-th event of interest based on the range d i, j1 of the current and the distance
- the distance between the gamma ray source and the detector 11 is relatively small, the gamma rays from the gamma ray source are likely to hit the detector 11 and thus are easily detected by the detector 11, and the distance between the gamma ray source and the detector 11 is relatively long.
- the gamma ray from the gamma ray source is hard to hit the detector 11, so it is difficult for the detector 11 to detect it. If a distribution image is generated without considering such ease of detection / hardness of detection, the presence of a gamma ray source that is far from the detector 11 is less likely to be reflected in the distribution image.
- the detection sensitivity parameter (s ij ) is set in consideration of not only the position of each pixel but also the interaction position R i1 in the detector 11, thereby detecting without considering the interaction position R i1.
- the detection sensitivity parameter (s ij )
- the detection sensitivity parameter (s ij ) is set in consideration of not only the position of each pixel but also the interaction position R i1 in the detector 11, thereby detecting without considering the interaction position R i1.
- the sensitivity parameter (FIG. 19A)
- more appropriate sensitivity correction can be performed.
- the distribution image more accurately represents the distribution of the gamma ray source.
- the gamma ray attenuation (exp [ ⁇ t (E 0 ) d i, j1 ]) in the detector 11 is also considered using the interaction position R i1 , the validity of the distribution image is considered. (The accuracy of estimating the gamma source distribution increases.)
- d ⁇ c, ij / d ⁇ represents the probability that the gamma ray from the pixel j causes Compton scattering in the i-th event. Since this probability varies depending on the Compton scattering angle, “d ⁇ c, ij / d ⁇ ” depends on the pixel j.
- E 0 ′ represents the energy of gamma rays after Compton scattering. Therefore, E 0 ′ in the equation (D5) corresponds to a value obtained by subtracting the detected energy E 1 at the i-th event from the initial energy E 0 of the gamma ray.
- ⁇ t (E 0 ′) represents the total interaction cross section between the gamma ray having the energy of E 0 ′ and the detector substance
- d i, 12 represents the gamma ray after Compton scattering at the i-th event as the detector.
- 12 represents the range of the gamma ray detectors 11 and 12 until photoelectric absorption is performed.
- d i, 12 represents the distance that the gamma rays after Compton scattering at the i-th event have passed through the detectors 11 and 12 until they are photoelectrically absorbed by the detector 12. Therefore, “exp [ ⁇ t (E 0 ′) d i, 12 ]” interacts with the detector substance in the detectors 11 and 12 until the gamma ray after Compton scattering at the i-th event reaches the photoelectric absorption position R i2. (In other words, attenuation that occurs in the detectors 11 and 12 before the gamma ray after Compton scattering in the i-th event is photoelectrically absorbed by the detector 12 at the position R i2 ).
- ⁇ 2 represents an approximation of the geometric efficiency (solid angle) of the photoelectric absorption position R i2 viewed from the Compton scattering position R i1 .
- ⁇ p (E 0 ′) represents the photoelectric absorption cross section between the gamma ray having the energy of E 0 ′ and the detector material, that is, the probability that the gamma ray after Compton scattering causes photoelectric absorption.
- the main calculation unit 30 can derive and set the system response function t ij for each event and for each pixel based on the measurement data (R i1 , R i2 , E i1 , E i2 ) for each event.
- the measurement data R i1 , R i2 , E i1 , E i2
- D5 the value of the physical quantity that does not depend on the measurement data of the i-th event is known to the imaging apparatus 1.
- FIG. 22 is a flowchart for generating a distribution image based on the measurement data of each event.
- a procedure for generating a distribution image based on measurement data of each event will be described.
- the main calculation unit 30 substitutes a predetermined common initial value (for example, 1) for the pixel value ⁇ k (0) of all the pixels of the distribution image, and sets a variable l indicating the number of iterations to zero. substitute.
- the main arithmetic unit 30 assigns 1 to the variable i in the subsequent step S22, and then sequentially executes the processes of steps S23 to S28.
- the main operation unit 30 reads the measurement data (R i1 , R i2 , E i1 , E i2 ) of the i-th event from the storage unit 44 in step S23, and measures the i-th event in step S24. Based on the data, the Compton scattering angle ⁇ C is calculated according to the above formula (A1) and the Compton conical surface is set. In subsequent step S25, the main computing unit 30 sets the Compton conical surface area of the i-th event by giving the thickness of the Compton conical surface of the i-th event based on the scattering angle distribution function f O ( ⁇ ).
- step S26 the main calculation unit 30 regards each pixel in the Compton conic surface area of the i-th event as a target pixel, and for each target pixel, v ij , Calculate and set s ij and t ij .
- the system response function t ij may be set to zero for each pixel (non-target pixel) outside the Compton conical surface area.
- step S27 subsequent to step S26, the main arithmetic unit 30 adds v ij , s ij and t ij obtained for the i-th event to the sum calculation result of the equation (D1b) obtained up to the (i ⁇ 1) -th event.
- step S28 it is confirmed whether or not the measurement data for all events has been read from the storage unit 44.
- step S28 If there is measurement data that has not yet been read (N in step S28), 1 is added to the variable i in step S29, and the process returns to step S23, and the processes in steps S23 to S28 are repeated. If measurement data for all events has been read from the storage unit 44 (Y in step S28), the process proceeds to step S30.
