JP5342672B2 - Semiconductor detector interpolation apparatus and method - Google Patents

Description

  The present invention relates to an interpolating apparatus for a semiconductor detector used in an X-ray semiconductor imaging apparatus such as a gamma-ray imaging used for medical nuclear medicine inspection, a non-destructive inspection using industrial gamma-rays, and an observation satellite, and a method therefor.

  A semiconductor detector is used for gamma ray imaging. This semiconductor detector detects gamma rays and outputs an energy signal thereof. One semiconductor chip is used for this semiconductor detector. In this semiconductor chip, a large number of pixel-type semiconductor elements are provided in the vertical and horizontal directions. Each of these pixel type semiconductor elements detects gamma rays and outputs an energy signal. The pixel type semiconductor element outputs a position identification signal. Each of these pixel-type semiconductor elements corresponds to each pixel, and each energy signal output from these semiconductor elements is obtained as each pixel value (pixel data).

  However, these pixel-type semiconductor elements currently have large variations in pixel values. For this reason, if the average value of all the pixel values is obtained and the pixel value is far from the average value, the pixel is judged to be missing and the pixel value is discarded, or each pixel around the missing pixel Is used to interpolate the missing pixel value.

  When the number of defective pixels increases in the semiconductor detector, the image quality of the image data acquired from the energy signal output from the semiconductor detector decreases. When the image quality degradation of the image data is large, the semiconductor detector cannot be used as a defective product. On the other hand, the pixel-type semiconductor element has a poor manufacturing yield. If only a pixel-type semiconductor element with good performance is selected, the cost of the pixel-type semiconductor element itself increases. Under such circumstances, there is a demand for a method for using the current pixel type semiconductor device more effectively.

JP 11-337645 A

  SUMMARY OF THE INVENTION An object of the present invention is to provide an interpolation apparatus and method for a semiconductor detector, which can use the current pixel type semiconductor device better and effectively.

The present invention provides a radiation imaging apparatus having a semiconductor detector forming each pixel by a plurality of semiconductor elements, each energy signal as an energy spectrum data showing the energy intensity distribution of the radiation output from each of the respective semiconductor elements An analysis unit for analyzing the shape , a classification unit for classifying the performance of each semiconductor element into three based on an analysis result of the energy spectrum data by the analysis unit, and among the three classification results, and a correction unit for each interpolation the value of each pixel of the two said classification performance is classified as low.

The present invention relates to an interpolation method for a semiconductor detector provided in a radiation imaging apparatus, wherein each pixel is formed by a plurality of semiconductor elements, as energy spectrum data indicating the energy intensity distribution of radiation output from each semiconductor element. Analyzing the shape of each energy signal, classifying the performance of each semiconductor element into three based on the analysis result of the energy spectrum data, and classifying the performance of each semiconductor element as low among the three classification results. The values of the respective pixels of the two classifications are interpolated.

  ADVANTAGE OF THE INVENTION According to this invention, the interpolation apparatus and method of a semiconductor detector which can use the present pixel type semiconductor element better and effectively can be provided.

The block diagram which shows one Embodiment of the interpolation apparatus of the semiconductor detector used for the gamma ray imaging apparatus which concerns on this invention. The schematic diagram which shows the pixel value of the ideal semiconductor detector acquired by the pixel energy spectrum analyzer of the apparatus. The schematic diagram which shows the pixel value of the actual semiconductor detector acquired by the pixel energy spectrum analyzer of the apparatus. The schematic diagram which shows the classification | category of the normal, attention, and defect pixel of each pixel (pixel type semiconductor device) by the classification device of the same apparatus. The schematic diagram which shows the classification | category in the attention pixel by the classification device of the apparatus. The schematic diagram of the data table which memorize | stores the classification result by the classification device of the apparatus. The schematic diagram which shows the attention pixel and peripheral pixel for demonstrating the interpolation process by the interpolation processing apparatus of the apparatus. The schematic diagram of the weighting data used for the interpolation processing apparatus of the apparatus. The schematic diagram for demonstrating the specific example of the interpolation process by the interpolation processing apparatus of the apparatus.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 shows a configuration diagram of an interpolation device for a semiconductor detector. The semiconductor detector 1 is formed by arranging a plurality of pixel-type semiconductor elements 1a in the vertical and horizontal directions. This semiconductor detector 1 is used for a gamma ray imaging apparatus used for a medical nuclear medicine examination, for example. Each pixel type semiconductor element 1a detects a gamma ray g and outputs an energy signal. The semiconductor detector 1 is provided with a surface ray source 2 facing the semiconductor detector 1. The surface ray source 2 is provided to face the semiconductor detector 1 when classifying the performance of the plurality of pixel type semiconductor elements 1a. The surface ray source 2 emits gamma rays g uniformly to the opposing surface. As a result, uniform gamma rays g are incident on the semiconductor detector 1. The semiconductor detector 1 detects gamma rays g emitted from the surface source 2 with each pixel type semiconductor element 1a, and outputs each energy signal e1 from the pixel type semiconductor element 1a. These energy signals e1 are sent to the pixel energy spectrum analyzer 3 as uniformity measurement data. Each energy signal e1 is sent to the pixel energy spectrum analyzing apparatus 3 as energy spectrum data. The energy spectrum data indicates the energy intensity distribution of the gamma ray g.

