JP2013053992A - Radiation imaging apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a radiation imaging apparatus that improves correction accuracy, and realizes high-speed image reconstruction processing.SOLUTION: A radiation imaging apparatus 1 includes: a plurality of detectors 21 that operate for each of through-holes 27 of a collimator 26; separation image generating means 12 for separating an image obtained by the detectors 21 for every detector 21 having equal relative position to the through-hole 27 to generate separation images; and image processing means 12 for performing image processing for every separation image generated by the separation image generating means 12.

Description

  The present invention relates to a radiation imaging apparatus that images an incident radiation distribution.

  As an apparatus that applies a radiation measuring apparatus to the nuclear medicine field, there is a single photon emission computed tomography apparatus (hereinafter referred to as a SPECT (Single Photon Emission Computed Tomography) apparatus) using a gamma camera. This SPECT apparatus measures the distribution of a compound containing a radioisotope and provides an image of a tomographic plane.

  The current SPECT apparatus mainly uses a detector combining a scintillator made of a single crystal and a plurality of photomultiplier tubes as a radiation (gamma ray) detector. In these SPECT apparatuses, the incident position of radiation is obtained by calculating the center of gravity. However, the position resolution of about 10 mm is the limit in the center of gravity calculation, and a SPECT apparatus having a higher position resolution is required for improving the diagnostic ability.

  In recent years, pixel-type detectors have been developed as detectors with higher resolution. The pixel-type detector acquires a position signal in small detector units, that is, pixel units, and the pixels are composed of divided scintillators and semiconductors. The intrinsic resolution of the detector is determined by the pixel size, and a pixel size of 1 to 2 mm is currently being developed.

  Further, as a technique for increasing the resolution, a tomographic image reconstruction method has been developed and improved. Up to now, the filtered backprojection method (FBP method: Filtered Back-Projection method), the successive approximation method without resolution correction (maximum likelihood estimation expectation maximization method (MLEM method: Maximum Likelihood Expectation Maximization method), and the subset are used. The expected value maximization method (OSEM method: Ordered Subset Expectation Maximization method) has been used, but in recent years, a successive approximation method with resolution correction has been developed (see Non-Patent Document 1). Although this method can reconstruct an image in consideration of the geometric shape of the collimator and the detector and provide a more accurate image, there are drawbacks such as a complicated calculation.

In the clinic, a SPECT apparatus having high sensitivity in addition to high position resolution is required. In order to obtain high position resolution, it is necessary to limit the direction of radiation incident on the detector with a collimator. For this purpose, the field of view in which the detector looks at the measurement object may be narrowed by a collimator. As such a collimator, for example, a LEHR (Low Energy High Resolution) collimator is known. However, in the LEHR collimator, sensitivity is sacrificed due to the limited field of view. On the other hand, in order to obtain high sensitivity, it is necessary to increase the size of the through hole of the collimator. As such a collimator, a LEGP (Low Energy General Purpose) collimator and a LEHS (Low Energy High Sensitivity) collimator are known. Yes. However, in these collimators, the position resolution is deteriorated.
Thus, in the conventional SPECT apparatus, position resolution and sensitivity are not compatible due to trade-off. For this reason, the appearance of a new technology that achieves both position resolution and sensitivity is desired.

  Therefore, a SPECT apparatus of a type in which a plurality of detectors are included for one rectangular through hole of a collimator has been invented as a SPECT apparatus that achieves both sensitivity and position resolution. In this SPECT apparatus, when the through-holes of the collimator are equal in size, it is possible to obtain a higher position resolution than the conventional SPECT apparatus in which the through-hole and the detector have a one-to-one correspondence (Patent Document 1, Non-Patent Document 2). reference).

  In addition, pixel-type detectors composed of a plurality of pixels (detectors) have output variations between pixels even when they receive the same level of radiation. In Patent Document 2, there are adjacent pixels (adjacent detection) for defective pixels (bad pixels) such as inactive pixels whose output level is significantly lower than other regions and active pixels whose output level is significantly higher than other regions. Pixel interpolation using a synthesizer) is disclosed.

Special table 2010-507090 gazette JP 2003-023572 A

Panin VY, et al, "Fully 3-D PET reconstruction with system matrix derived from point source measurements," IEEE Trans Med Imaging. 2006 Jul; 25 (7): 907-921 C. Robert et al. (2008) 2008 IEEE Nuclear Science Symposium Conference Record Vol6 pp.4246-4251

  However, when using a type of collimator that includes multiple detectors for a single through-hole, it is necessary to perform image reconstruction by combining resolution correction, and project each individual detector with a different response function. It is necessary to calculate. For this reason, there is a problem that the correction accuracy is reduced and the reconstruction time is increased due to the complicated calculation.