- step S30 the main calculation unit 30 performs a pixel value update process.
- the updated pixel value ⁇ k (l + 1) is obtained according to the equation (D1). That is, the pixel value ⁇ k (l + 1) of each pixel of the distribution image is obtained by multiplying the sum total calculation result of the latest formula (D1b) obtained in step S27 by ⁇ k (l) .
- the main arithmetic unit 30 confirms whether or not a predetermined update end condition is satisfied. If the update end condition is satisfied, the process proceeds to step S33, the distribution image having the latest pixel value ⁇ k is stored in the storage unit 44, and the operation of FIG. If the update end condition is not satisfied, 1 is added to the variable l in step S32, and then the process returns to step S22, and the processes after step S22 are repeated.
- the update end condition is satisfied, for example, when the number of executions of the update process in step S30 reaches a predetermined number.
- the update end condition may be satisfied when a predetermined end instruction operation is input to the operation unit 43 during the repeated execution of the processes of steps S22 to S32.
- the amount of change in each pixel value ⁇ k by the update process in step S30 is sufficiently small (for example, the sum of the pixel values ⁇ k (l + 1) in all the pixels of the distribution image and the pixel value ⁇ k When the difference from the sum of (l) is equal to or less than a predetermined value), the update end condition may be satisfied.
- Second Improvement Method Regarding Distribution Image Generation A second improvement method relating to the generation of the distribution image will be described.
- a distribution image is generated based on the following formula (E1).
- equation (E1) v ij is introduced based on the reference method, but the introduction of s ij is postponed.
- the second improvement method is the same as the first improvement method except that s j is incorporated in the calculation formula of the pixel value ⁇ j instead of s ij .
- the detection sensitivity parameter s j is obtained for each pixel, for example, according to the following formula (E2).
- ⁇ t (E 0 ) represents the total interaction cross section between the gamma ray having energy of E 0 and the detector material
- r j represents the position (center position) of the pixel j.
- p is a detector element in the first to (m ⁇ n ⁇ o) th detector elements forming the detector 11 (see FIGS. 21A and 21B).
- R p represents the center position of the p-th detector element.
- d p, j1 is the length of the path located in the detector 11 in the straight path from the position r j to the position R p . That is, d p, j1 is similar to d i, j1 in FIG.
- Equation (E2) Equation (E2) interacts with the detector substance in the detector 11 until the gamma ray from the pixel j reaches the p-th detector element. Represents the probability of reaching without.
- ⁇ 2 represents an approximation of the geometric efficiency of the p th detector element as viewed from pixel j.
- s j in the expression (E2) is calculated as the sum of detection sensitivity (gamma ray detection sensitivity) set for all detector elements. Is done. Since the number of detector elements is a fixed value, the sum of the detection sensitivities can be regarded as an average value of the detection sensitivities of all the detector elements.
- the size of each detector element is the product (for example, 3 ⁇ 3 ⁇ 1 [mm 3 ]) of the X, Y, and Z axis components of the position detection accuracy of the detector 11.
- the second improved method an advantageous effect is achieved by introducing v ij in comparison with the reference method.
- the first improved method that also introduces s ij is more suitable for generating a more accurate distribution image than the second improved method.
- a third improved method relating to the generation of the distribution image will be described.
- a distribution image is generated based on the following formula (F1) or formula (F2).
- s ij is introduced based on the reference method, but the introduction of v ij is postponed.
- the third improvement method is the same as the first improvement method, except that Y i or v i is incorporated in the calculation formula of the pixel value ⁇ j instead of v ij .
- Y i may be “1” in all events.
- the probability parameter v i is set for each event, and one probability parameter v i is commonly applied to each pixel.
- the probability parameter v i is calculated according to the following formula (F3).
- the right side of the equation (F3) represents the sum of p ij for all pixels in the Compton conic surface area of the i-th event.
- the third improved method an advantageous effect is achieved by introducing s ij in comparison with the reference method.
- the first improved method that also introduces v ij is more suitable for generating a more accurate distribution image than the third improved method.
- Using v i instead of v ij also reduces artifacts in comparison with the reference method, but as can be seen from the above description with reference to FIGS. 15 and 16 (a) and (b),
- the first improved method of setting the optimal v ij for each pixel in each event has a higher artifact reduction effect.
- FIG. 23 is a schematic plan view of a tumor-bearing mouse used in the experiment.
- An imaging target 200 is obtained by administering a 64 Cu-labeled anti-HER2 antibody to a tumor bearing mouse transplanted with three tumor tissues A431, 4T1, and C6.
- the 64 Cu-labeled anti-HER2 antibody is an antibody that contains 64 Cu and accumulates in a tumor containing HER2.
- HER2 is most highly expressed in the tumor tissue A431, and therefore high antibody accumulation is expected against the tumor tissue A431.
- 64 Cu in the tumor-bearing mouse functions as the gamma ray source 201
- 511 keV (kiloelectron volt) gamma rays accompanying the annihilation of positrons emitted from 64 Cu function as the gamma rays from the gamma ray source 201.
- the imaging time using the Compton camera 10 is 9 hours, and the measured number of events is 1.9 ⁇ 10 6 .