  On the other hand, the semiconductor detector 1 is provided, for example, in a gamma-ray imaging apparatus used for a medical nuclear medicine examination and performs gamma-ray g from a medicine administered into a subject such as a human body for each pixel type. Detection is performed by the semiconductor element 1a, and each energy signal e1 is output from the pixel type semiconductor element 1a.

  The pixel energy spectrum analyzer 3 analyzes each energy signal e1 output from each pixel type semiconductor element 1a of the semiconductor detector 1. The pixel energy spectrum analyzer 3 collects and counts each energy signal e1 output from each pixel type semiconductor element 1a of the semiconductor detector 1, and obtains each pixel value (pixel data) from the collected count value. .

  If the semiconductor detector 1 is an ideal semiconductor detector and normal energy signals are output from all the pixel-type semiconductor elements 1a, the values of all the pixels of the semiconductor detector 1 are as shown in FIG. A constant value, for example, “100”. In the figure, each pixel of the semiconductor detector 1 is denoted by 1a. The value “100” of the pixel 1a is an average value of, for example, one-half to two-thirds of the total number of pixels in the semiconductor detector 1, and is assumed to be a pixel value for easy understanding. It is “100” and may be “50” or “30”. The actual semiconductor detector 1 varies for each pixel 1a, and takes values such as pixel values “100” “0” “50” “140” “1000” as shown in FIG.

  The pixel energy spectrum analyzer 3 analyzes the shape of each energy signal e1 output from each pixel type semiconductor element 1a on a display or the like, displays the peak position of each energy signal e1, and counts the peak positions. Analyze value, half width (FWHM).

  The classification device 4 classifies the performance of each pixel type semiconductor element 1a into at least three based on the analysis result of the pixel energy spectrum analysis device 3, for example, each pixel value for each pixel of the semiconductor detector 1 shown in FIG. (Hereinafter referred to as pixel classification). This classification device 4 can be used by correcting the first classification in which the performance of each pixel type semiconductor element 1a is normal and the energy signal e1 output from the pixel type semiconductor element 1a as three classifications, for example. The pixels are classified into the second classification and the third classification in which the energy signal e1 output from the pixel type semiconductor element 1a cannot be corrected and is abnormal.

Specifically, in the first classification C1, the energy signal e1 is normally output from the pixel-type semiconductor element 1a, and a normal image can be acquired by correcting the energy window setting for each image and the intensity of the energy signal e1. In the first classification C1, as shown in FIG. 4, the average value AV of the values of, for example, one-half to two-thirds of all the numbers of pixels in the semiconductor detector 1 is set as an intermediate value and straddles it. This is the set pixel value range. The pixels of the first classification C1 are referred to as normal pixels.
The second classification C2 has a large variation in energy resolution and intensity of the energy signal e1 compared to the first classification C1, but an image is acquired by performing interpolation using a plurality of pixel values around the pixel. Is possible. The second classification C2 is each pixel value range in which the pixel value is set larger on the positive and negative side than the pixel value range of the first classification C1 as shown in FIG. The pixels of the second category C2 are referred to as attention pixels.
In the third classification C3, the energy signal e1 is not output from the pixel-type semiconductor element 1a, or the energy signal e1 having an abnormal level is continuously output. The third category C3 is each pixel value range in which the pixel value is set larger on the positive and negative side than the pixel value range of the second category C2, as shown in FIG. The pixel of the third classification C3 is a pixel that cannot output an output signal corresponding to data at all, and is called a defective pixel.