  Further, when a collimator of a type including a plurality of detectors for one through hole is used, since the response functions of the individual detectors are not the same, Patent Document 2 deals with defective pixels (defective pixels). In the pixel interpolation using the simple adjacent pixel (adjacent detector) disclosed in FIG.

  Therefore, an object of the present invention is to provide a radiation imaging apparatus capable of improving correction accuracy and speeding up image reconstruction processing.

  In order to solve such a problem, the present invention provides a radiation imaging apparatus in which a plurality of detectors correspond to a through-hole of a collimator, and an image acquired by the detector is relative to the through-hole. A separate image generating unit that generates a separated image by separating the detectors at the same position; and an image processing unit that performs image processing for each separated image generated by the separated image generating unit. A radiation imaging apparatus.

  According to the present invention, it is possible to provide a radiation imaging apparatus capable of improving correction accuracy and speeding up image reconstruction processing.

It is a block diagram of the SPECT apparatus which concerns on this embodiment. It is the perspective view which showed arrangement | positioning of a detector and a collimator. (A) is the figure which planarly viewed arrangement | positioning of a detector and a collimator from the orthogonal | vertical direction with respect to the entrance plane of a detector, (b) is the figure which showed the partial detector group. It is the figure which showed the example of the response function of a detector. It is the perspective view which showed other arrangement | positioning of a detector and a collimator. (A) is the figure which planarly viewed other arrangement | positioning of a detector and a collimator from the orthogonal | vertical direction with respect to the entrance plane of a detector, (b) is the figure which showed the partial detector group.

  Hereinafter, modes for carrying out the present invention (hereinafter referred to as “embodiments”) will be described in detail with reference to the drawings as appropriate. In each figure, common portions are denoted by the same reference numerals, and redundant description is omitted.

<< SPECT apparatus (radiation imaging apparatus) 1 >>
An overall configuration of a SPECT apparatus (radiation imaging apparatus) 1 according to the present embodiment will be described with reference to FIG. FIG. 1 is a configuration diagram of a SPECT apparatus 1 according to the present embodiment.

The SPECT device 1 includes a gantry 10, cameras 11A and 11B, a data processing device 12, a display device 13, and a bed 14. The subject 15 is administered a radioactive drug, for example, a drug containing ^99m Tc with a half-life of 6 hours. Gamma rays (radiation) emitted from ^99m Tc in the body of the subject 15 placed on the bed 14 are detected by the cameras 11A and 11B supported by the gantry 10, and a tomographic image is taken.

<Cameras 11A and 11B>
The cameras 11 </ b> A and 11 </ b> B can move in the radial direction and the circumferential direction of a cylindrical opening provided in the central portion of the gantry 10. At the time of tomographic imaging, the camera 11 rotates around the subject 15 around an attachment portion (not shown) with the gantry 10 and is generated from a radiopharmaceutical accumulated in a tumor or the like in the body of the subject 15. Detect lines.

The cameras 11A and 11B incorporate a collimator 26 and a number of detectors 21. The collimator 26 has a through-hole 27 and a septa 28 that partitions the through-hole 27, and selects γ-rays emitted from ^99m Tc in the body of the subject 15 (regulates the incident angle), and in a certain direction. It has the role of passing only γ rays. And the gamma ray which passed the collimator 26 (through-hole 27) is detected by the detector 21. FIG.

  The cameras 11A and 11B include an application specific integrated circuit (hereinafter referred to as an ASIC (Application Specific Integrated Circuit)) 25 for measuring a γ-ray detection signal, a detector board 23, an ASIC board 24, It has. The detection signal of the γ-ray detected by the detector 21 is input to the ASIC 25 via the detector substrate 23 and the ASIC substrate 24, and the ID of the detector 21 that detected the γ-ray, the peak value of the detected γ-ray, Converted to detection time data. The data (packet data) converted by the ASIC 25 is transmitted to the data processing device 12.

  The cameras 11 </ b> A and 11 </ b> B include a light shielding / γ-ray / electromagnetic shield 29. The light shielding / γ-ray / electromagnetic shield 29 is made of iron, lead, or the like, and surrounds the detector 21, the detector substrate 23, the ASIC substrate 24, the ASIC 25, and the collimator 26. It is designed to block electromagnetic waves.