- the three-dimensional distribution images obtained as a result of the experiment to the reference method, the second improvement method, and the first improvement method are referred to as three-dimensional distribution images 610, 620, and 630, respectively (see FIG. 24).
- the number of iterations of the iterative calculation in the experiment is uniformly 60 times. That is, under the conditions of the experiment, the three-dimensional distribution images 610, 620, and 630 have ⁇ j (60) in the expressions (B1), (E1), and (D1) as the pixel value of the pixel j, respectively. It is a three-dimensional distribution image.
- the MIP images 611, 612, and 613 are planar images obtained by projecting the three-dimensional distribution image 610 onto the XY plane, the ZX plane, and the YZ plane, respectively, using a maximum value projection method (maximum intensity projection).
- the MIP images 621, 622, and 623 are planar images obtained by projecting the three-dimensional distribution image 620 onto the XY plane, the ZX plane, and the YZ plane, respectively, using the maximum value projection method.
- the MIP images 631, 632, and 633 are planar images obtained by projecting the three-dimensional distribution image 630 onto the XY plane, the ZX plane, and the YZ plane, respectively, using the maximum value projection method. That is, for example, in the three-dimensional distribution image 610, the unit operation for projecting the pixel having the maximum pixel value on the XY plane from the group of pixels having the same position in the X and Y axis directions is performed on the entire three-dimensional distribution image 610. As a result, the MIP image 611 is formed on the XY plane that is the projection plane. The same applies to the MIP images 612 and 613, and the same applies to the MIP images 621 to 623 and 631 to 633.
- VOI volume of interest
- Regions 620 [A431], 620 [4T1], and 620 [C6] in which tumor tissues A431, 4T1, and C6 exist are set in the image 620, and tumor tissues A431, 4T1, and C6 exist in the three-dimensional distribution image 630 Areas 630 [A431], 630 [4T1], and 630 [C6] are set.
- the size of the region set for the tumor tissue A431 (that is, the regions 610 [A431], 620 [A431], and 630 [A431]) is the same between the images 610, 620, and 630. The same applies to the regions set for the tumor tissues 4T1 and C6.
- FIG. 27 shows the measurement results of the radioactivity of 64 Cu from each tissue of the tumor-bearing mouse after the imaging experiment.
- FIG. 28 in the bar graph composed of the bars 651 to 653, after normalizing the measured radioactivity from the tumor tissue A431 shown by the bar 651 as 1, the measured radioactivity from the tumor tissues 4T1 and C6 is shown. Represented by bars 652 and 653, respectively.
- the relative intensities indicated by the bars 671 to 673 in FIG. 28 represent the total values of the pixel values in the regions 630 [A431], 630 [4T1], and 630 [C6] based on the first improved method, respectively. However, the relative intensity indicated by the bars 671 to 673 is normalized by setting the relative intensity corresponding to the region 630 [A431] to 1.0.
- the relative intensities indicated by the bars 681 to 683 in FIG. 28 represent the total values of the pixel values in the regions 620 [A431], 620 [4T1], and 620 [C6] based on the second improved method, respectively.
- the relative intensity indicated by the bars 681 to 683 is normalized by setting the relative intensity corresponding to the region 620 [A431] to 1.0.
- the relative intensities indicated by the bars 691 to 693 in FIG. 28 represent the total values of the pixel values in the regions 610 [A431], 610 [4T1], and 610 [C6] based on the reference method, respectively.
- the relative intensity indicated by the bars 691 to 693 is normalized by setting the relative intensity corresponding to the region 610 [A431] to 1.0.
- the relative intensities of the tumor tissues 4T1 and C6 viewed from the relative intensity (1.0) of the tumor tissue A431 are the measured radioactivity (intensities of the bars 652 and 653 viewed from the intensity of the bar 651). It can be said that the closer to, the better the quantitativeness of the image.
- the strength of the bars 672 and 673 viewed from the strength of the bar 671 and the strength of the bars 682 and 683 viewed from the strength of the bar 681 are both
- the strength of the bars 652 and 653 viewed from the strength of the bar 651 is close. That is, in comparison with the reference method, the first or second improvement method improves the quantitativeness of the image.
- the target device that is the imaging device 1 itself or the target device that is a part of the imaging device 1 may be configured by hardware such as an integrated circuit or a combination of hardware and software. it can.
- Arbitrary specific functions that are all or part of the functions realized in the target device may be described as a program, and the program may be stored in a memory (flash memory or the like) that can be mounted on the target device. . Then, the specific function may be realized by executing the program on a program execution device (for example, a microcomputer that can be mounted on the target device).
- the program can be stored and fixed on an arbitrary recording medium.
- the recording medium for storing and fixing the program may be mounted or connected to a device (such as a server device) different from the target device.
- the imaging apparatus 1 includes event detection means for detecting an event and calculation means for generating a distribution image.
- the event detection means is mainly realized by, for example, the event measurement data acquisition unit 20, and the calculation means is mainly realized by, for example, the main calculation unit 30 (see FIG. 3).
- the computing means (30) is a conical surface setting means (S21; FIG. 22) for setting a Compton conical surface for each event, a conical surface area setting means (S22; FIG. 22) for setting a Compton conical surface region for each event, Probability parameter setting means (S23; FIG.