  The classification device 4 further converts the pixel type semiconductor element 1a classified in the second classification C2 into a plurality of stages in accordance with the performance level of the pixel type semiconductor element 1a within the pixel value range of the second classification C2. Classify pixels. As shown in FIG. 5, the classification device 4 classifies pixels within the pixel value range of the second classification C2 into, for example, three stages d1, d2, and d3. The first stage d1 is adjacent to the pixel value range of the first classification C1 within each pixel value range of the second classification C2. The second stage d2 is set in the middle of each pixel value range of the second classification C2. The third stage d3 is adjacent to each pixel value range of the third category C3 within each pixel value range of the second category C2.

  The classification device 4 includes a memory 5. The classification device 4 stores the result of pixel classification in the first, second, or third classification C1, C2, C3 in the memory 5 in a data table. Note that the memory 5 may store only the results of the classification into the second classification C2 and the third classification C3. FIG. 6 is a schematic diagram of the data table 10. The data table 10 is formed by arranging a plurality of columns 10 a in the vertical and horizontal directions, and each column 10 a corresponds to each pixel 1 a of the semiconductor detector 1. Here, the pixel 1a pixel-classified into the first classification C1 is designated as “0”, the pixel 1a classified into the second classification C2 is designated as “1”, and the pixel 1a is classified into the third classification C3. Is denoted by “2”. Since each column 10a of the data table 10 corresponds to each pixel 1a of the semiconductor detector 1, the coordinates (x, y) of the pixel can be known from the position of the column 10a.

  The data table 10 may be described as “x coordinate, y coordinate, pixel classification”. For example, if “0, 2, code 0”, a pixel having an x coordinate of “0” and a y coordinate of “2” is a normal pixel. In the case of “10, 25, code 1”, a pixel whose x coordinate is “10” and y coordinate is “25” is a caution pixel. In the case of “20, 20, code 2”, a pixel whose x coordinate is “20” and y coordinate is “20” is a defective pixel. In the case of “30, 10, code 1”, a pixel whose x coordinate is “30” and y coordinate is “10” is a caution pixel.

  The data table 10 may describe pixel classifications according to a preset order of x coordinate and y coordinate. For example, “0, 0, 1, 0, 0, 2, 2, 0, 0, 0, 2,..., 0”. In this case, the (x, y) coordinates are “(x1, y1), (x1, y2) (x1, y3)... (Xn, ym)”. Note that the format of the data table 10 may be any format as long as the xy coordinates and pixel classification of each pixel are known.

  The data table 10 is not limited to the first, second, or third classification C1, C2, C3 as a result of the pixel classification, for example, three stages d1 of the result of pixel classification within each pixel value range of the second classification C2. , D2, and d3 are also stored. The classification results of these three stages d1, d2, and d3 are stored in each column 10a of the corresponding coordinates shown in FIG. 6, for example.

  The interpolation processing device 6 interpolates, for example, the values of the respective pixels of the second classification C2 and the third classification C3 that are classified as having a low performance of the pixel type semiconductor element 1a among the pixel classification results of the classification device 4. . For example, as illustrated in FIG. 7, the interpolation processing device 6 includes a pixel value of the target pixel p5 among the pixels p1, p2,..., P9 and a plurality of peripheral pixels p1, which are arranged around the target pixel p5, An average value process is performed on the pixel values of p2, p3, p4, p6, p7, p8, and p9, and this average value is interpolated to the target pixel p5. In this case, the interpolation processing device 6 includes the target pixel p5 in the average value process if the target pixel p5 is the attention pixel classified into the second classification C2. Further, the interpolation processing unit 6 excludes the target pixel p5 from the average value processing if the target pixel p5 is a missing pixel classified into the third classification C3.