<Data processing device 12, display device 13>
The data processing device 12 includes a storage device (not shown) and a tomogram information creation device (not shown). The data processing device 12 takes in the packet data including the detected peak value of γ-rays, detection time data and detector (channel) ID from the ASIC 25, generates a plane image, and further generates tomographic image information by image reconstruction. Can be done. Further, the data processing device 12 can display the generated plane image and tomographic image information on the display device 13.

In this way, the SPECT apparatus 1 can image the radiopharmaceutical (γ rays emitted from ^99m Tc) accumulated in the tumor or the like in the body of the subject 15 and identify the position of the tumor or the like. It has become.

≪Detector 21 and collimator 26≫
Next, the detector 21 and the collimator 26 used in the cameras 11A and 11B of the SPECT apparatus 1 according to the present embodiment will be further described with reference to FIGS. FIG. 2 is a perspective view showing the arrangement of the detector 21 and the collimator 26.

<Detector 21>
The detector 21 is a pixel type semiconductor detector using a CdTe semiconductor, for example, and is divided for each pixel as shown in FIG. Each detector 21 constitutes one pixel, and a large number of detectors 21 are arranged two-dimensionally to constitute a detector group 22. Therefore, unlike the scintillator made of one large crystal, the γ-ray detection signal is collected for each detector 21 unit, that is, for each pixel unit. For convenience of obtaining the response function, the detector group 22 (array of detectors 21) preferably has a periodic structure.

  In the description of the pixel-type detector, the terms “detector” and “detector group” are used. The detector means one pixel having an arbitrary shape, and the detector group is a detector. Let's say an ordered collection.

  Although the SPECT apparatus 1 according to the present embodiment will be described assuming that the detector group 22 is physically divided into pixels as shown in FIG. 2, the present invention is not limited to this. For example, a plurality of electrodes may be arranged and a large detector may be electrically divided into pixels, and a readout method such as a double-sided silicon strip detector (DSSD) should be devised. It may also be a detector that pixelates.

<Collimator 26>
The collimator 26 is made of lead, and has a through hole 27 in a direction that can be seen when viewed in a plan view from the direction perpendicular to the incident surface of the detector 21 (incident direction of gamma rays, minus direction of the z axis shown in FIG. 2). The through holes 27 are arranged in a grid pattern. Each through hole 27 is partitioned by a septa 28.

FIG. 3A shows the arrangement of the detector 21 (detector group 22) and the collimator 26 in a direction perpendicular to the incident surface of the detector 21 (detector group 22) (incident direction of gamma rays, z shown in FIG. It is the figure seen planarly from the axis | shaft minus direction.
In the present embodiment, as shown in FIG. 3A, a case where four detectors 21 (21a, 21b, 21c, 21d) are included in one through hole 27 of the collimator 26 will be described.

  Note that the arrangement of the through holes 27 and the detectors 21 of the collimator 26 of the SPECT apparatus 1 according to the present embodiment is 2 × 2 with respect to one through hole 27 as shown in FIGS. However, the present invention is not limited to this. For example, an integer number of detectors 21 may be included in one through-hole 27 such as 9 in 3 × 3 or 8 in 2 × 4. More generally, if the through-hole 27 and the detector 21 maintain an integer ratio, a block composed of a plurality of through-holes 27 is regarded as one through-hole, with respect to one through-hole. An integer number of detectors 21 may be included.

<< Response function >>
In the SPECT apparatus 1 according to the present embodiment, as shown in FIG. 3A, four detectors 21 (21 a, 21 b, 21 c, 21 d) correspond to one through hole 27 of the collimator 26. .
The data processing device 12 of the SPECT apparatus 1 according to the present embodiment has a relative position with respect to the through hole 27 as shown in FIGS. 3A and 3B from the planar image generated by the data processing device 12. Select a detector having the same relative position as the detector 21a (in FIG. 3 (a), a shaded detector; an upper right detector among the four detectors included for each through-hole 27). The partial detector group 22A is taken out, and the data of the partial detector group 22A is taken out and a partial image (separated image) of the partial detector group 22A is created.
Similarly, a detector whose relative position with respect to the through hole 27 has the same relative position as that of the detector 21b (in FIG. 3A, the upper left detector among the four detectors included for each through hole 27). Is selected as the partial detector group 22B, the data of the partial detector group 22B is taken out, and a partial image (separated image) of the partial detector group 22B is created. Select a detector having the same relative position to the through hole 27 as the detector 21c (in FIG. 3A, the lower right detector among the four detectors included for each through hole 27). Thus, the partial detector group 22C is obtained, and the data of the partial detector group 22C is taken out to create a partial image (separated image) of the partial detector group 22C. A detector having the same relative position to the through hole 27 as the detector 21d (in FIG. 3A, the lower left detector among the four detectors included for each through hole 27) is selected. The partial detector group 22D is taken out, and the data of the partial detector group 22D is taken out to create a partial image (separated image) of the partial detector group 22D.