- the sensitivity parameter setting means (S23; FIG. 22) for setting a probability parameter v ij for each pixel (target pixel) and detection for setting a detection sensitivity parameter s ij for each event and for each pixel (target pixel) It can be said that the sensitivity parameter setting means (S23; FIG. 22) is included.
Landscapes
- Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Molecular Biology (AREA)
- High Energy & Nuclear Physics (AREA)
- General Physics & Mathematics (AREA)
- Biomedical Technology (AREA)
- Optics & Photonics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Medical Informatics (AREA)
- General Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Nuclear Medicine (AREA)
- Measurement Of Radiation (AREA)
Abstract
Description
図1(a)及び(b)を参照して、コンプトン散乱を利用したガンマ線源の分布推定方法を説明する。図1(a)及び(b)において、11及び12は、夫々、コンプトンカメラ10を形成する前段検出器及び後段検出器を表している。検出器11及び12の各々は相互作用位置及びエネルギを測定可能な放射線検出器であり、半導体や発光物質(シンチレーター)などで形成される。201は、ガンマ線を放射するガンマ線源を表している。200は、ガンマ線源201を内包する撮像対象を表している。検出器11及び12は、典型的には互いに離間した状態で平行に配置されている。ガンマ線源201を内包する撮像対象200から見て、前段検出器11、後段検出器12の順に並んでいる。即ち、ガンマ線源201を内包する撮像対象200と後段検出器12との間に前段検出器11が配置される。尚、図1(b)には、撮像対象200内にガンマ線源201が1つしか示されていないが、多数のガンマ線源201が撮像対象200内に散在しうる。
図3は、本実施形態に係るガンマ線源分布画像化装置1の概略的な全体構成図である。画像化装置1は、コンプトンカメラ10に加え、符号20、30及び41~44によって参照される各部位を備える。但し、コンプトンカメラ10は画像化装置1の構成要素に含まれない、と考えることも可能である。
主演算部30は、list-mode maximum-likelihood expectation-maximization(LM-ML-EM)法を元にした画像再構成法により、ガンマ線源の分布画像を生成することができる。これは、観測されたガンマ線の測定データが得られる確率が、統計的に最も高くなるように、空間中のガンマ線源分布を推定する手法である。
各イベントにおいて、前段検出器11に向かうガンマ線の到来方向はコンプトン円錐面の母線方向に限定される。しかし、コンプトン散乱角はコンプトン散乱時のエネルギによって求められるため、ガンマ線検出時の物理現象に由来する不確定性を持つ。この不確定性を示す、コンプトン散乱角の分布関数(コンプトン散乱角の誤差分布の分布関数)を散乱角分布関数と呼ぶ。コンプトン円錐面も散乱角分布関数に従った不確定性を有しており、ガンマ線源は散乱角分布関数に由来する広がりの中に確率的に存在すると推定される。
分布画像の生成に関する第1改良方法を説明する。第1改良方法並びに後述の各改良方法も、上述の基準方法と同じくLM-ML-EM法を元にしている。但し、第1改良方法では、下記式(D1)に基づき分布画像を生成する。
式(B1)による基準方法では、計算によって推定されるガンマ線検出数(Σktikλk)が極めて小さい場合などにおいて、画素値が非常に大きな値に更新されるオーバーフィッティングが生じる。これは、図12に示す如く、円錐状に推定されるガンマ線到来方向が画像空間ISに対してかすめるように交差した場合などに生じ、画像の辺縁部でアーチファクト(虚像)が生じる原因をとなる。
検出感度を構成する因子で最も大きな影響を持つのは、個々の画素jから見た検出器11の幾何学的効率であり、その幾何学的効率は、近似的には画素j及び検出器11間の距離の逆二乗に比例する。しかし、式(B1)による基準方法では、検出感度(sj)が画素jの位置に対してのみ依存すると仮定しているため、コンプトン散乱位置(検出器11での相互作用位置)の違いによる幾何学的効率の違いが平均化される。
システム応答関数tijは下記式(D5)に従って決定される。
図22は、各イベントの測定データに基づく分布画像の生成動作フローチャートである。図22を参照して、各イベントの測定データに基づく分布画像の生成動作手順を説明する。図4による各イベントの測定データの取得後、ステップS21に至る。ステップS21において、主演算部30は、分布画像の全画素の画素値λk (0)に対し所定の共通初期値(例えば1)を代入すると共に、反復計算の回数を示す変数lにゼロを代入する。主演算部30は、続くステップS22にて変数iに1を代入した後、ステップS23~S28の処理を順次実行する。
分布画像の生成に関する第2改良方法を説明する。第2改良方法では、下記式(E1)に基づき分布画像が生成される。式(E1)では、基準方法を元にしてvijが導入されてはいるが、sijの導入は見送られている。sijの代わりにsjが画素値λjの算出式に組み込まれている点を除き、第2改良方法は第1改良方法と同様である。検出感度パラメータsjは、例えば下記式(E2)に従って画素ごとに求められる。
分布画像の生成に関する第3改良方法を説明する。第3改良方法では、下記式(F1)又は式(F2)に基づき分布画像が生成される。式(F1)又は式(F2)では、基準方法を元にしてsijが導入されてはいるが、vijの導入は見送られている。vijの代わりにYi又はviが画素値λjの算出式に組み込まれている点を除き、第3改良方法は第1改良方法と同様である。
基準方法、第1改良方法及び第2改良方法を利用した実験の内容及び結果を説明する。図23は、当該実験で利用された担癌マウスの概略平面図である。3つの腫瘍組織A431、4T1及びC6が移植された担癌マウスに64Cu標識抗HER2抗体を投与したものを撮像対象200とする。64Cu標識抗HER2抗体は、64Cuを内包する抗体であって、HER2を含む腫瘍に集積する抗体である。3つの腫瘍組織の内、腫瘍組織A431においてHER2が最も高発現であり、故に腫瘍組織A431に対して高い抗体の集積が期待される。本実験では、担癌マウス中の64Cuがガンマ線源201として機能し、64Cuより放出される陽電子の対消滅に伴う511keV(キロエレクトロンボルト)のガンマ線が、ガンマ線源201からのガンマ線として機能する。当該実験において、コンプトンカメラ10を用いた撮像時間は9時間であり、測定されたイベント数は1.9×106である。