  As shown in FIG. 7, the interpolation processing device 6 weights the pixel value of the target pixel p5 and the pixel values of the plurality of peripheral pixels p1, p2, p3, p4, p6, p7, p8, and p9. For example, the interpolation processing device 6 includes the pixel value of the pixel of interest p5 and a plurality of surroundings among the pixel values corresponding to the pixel-type semiconductor elements 1a classified into the three stages d1, d2, and d3 in the second classification C2. The pixel values of the pixels p1, p2, p3, p4, p6, p7, p8, and p9 are weighted, and the average value processing is performed on the weighted pixel values.

  A weighting data memory 7 is connected to the interpolation processing device 6. The weighting data memory 7 stores weighting data. FIGS. 8A, 8B, and 8C are schematic diagrams illustrating examples of the weighting data G1, G2, and G3. These weighting data G1, G2, and G3 are for weighting each pixel classified into three stages d1, d2, and d3 in the second classification C2, and according to each stage d1, d2, and d3. The weight value is different.

  The weighting data G1 shown in FIG. 8A weights the pixel corresponding to the pixel type semiconductor element 1a classified in the first stage d1 in the second classification C2 as a target pixel. In the weighting data G1, the weight for the target pixel is set to be larger than the weights of the surrounding pixels. The first stage d1 is provided in contact with the normal pixels of the first classification C1. Therefore, the accuracy with respect to the target pixel can be treated as being approximate to that of a normal pixel. Thereby, the weight for the pixel of interest is set larger than the weight of the surrounding pixels, and the contribution ratio for the pixel of interest is increased.

The weighting data G2 shown in FIG. 6B weights the pixel corresponding to the pixel type semiconductor element 1a classified in the second stage d2 in the second classification C2 as a target pixel. In this weighting data G2, the weights of the target pixel and the peripheral pixels are set to be the same.
The weighting data G3 shown in FIG. 6C weights the pixel corresponding to the pixel type semiconductor element 1a classified in the third stage d3 in the second classification C2 as a target pixel. In the weighting data G3, the weight for the target pixel is set to be smaller than the weights of the surrounding pixels.

The third stage d3 is provided in contact with the missing pixel of the third classification C3. Accordingly, the accuracy with respect to the target pixel can be treated as being approximate to that of the defective pixel. Thereby, the weight for the pixel of interest is set smaller than the weight of the surrounding pixels, and the contribution rate to the pixel of interest is reduced.

  The interpolation processing device 6 functions as an image generation unit that generates an image S of the subject subjected to the interpolation processing, that is, generates a radiation distribution image.

On the other hand, for example, when a nuclear medicine examination is performed by being provided in a gamma ray imaging apparatus used for medical nuclear medicine examination, the semiconductor detector 1 uses gamma rays g from a medicine administered into a subject such as a human body for each pixel type. Detection is performed by the semiconductor element 1a, and each energy signal e1 is output from the pixel type semiconductor element 1a. The data collection device 8 collects and counts each energy signal e1 output from each pixel type semiconductor element 1a, and obtains each pixel value (pixel data) of the image of the subject from the collected count value.
The uniformity correction device 9 performs, for example, sensitivity uniformity correction on each pixel value of the subject image, and sends the subject image subjected to the uniformity correction to the interpolation processing device 6.

Next, an interpolation process for the semiconductor detector 1 by the apparatus configured as described above will be described.
The surface ray source 2 uniformly emits gamma rays g. The semiconductor detector 1 detects a uniform gamma ray g by each pixel type semiconductor element 1a, and outputs each energy signal e1 from these pixel type semiconductor elements 1a. These energy signals e1 are sent to the pixel energy spectrum analyzer 3 as uniformity measurement data.

  The pixel energy spectrum analyzer 3 collects and counts each energy signal e1 output from each pixel type semiconductor element 1a of the semiconductor detector 1, and obtains each pixel value (pixel data) from the collected count value. .