  Since this partial image (separated image) is created for each relative position between the detector 21 and the through-hole 27 of the collimator 26, the number of the detectors 21 included in one through-hole 27 (integer number) and The same number of partial images (separated images) are created. That is, in the SPECT apparatus 1 according to the present embodiment, four partial images (separated images) are created.

  FIG. 4 is a diagram illustrating an example of the response function 30 (30A, 30B) of the detector 21. Here, for the sake of simplicity, a one-dimensional system will be discussed as shown in FIG. The actual detector 21 and collimator 26 are two-dimensional, and this one-dimensional consideration may be extended to two dimensions.

4A is a diagram for explaining a response function 30A of the partial detector group 22A including the detector 21a (see FIG. 3B), and FIG. 4B is a partial detector group including the detector 21b. It is a figure explaining the response function 30B of 22B (refer FIG.3 (b)).
As shown in FIG. 4A, the response function 30A of the detector 21 included in the partial detector group 22A has the same shape in all pixels. Similarly, as shown in FIG. 4B, the response function 30B of the detector 21 included in the partial detector group 22B has the same shape in all the pixels.

The response function 30 is determined by the geometry of the detector 21 and the collimator 26 (through hole 27). For this reason, in the detector group 22 (see FIG. 2) as a whole, the shape of the response function 30 is inverted between the odd-numbered and even-numbered detectors 21, so that the response functions 30 are not all the same.
On the other hand, by separating each of the partial detector groups 22A and 22B, the response functions 30A in the partial image (separated image) of the partial detector group 22A are all the same, and the partial image of the partial detector group 22B ( The response functions 30B in the separated image) are all the same.

<Pixel interpolation>
The correction (pixel interpolation) of the detector 21 in the SPECT apparatus 1 according to the present embodiment will be described.
The SPECT apparatus 1 uses several thousand (for example, 64 × 64) detectors 21 to acquire an image. It is difficult to make all of such a large number of detectors 21 operate normally, and a defect (inactive or excessively active) of the detector 21 causes a lack of image.

  In order to correct this chipping, the data processing device 12 of the SPECT apparatus 1 according to the present embodiment is a partial image (separated image) of the partial detector group (22A, 22B, 22C, 22D) having the same relative position to the through hole 27. Interpolation is performed from adjacent detectors 21 in FIG. The interpolation method is a conventionally known interpolation such as an average value of adjacent detectors 21 except that the values of adjacent detectors 21 in the partial detector group (22A, 22B, 22C, 22D) are used. The method can be used.

Note that the correction (pixel interpolation) of the conventional SPECT apparatus as shown in Patent Document 2 is to perform interpolation based on data of physically adjacent detectors 21.
However, in the configuration in which an integer number of detectors 21 correspond to one through hole 27 as in the SPECT apparatus 1 according to the present embodiment, the response functions are different in physically adjacent detectors 21. If the correction (pixel interpolation) of the conventional SPECT apparatus (see FIG. 4) is used as it is, correct correction cannot be performed.

  On the other hand, the correction (pixel interpolation) of the SPECT apparatus 1 according to the present embodiment uses the detector 21 having the same position relative to the through-hole 27, so that the same response function 30 as the detector 21 to be corrected is used. Since the data of the detector 21 in the vicinity is used, the correction accuracy is improved.

<Image reconstruction>
Next, image reconstruction for creating a tomographic image in the SPECT apparatus 1 according to the present embodiment will be described.
In image reconstruction that performs resolution correction (for example, see Non-Patent Document 1), a response function 30 determined by the arrangement of the detector 21 and the collimator 26 is obtained by experiment or ray-trace simulation, and a sequential image reconstruction loop is performed. Include. In the sequential image reconstruction, a certain hypothetical image is determined, the projection of the hypothetical image is compared with the actually obtained image, and the operation of making corrections so as to match is repeated.