図28のバー681~683が示す相対強度は、夫々、第2改良方法に基づく領域620[A431]、620[4T1]、620[C6]内の画素値の合計値を表す。但し、バー681~683が示す相対強度に関し、領域620[A431]に対応する相対強度を1.0とする正規化を行っている。
図28のバー691~693が示す相対強度は、夫々、基準方法に基づく領域610[A431]、610[4T1]、610[C6]内の画素値の合計値を表す。但し、バー691~693が示す相対強度に関し、領域610[A431]に対応する相対強度を1.0とする正規化を行っている。
本発明の実施形態は、特許請求の範囲に示された技術的思想の範囲内において、適宜、種々の変更が可能である。以上の実施形態は、あくまでも、本発明の実施形態の例であって、本発明ないし各構成要件の用語の意義は、以上の実施形態に記載されたものに制限されるものではない。上述の説明文中に示した具体的な数値は、単なる例示であって、当然の如く、それらを様々な数値に変更することができる。
10 コンプトンカメラ
11 前段検出器
12 後段検出器
20 イベント測定データ取得部
30 主演算部
200 撮像対象
201 ガンマ線源
R1、Ri1 相互作用位置(コンプトン散乱位置)
R2、Ri2 相互作用位置(全エネルギ吸収位置)
θC コンプトン散乱角
IS 画像空間
Claims (16)
- 後段検出器と、
ガンマ線源と前記後段検出器との間に配置された前段検出器と、
前記前段検出器でコンプトン散乱されたガンマ線が前記後段検出器にて光電吸収されるイベントを検出するイベント検出手段と、
複数のイベントにおける各検出器とガンマ線との相互作用の測定データに基づき、前記ガンマ線源の空間分布を示す分布画像を、前記ガンマ線源を内包する画像空間内の画像として生成する演算手段と、を備えた画像化装置において、
前記演算手段は、各イベントでコンプトン散乱したガンマ線が前記画像空間内から到来した確率を示す確率パラメータを各イベントの前記測定データに基づき前記分布画像内の複数の画素の夫々に対して個別に設定し、設定した前記確率パラメータを用いて前記分布画像を生成する
ことを特徴とする画像化装置。 - 前記演算手段は、
前記測定データに基づいて、前記イベントごとに、コンプトン散乱位置を頂点とし且つコンプトン散乱角を半頂角とし且つコンプトン散乱後のガンマ線の飛行経路の直線上に中心軸を持った円錐面を設定する円錐面設定手段と、
前記イベントごとに、前記コンプトン散乱角の不確定性を示す散乱角分布関数に基づき前記円錐面に厚みを持たせることで円錐面領域を設定する円錐面領域設定手段と、
前記イベントごとに且つ前記画素ごとに、前記コンプトン散乱位置を中心とし当該画素の位置及び前記コンプトン散乱位置間の距離を半径とする球面と前記円錐面領域と前記画像空間との交差状態に基づいて、前記確率パラメータを設定する確率パラメータ設定手段と、を有する
ことを特徴とする請求項1に記載の画像化装置。 - 前記確率パラメータ設定手段は、前記イベントごとに且つ前記画素ごとに、前記球面と前記円錐面領域と前記画像空間との交差領域に配置された各画素を抽出して、抽出した各画素に対して前記散乱角分布関数の強度に応じた指標を導出し、抽出した各画素について導出された指標の総和に基づき前記確率パラメータを設定する
ことを特徴とする請求項2に記載の画像化装置。 - 前記演算手段は、前記イベントごとに前記円錐面領域内の各画素に対して前記確率パラメータを設定する
ことを特徴とする請求項2又は3に記載の画像化装置。 - 前記演算手段は、各イベントでのコンプトン散乱の検出感度を示す検出感度パラメータを各イベントの前記測定データに基づき前記複数の画素の夫々に対して個別に設定し、設定した前記検出感度パラメータ及び前記確率パラメータを用いて前記分布画像を生成する
ことを特徴とする請求項1~4の何れかに記載の画像化装置。 - 前記演算手段は、前記イベントごとに且つ前記画素ごとに、コンプトン散乱位置と当該画素の位置とに応じて前記検出感度パラメータを設定する検出感度パラメータ設定手段を有する
ことを特徴とする請求項5に記載の画像化装置。 - 前記検出感度パラメータ設定手段は、前記イベントごとに且つ前記画素ごとに、前記コンプトン散乱位置及び当該画素の位置に基づく前記前段検出器内におけるコンプトン散乱前のガンマ線の飛程と、前記コンプトン散乱位置及び当該画素の位置間の距離とに基づき、前記検出感度パラメータを設定する
ことを特徴とする請求項6に記載の画像化装置。 - 前記演算手段は、各イベントでのコンプトン散乱の検出感度を示す検出感度パラメータを各イベントの前記測定データに基づき前記複数の画素の夫々に対して個別に設定する検出感度パラメータ設定手段を有して、前記検出感度パラメータ及び前記確率パラメータを用いて前記分布画像を生成し、
前記検出感度パラメータ設定手段は、前記イベントごとに且つ前記画素ごとに、前記コンプトン散乱位置及び当該画素の位置に基づく前記前段検出器内におけるコンプトン散乱前のガンマ線の飛程と、前記コンプトン散乱位置及び当該画素の位置間の距離とに基づき、前記検出感度パラメータを設定し、
前記演算手段は、
λj (l)、λj (l+1)は、夫々、第l回目、第(l+1)回目の反復計算によって得られる、前記分布画像の第j番目の画素の画素値を表し、
tij、vij、sijは、夫々、第iイベントの前記測定データに基づき、第iイベントにおける第j番目の画素に対して設定されるシステム応答関数、前記確率パラメータ、前記検出感度パラメータを表す
ことを特徴とする請求項2~4の何れかに記載の画像化装置。 - 各イベントにおいて、前記測定データにより、前記前段検出器内におけるガンマ線のコンプトン散乱位置及びコンプトン散乱により当該ガンマ線が失ったエネルギ、並びに、前記後段検出器内におけるガンマ線の光電吸収位置及び光電吸収エネルギが示される
ことを特徴とする請求項1~8の何れかに記載の画像化装置。 - 後段検出器と、
ガンマ線源と前記後段検出器との間に配置された前段検出器と、
前記前段検出器でコンプトン散乱されたガンマ線が前記後段検出器にて光電吸収されるイベントを検出するイベント検出手段と、
複数のイベントにおける各検出器とガンマ線との相互作用の測定データに基づき、前記ガンマ線源の空間分布を示す分布画像を生成する演算手段と、を備えた画像化装置において、
前記演算手段は、各イベントでのコンプトン散乱の検出感度を示す検出感度パラメータを各イベントの前記測定データに基づき前記分布画像内の複数の画素の夫々に対して個別に設定し、設定した前記検出感度パラメータを用いて前記分布画像を生成する