  The pixel energy spectrum analyzer 3 analyzes each energy signal e1 output from each pixel type semiconductor element 1a of the semiconductor detector 1. The pixel energy spectrum analyzer 3 collects and counts each energy signal e1 output from each pixel type semiconductor element 1a of the semiconductor detector 1, and obtains each pixel value (pixel data) from the collected count value. . FIG. 3 shows the value of each pixel of the semiconductor detector 1, and there are variations among the pixels, for example, the pixel values “100” “0” “50” “140” “1000”, and the like. The pixel energy spectrum analyzer 3 analyzes the shape of each energy signal e1 output from each pixel type semiconductor element 1a and displays it on a display or the like, or displays the peak position of each energy signal e1 and this peak position. The count value and half-value width (FWHM) are analyzed.

  The classification device 4 receives the analysis result of the pixel energy spectrum analysis device 3, for example, the pixel value of each pixel of the semiconductor detector 1 shown in FIG. 3, and classifies these pixels into three pixels, for example, as shown in FIG. The first classification C1 in which the performance of each pixel type semiconductor element 1a is normal, the second classification C2 that can be used by correcting the energy signal e1 output from the pixel type semiconductor element 1a, and the pixel type semiconductor element The pixel classification is made into the third classification C3 in which the energy signal e1 output from 1a cannot be corrected and is abnormal. Of these, the pixels of the first classification C1 are normal pixels. The pixels of the second category C2 are attention pixels. The pixel of the third classification C3 is a defective pixel.

  The classification device 4 further converts the pixel type semiconductor element 1a classified in the second classification C2 into a plurality of stages in accordance with the performance level of the pixel type semiconductor element 1a within the pixel value range of the second classification C2. Classify pixels. As shown in FIG. 5, the classification device 4 classifies pixels within the pixel value range of the second classification C2 into, for example, three stages d1, d2, and d3. The classification device 4 converts the pixel classification results into the first, second, or third classification C1, C2, and C3 into a data table as shown in FIG. 6, for example, and stores it in the memory 5 as well as the second classification C2. For example, three stages d1, d2, and d3 of the result of pixel classification within each pixel value range are also stored in the memory 5.

  On the other hand, for example, when performing a nuclear medicine examination provided in gamma-ray imaging used for medical nuclear medicine examination, the semiconductor detector 1 uses gamma rays g from a medicine administered into a subject such as a human body for each pixel type semiconductor. It detects with the element 1a and outputs each energy signal e1 from these pixel type semiconductor elements 1a. The data collection device 8 collects and counts each energy signal e1 output from each pixel type semiconductor element 1a, and obtains each pixel value (pixel data) of the image of the subject from the collected count value. The uniformity correction device 9 performs, for example, sensitivity uniformity correction on each pixel value of the subject image, and sends the subject image subjected to the uniformity correction to the interpolation processing device 6.

  The interpolation processing device 6 receives the image of the subject whose uniformity has been corrected by the uniformity correction device 9, and is classified as having a low performance of the pixel type semiconductor element 1a in the pixel classification result by the classification device 4 for this image. For example, each pixel value of the second classification C2 and the third classification C3 is interpolated. For example, as illustrated in FIG. 7, the interpolation processing device 6 includes a pixel value of the target pixel p5 among the pixels p1, p2,..., P9 and a plurality of peripheral pixels p1, which are arranged around the target pixel p5, An average value process is performed on the pixel values of p2, p3, p4, p6, p7, p8, and p9, and this average value is interpolated to the target pixel p5. In this case, the interpolation processing device 6 includes the target pixel p5 in the average value process if the target pixel p5 is the attention pixel classified into the second classification C2. Further, the interpolation processing unit 6 excludes the target pixel p5 from the average value processing if the target pixel p5 is a missing pixel classified into the third classification C3.

Here, a specific example of the interpolation process will be described. FIG. 9 shows each pixel p1, p2,..., P9 corresponding to each pixel type semiconductor element 1a in the semiconductor detector 1. In the figure, an interpolation process for the first to fourth regions H1 to H4 will be described. In the regions H1 to H4, each pixel is denoted by reference numerals p1, p2,.
First, in the first region H1, the pixel p5 is a target pixel, and the pixels p1, p2, p3, p4, p6, p7, p8, and p9 are peripheral pixels. When the target pixel p5 is classified as a target pixel, the interpolation processing device 6 performs an average value process including the target pixel p5 when performing the interpolation process on the target pixel p5. That is, the interpolation process for the pixel of interest p5 is
(P1 + p2 + p3 + p4 + p5 + p6 + p7 + p8 + p9) /9.0 (1)
And the calculated value is replaced with the target pixel p5 as the pixel value of the target pixel p5. In the above formula (1), each pixel p1, p2, p3, p4, p5, p6, p7, p8, and p9 represents a pixel value.
When the target pixel p5 is classified as a defective pixel, the interpolation processing device 6 performs an average value process except for the target pixel p5 when performing the interpolation process on the target pixel p5. That is, the interpolation process for the pixel of interest p5 is
(P1 + p2 + p3 + p4 + p6 + p7 + p8 + p9) /8.0 (2)
And the calculated value is replaced with the target pixel p5 as the pixel value of the target pixel p5.