  In conventional image reconstruction (FBP method, MLEM method, OSEM method, etc.), resolution degradation of the detector 21 is not taken into account when the image is projected. On the other hand, in image reconstruction with resolution correction (for example, see Non-Patent Document 1), degradation of resolution is taken into account, that is, a response determined by the arrangement of the detector 21 and the collimator 26 with respect to an assumed image. A projection image is created by convolution integration of the function 30.

  Note that the image reconstruction for creating the tomographic image in the conventional SPECT apparatus regards the entire image of the detector group 22 (see FIG. 2) as one image, and the response function 30 of the detector 21 is the same. Therefore, it is necessary to calculate as defined in order to perform convolution integration, and the amount of calculation is very large.

On the other hand, in the SPECT apparatus 1 according to the present embodiment, the hypothetical image and the response function 30 are Fourier-transformed using Fast Fourier Transform (FFT) or the like, and a product is obtained on the frequency space, and then the inverse is performed. The image is returned to the projected image using Fourier transform.
Here, the SPECT apparatus 1 according to the present embodiment performs convolution integration for each partial image (separated image) of the partial detector group (22A, 22B, 22C, 22D). In the partial detector groups (22A, 22B, 22C, 22D), since the response functions 30 are equal, an algorithm using a fast Fourier transform or the like can be used, and high-speed convolution integration is possible. That is, in the SPECT apparatus 1 according to the present embodiment, the reconfiguration time can be shortened.

≪Modification≫
The SPECT apparatus 1 according to the present embodiment is not limited to the configuration of the above-described embodiment, and various modifications can be made without departing from the spirit of the invention.

  As shown in FIG. 3A, the SPECT apparatus 1 according to this embodiment is perpendicular to the incident surface of the detector 21 (detector group 22) (incident direction of gamma rays, z-axis shown in FIG. Although it has been described that the septa 28 of the collimator 26 is arranged on the boundary line between the detectors 21 when viewed in plan from the minus direction), the present invention is not limited to this.

As shown in FIG. 5 and FIG. 6A, it is seen in plan view from the direction perpendicular to the incident surface of the detector 21 (detector group 22) (incident direction of gamma rays, minus direction of z-axis shown in FIG. 5). At this time, the configuration is such that the septa 28 of the collimator 26 passes through the center of the detector 21 and the boundary line between the detectors 21 and the septa 28 are orthogonal to each other (see JP 2011-106887 A). Also good.
Even in such an arrangement of the detector 21 (detector group 22) and the collimator 26, as shown in FIGS. 6A and 6B, the detector 21 and the collimator 26 (through hole 27). The partial detector group 22A can be extracted by collecting the detectors 21 having the same correspondence relationship with each other, that is, having the same response function. Thereby, it can process similarly to the SPECT apparatus 1 which concerns on this embodiment.

  Moreover, the structure arrange | positioned so that the arrangement | sequence of the through-hole 27 of the collimator 26 may cross | intersect the arrangement | sequence of the detector 21 at a predetermined angle may be sufficient.

1 SPECT equipment (radiation imaging equipment, single photon emission computed tomography equipment)
10 Gantry 11A, 11B Camera 12 Data processing device (separated image generating means, image processing means)
13 Display device 14 Bed 15 Subject 21, 21a, 21b, 21c, 21d Detector 22 Detector group 22A, 22B, 22C, 22D Partial detector group 23 Detector board 24 ASIC board 25 ASIC (Application-specific integrated circuit) )
26 Collimator 27 Through-hole 28 Septa 29 Light shielding / γ-ray / electromagnetic shield 30, 30A, 30B Response function

Claims (4)

A radiation imaging apparatus in which a plurality of detectors correspond to the through-hole of the collimator,
A separated image generating means for separating the image acquired by the detector for each detector having the same relative position with respect to the through-hole, and generating a separated image;
An image processing unit that performs image processing for each separated image generated by the separated image generating unit. The radiation imaging apparatus according to claim 1, wherein interpolation of the detector is performed based on a separated image separated for each detector having the same relative position to the through hole. The image processing means includes
The radiographic imaging apparatus according to claim 1, wherein during image reconstruction for creating a tomographic image, a back projection operation is performed for each separated image separated for each detector having the same relative position to the through hole. . The image processing means includes
Convolution integration using Fourier transform for each separated image separated for each detector with the same relative position with respect to the through-hole, using resolution correction for image reconstruction to create a tomographic image, during back projection and forward projection The radiation imaging apparatus according to claim 3, wherein:

Images