ことを特徴とする画像化装置。 - 前記演算手段は、前記イベントごとに且つ前記画素ごとに、コンプトン散乱位置と当該画素の位置とに応じて前記検出感度パラメータを設定する検出感度パラメータ設定手段を有する
ことを特徴とする請求項10に記載の画像化装置。 - 前記検出感度パラメータ設定手段は、前記イベントごとに且つ前記画素ごとに、前記コンプトン散乱位置及び当該画素の位置に基づく前記前段検出器内におけるコンプトン散乱前のガンマ線の飛程と、前記コンプトン散乱位置及び当該画素の位置間の距離とに基づき、前記検出感度パラメータを設定する
ことを特徴とする請求項11に記載の画像化装置。 - 前記演算手段は、
前記測定データに基づいて、前記イベントごとに、コンプトン散乱位置を頂点とし且つコンプトン散乱角を半頂角とし且つコンプトン散乱後のガンマ線の飛行経路の直線上に中心軸を持った円錐面を設定する円錐面設定手段と、
前記イベントごとに、前記コンプトン散乱角の不確定性を示す散乱角分布関数に基づき前記円錐面に厚みを持たせることで円錐面領域を設定する円錐面領域設定手段と、を有し、前記イベントごとに前記円錐面領域内の各画素に対して前記検出感度パラメータを設定する
ことを特徴とする請求項10~12の何れかに記載の画像化装置。 - 各イベントにおいて、前記測定データにより、前記前段検出器内におけるガンマ線のコンプトン散乱位置及びコンプトン散乱により当該ガンマ線が失ったエネルギ、並びに、前記後段検出器内におけるガンマ線の光電吸収位置及び光電吸収エネルギが示される
ことを特徴とする請求項10~13の何れかに記載の画像化装置。 - 後段検出器及びガンマ線源と前記後段検出器との間に配置された前段検出器を備えたコンプトンカメラに利用される方法であって、前記前段検出器でコンプトン散乱されたガンマ線が前記後段検出器にて光電吸収される各イベントでの各検出器とガンマ線との相互作用の測定データに基づき、前記ガンマ線源の空間分布を示す分布画像を、前記ガンマ線源を内包する画像空間内の画像として生成する画像化方法において、
各イベントでコンプトン散乱したガンマ線が前記画像空間内から到来した確率を示す確率パラメータを各イベントの前記測定データに基づき前記分布画像内の複数の画素の夫々に対して個別に設定し、設定した前記確率パラメータを用いて前記分布画像を生成する
ことを特徴とする画像化方法。 - 後段検出器及びガンマ線源と前記後段検出器との間に配置された前段検出器を備えたコンプトンカメラに利用される方法であって、前記前段検出器でコンプトン散乱されたガンマ線が前記後段検出器にて光電吸収される各イベントでの各検出器とガンマ線との相互作用の測定データに基づき、前記ガンマ線源の空間分布を示す分布画像を生成する画像化方法において、
各イベントでのコンプトン散乱の検出感度を示す検出感度パラメータを各イベントでの前記測定データに基づき前記分布画像内の複数の画素の夫々に対して個別に設定し、設定した前記検出感度パラメータを用いて前記分布画像を生成する
ことを特徴とする画像化方法。
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2016546612A JP6607576B2 (ja) | 2014-09-05 | 2015-08-28 | 画像化装置及び方法 |
EP15837605.3A EP3190431B1 (en) | 2014-09-05 | 2015-08-28 | Imaging device and method |
US15/509,112 US9784857B2 (en) | 2014-09-05 | 2015-08-28 | Imaging device and method |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2014181498 | 2014-09-05 | ||
JP2014-181498 | 2014-09-05 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2016035706A1 true WO2016035706A1 (ja) | 2016-03-10 |
Family
ID=55439769
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2015/074427 WO2016035706A1 (ja) | 2014-09-05 | 2015-08-28 | 画像化装置及び方法 |
Country Status (4)
Country | Link |
---|---|
US (1) | US9784857B2 (ja) |
EP (1) | EP3190431B1 (ja) |
JP (1) | JP6607576B2 (ja) |
WO (1) | WO2016035706A1 (ja) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2018036234A (ja) * | 2016-09-02 | 2018-03-08 | 国立大学法人群馬大学 | 情報処理装置、イメージング方法、及び、イメージングプログラム |
JP2020016508A (ja) * | 2018-07-24 | 2020-01-30 | キヤノン株式会社 | 放射線検出器 |
JP2020529903A (ja) * | 2017-06-15 | 2020-10-15 | コンセッホ スペリオル デ インベスティガシオンス サイエンティフィカス(シーエスアイシー) | 腫瘍学的診断およびリアルタイムガイド下生検に適したデュアル画像システム |
CN112512427A (zh) * | 2018-08-07 | 2021-03-16 | 美国西门子医疗系统股份有限公司 | 用于医学成像的自适应康普顿摄像机 |
JP7578450B2 (ja) | 2020-10-07 | 2024-11-06 | キヤノンメディカルシステムズ株式会社 | 核医学診断装置 |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP6551003B2 (ja) * | 2015-07-24 | 2019-07-31 | 三菱重工業株式会社 | 放射線測定装置及び放射線測定方法 |
CN111880211B (zh) * | 2020-06-08 | 2022-11-15 | 中国科学院高能物理研究所 | 一种利用康普顿散射事例统计进行放射性核素识别的方法 |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2010048699A (ja) * | 2008-08-22 | 2010-03-04 | Gunma Univ | コンプトンカメラ |
Family Cites Families (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5861627A (en) * | 1997-06-24 | 1999-01-19 | The University Of Utah | Image reconstruction for compton camera including spherical harmonics |