Next, the second region H2 is a case where there is an image boundary in the periphery. In the second region H2, the pixel p5 is a target pixel, and the pixels p4, p7, and p8 are peripheral pixels. When the target pixel p5 is classified as a target pixel, the interpolation processing device 6 performs an average value process including the target pixel p5 when performing the interpolation process on the target pixel p5. That is, the interpolation process for the pixel of interest p5 is
(P4 + p5 + p7 + p8) /4.0 (3)
And the calculated value is replaced with the target pixel p5 as the pixel value of the target pixel p5.
When the target pixel p5 is classified as a defective pixel, the interpolation processing device 6 performs an average value process except for the target pixel p5 when performing the interpolation process on the target pixel p5. That is, the interpolation process for the pixel of interest p5 is
(P4 + p7 + p8) /3.0 (4)
And the calculated value is replaced with the target pixel p5 as the pixel value of the target pixel p5.

Next, the third region H3 is a case where there are missing pixels in the periphery. Here, p6 is a missing pixel. In the third region H3, the pixel p5 is the target pixel, and the pixels p1, p2, p3, p4, p6, p7, p8, and p9 are peripheral pixels. When the target pixel p5 is classified as a target pixel, the interpolation processing device 6 performs an average value process including the target pixel p5 when performing the interpolation process on the target pixel p5. Note that the missing pixel p6 is excluded. That is, the interpolation process for the pixel of interest p5 is
(P1 + p2 + p3 + p4 + p5 + p7 + p8 + p9) /8.0 (5)
And the calculated value is replaced with the target pixel p5 as the pixel value of the target pixel p5.
When the target pixel p5 is classified as a defective pixel, the interpolation processing device 6 performs an average value process except for the target pixel p5 when performing the interpolation process on the target pixel p5. The missing pixel p6 is also removed. That is, the interpolation process for the pixel of interest p5 is
(P1 + p2 + p3 + p4 + p7 + p8 + p9) /7.0 (6)
And the calculated value is replaced with the target pixel p5 as the pixel value of the target pixel p5.

Next, the fourth region H4 is a case where there are image boundaries in the periphery and missing pixels in the periphery. Here, let p1 and p7 be defective pixels. In the fourth region H4, the pixel p5 is the target pixel, and the pixels p1, p2, p4, p5, p7, and p8 are peripheral pixels. When the target pixel p5 is classified as a target pixel, the interpolation processing device 6 performs an average value process including the target pixel p5 when performing the interpolation process on the target pixel p5. Note that the missing pixels p1 and p7 are excluded. That is, the interpolation process for the pixel of interest p5 is
(P2 + p4 + p5 + p8) /4.0 (7)
And the calculated value is replaced with the target pixel p5 as the pixel value of the target pixel p5.
When the target pixel p5 is classified as a defective pixel, the interpolation processing device 6 performs an average value process except for the target pixel p5 when performing the interpolation process on the target pixel p5. The missing pixels p1, p7 are also removed. That is, the interpolation process for the pixel of interest p5 is
(P2 + p4 + p8) /3.0 (8)
And the calculated value is replaced with the target pixel p5 as the pixel value of the target pixel p5.

  Further, as shown in FIG. 7, the interpolation processing device 6 weights the pixel value of the target pixel p5 and the pixel values of a plurality of peripheral pixels p1, p2, p3, p4, p6, p7, p8, and p9. Do. In the specific description of the interpolation process, the weighting is all “1” for the target pixel p5 and the surrounding pixels p1, p2, p3, p4, p6, p7, p8, and p9. That is, it is the same as using the weighting data G2 shown in FIG.