US20050139775A1 (en) * | 2003-12-26 | 2005-06-30 | Riken | Gamma-ray detector and gamma-ray image pickup apparatus |
WO2006058432A1 (en) * | 2004-12-01 | 2006-06-08 | Triumf, Operating As A Joint Venture By The Governors Of The University Of Alberta, The University Of British Columbia, Carleton | System for selecting true coincidence events in positron emission tomography |
JP4486623B2 (ja) | 2006-08-11 | 2010-06-23 | 独立行政法人理化学研究所 | コンプトン撮像カメラ |
JPWO2008035708A1 (ja) * | 2006-09-21 | 2010-01-28 | 株式会社日立メディコ | 核医学診断装置 |
US8515011B2 (en) * | 2009-06-02 | 2013-08-20 | Mayo Foundation For Medical Education And Research | System and method for dose verification radiotherapy |
US10088580B2 (en) * | 2012-05-31 | 2018-10-02 | Minnesota Imaging And Engineering Llc | Detector systems for radiation imaging |
-
2015
- 2015-08-28 EP EP15837605.3A patent/EP3190431B1/en active Active
- 2015-08-28 WO PCT/JP2015/074427 patent/WO2016035706A1/ja active Application Filing
- 2015-08-28 JP JP2016546612A patent/JP6607576B2/ja active Active
- 2015-08-28 US US15/509,112 patent/US9784857B2/en active Active
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2010048699A (ja) * | 2008-08-22 | 2010-03-04 | Gunma Univ | コンプトンカメラ |
Non-Patent Citations (2)
Title |
---|
TAKAHIRO IDA ET AL.: "Development of a list- mode iterative reconstruction for Ge semiconductor Compton camera with an accurate model of the uncertainty of the Compton scattering angle", JSAP AUTUMN MEETING KOEN YOKOSHU, vol. 75 th, no. 18a-B2-9, 1 September 2014 (2014-09-01), XP009500813 * |
YUKIO OKABE ET AL.: "Study of Implementation of Maximum-likelihood Reconstruction for Compton Camera", IEICE TECHNICAL REPORT, vol. 109, no. 407, 21 January 2010 (2010-01-21), pages 47 - 51, XP055423561 * |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2018036234A (ja) * | 2016-09-02 | 2018-03-08 | 国立大学法人群馬大学 | 情報処理装置、イメージング方法、及び、イメージングプログラム |
JP2020529903A (ja) * | 2017-06-15 | 2020-10-15 | コンセッホ スペリオル デ インベスティガシオンス サイエンティフィカス(シーエスアイシー) | 腫瘍学的診断およびリアルタイムガイド下生検に適したデュアル画像システム |
JP7226827B2 (ja) | 2017-06-15 | 2023-02-21 | コンセッホ スペリオル デ インベスティガシオンス サイエンティフィカス(シーエスアイシー) | 被験者の第1の画像および第2の画像を生成するシステム及びシステムの作動方法 |
JP2020016508A (ja) * | 2018-07-24 | 2020-01-30 | キヤノン株式会社 | 放射線検出器 |
JP7182930B2 (ja) | 2018-07-24 | 2022-12-05 | キヤノン株式会社 | 放射線検出器 |
CN112512427A (zh) * | 2018-08-07 | 2021-03-16 | 美国西门子医疗系统股份有限公司 | 用于医学成像的自适应康普顿摄像机 |
CN112512427B (zh) * | 2018-08-07 | 2024-05-24 | 美国西门子医疗系统股份有限公司 | 用于医学成像的自适应康普顿摄像机 |
JP7578450B2 (ja) | 2020-10-07 | 2024-11-06 | キヤノンメディカルシステムズ株式会社 | 核医学診断装置 |
Also Published As
Publication number | Publication date |
---|---|
US20170261624A1 (en) | 2017-09-14 |
EP3190431A4 (en) | 2018-04-18 |
EP3190431A1 (en) | 2017-07-12 |
JP6607576B2 (ja) | 2019-11-20 |
JPWO2016035706A1 (ja) | 2017-07-13 |
EP3190431B1 (en) | 2019-03-20 |
US9784857B2 (en) | 2017-10-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP6607576B2 (ja) | 画像化装置及び方法 | |
JP5582370B2 (ja) | ガンマ線を利用する画像化装置、画像信号処理装置およびガンマ線測定データの画像処理方法 | |
US7888651B2 (en) | Method and system for using tissue-scattered coincidence photons for imaging | |
Tashima et al. | 3D Compton image reconstruction method for whole gamma imaging | |
Etxebeste et al. | CCMod: a GATE module for Compton camera imaging simulation | |
Pedemonte et al. | A machine learning method for fast and accurate characterization of depth-of-interaction gamma cameras | |
Feng et al. | Influence of Doppler broadening model accuracy in Compton camera list-mode MLEM reconstruction | |
JPWO2017209059A1 (ja) | ガンマ線画像取得装置およびガンマ線画像取得方法 | |
Ida et al. | Accurate modeling of event-by-event backprojection for a germanium semiconductor Compton camera for system response evaluation in the LM-ML-EM image reconstruction method | |
Yao et al. | Study of 3D fast Compton camera image reconstruction method by algebraic spatial sampling | |
Ding et al. | Charged‐particle emission tomography | |
Wu et al. | An accurate probabilistic model with detector resolution and Doppler broadening correction in list-mode MLEM reconstruction for Compton camera | |
Borja-Lloret et al. | Influence of the background in Compton camera images for proton therapy treatment monitoring | |
Wang et al. | Enhancing spatial resolution of 18F positron imaging with the Timepix detector by classification of primary fired pixels using support vector machine | |
Hirano et al. | Monte Carlo simulation of scintillation photons for the design of a high-resolution SPECT detector dedicated to human brain | |
El Bitar et al. | A detector response function design in pinhole SPECT including geometrical calibration | |
Lin et al. | Recovering the triple coincidence of non-pure positron emitters in preclinical PET | |
Saaidi et al. | Monte Carlo simulation of two siemens biograph PET/CT system using gate: image quality performance | |
Calderón et al. | Design, development, and modeling of a compton camera tomographer based on room temperature solid statepixel detector | |
Giovagnoli | Image reconstruction for three-gamma PET imaging | |
Liang et al. | Spatial resolution recovery utilizing multi-ray tracing and graphic processing unit in PET image reconstruction | |
Chiang et al. | Time of flight dual photon emission computed tomography | |
Feng | Modeling and regularization in tomographic reconstruction for Compton camera imaging | |
Kagaya et al. | Evaluating the capability of detecting recoil-electron tracks using an electron-tracking Compton camera with a silicon-on-insulator pixel sensor | |
Kalaitzidis | From Monte Carlo PET Simulations to Reconstructed Images: Modelling and Optimisation for 68Ga Theragnostics |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 15837605 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2016546612 Country of ref document: JP Kind code of ref document: A |
|
REEP | Request for entry into the european phase |
Ref document number: 2015837605 Country of ref document: EP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2015837605 Country of ref document: EP |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
WWE | Wipo information: entry into national phase |
Ref document number: 15509112 Country of ref document: US |