  In contrast, for example, the interpolation processing device 6 determines the pixel value of the target pixel p5 among the pixel values corresponding to the pixel type semiconductor elements 1a classified into the three stages d1, d2, and d3 in the second classification C2. A plurality of peripheral pixels p1, p2, p3, p4, p6, p7, p8, and p9 are weighted, and an average value process is performed on the weighted pixel values. For weighting, the weighting data G1, G2, and G3 stored in the weighting data memory 7 are used.

For example, when the target pixel p5 is classified into the first stage d1 in the attention pixel shown in FIG. 5 in the first region H1, the interpolation processing device 6 performs the interpolation process on the target pixel p5 as shown in FIG. ) To weight the pixel value of the pixel of interest p5 and the pixel values of the surrounding pixels p1, p2, p3, p4, p6, p7, p8, and p9 to include the pixel of interest p5. Average value processing is performed. That is, the interpolation process for the pixel of interest p5 is
(1 ・ p1 + 1 ・ p2 + 1 ・ p3 + 1 ・ p4 + 2 ・ p5 + 1 ・ p6
+ 1 · p7 + 1 · p8 + 1 · p9) /9.0 (9)
And the calculated value is replaced with the target pixel p5 as the pixel value of the target pixel p5.

When the target pixel p5 is classified into the third stage d3 in the attention pixel shown in FIG. 5 in the first region H1, the interpolation processing device 6 performs the interpolation process for the target pixel p5 as shown in FIG. ) To weight the pixel value of the pixel of interest p5 and the pixel values of the surrounding pixels p1, p2, p3, p4, p6, p7, p8, and p9 to include the pixel of interest p5. Average value processing is performed. That is, the interpolation process for the pixel of interest p5 is
(2 ・ p1 + 2 ・ p2 + 2 ・ p3 + 2 ・ p4 + 1 ・ p5 + 2 ・ p6
+ 2 · p7 + 2 · p8 + 2 · p9) /9.0 (10)
And the calculated value is replaced with the target pixel p5 as the pixel value of the target pixel p5.
However, the interpolation processing device 6 generates an image S of the subject subjected to the interpolation processing, that is, generates a radiation distribution image.

  As described above, according to the above embodiment, each pixel value (pixel data) is obtained by analyzing each energy signal e1 output from each pixel type semiconductor element 1a of the semiconductor detector 1 by the pixel energy spectrum analyzer 3. The pixel values are obtained and classified by the classification device 4 into, for example, first to third classifications C1, C2, and C3 according to the performance level of each pixel type semiconductor element 1a. For example, the interpolation processing device 6 performs interpolation processing on each pixel value of the attention pixel of the second classification C2 and the missing pixel of the third classification C3 classified as low.

  As a result, conventionally, only normal pixels are used, and defective pixels whose pixel values are far from the average value are discarded. However, for each of the attention pixel and the defective pixel, each of the attention pixel and the defective pixel Values can be exchanged by interpolation processing. For the attention pixel, interpolation processing is performed including the pixel value of the attention pixel, and for the missing pixel, interpolation processing is performed by excluding the pixel value of the defective pixel. Each pixel value can be replaced with a value close to the actual pixel value. The attention pixel is a pixel that can obtain an image by performing interpolation using a plurality of pixel values around the pixel, although the energy resolution and the intensity variation of the energy signal e1 are large compared to the normal pixel. Therefore, it can be included in the interpolation process. Therefore, the pixel value of the pixel-type semiconductor element 1a with insufficient performance that is a caution pixel or a defective pixel can be used effectively.

  However, the value of each pixel classified as a caution pixel or a defective pixel is obtained by interpolation processing, and can be used, for example, for acquiring an image. As a result, even if the number of defective pixels in the semiconductor detector 1 increases, it can be acquired without degrading the image quality of the image data, and the current semiconductor detector 1 can be used more effectively.

  For example, as shown in FIG. 9, the interpolation process is performed by using the second region H1 when there is no image boundary or missing pixel around the pixel of interest p5, the second region H2 having an image boundary around the target pixel p5, and the surrounding defect. The pixel of interest p5 can be interpolated for the third region H3 with pixels and the fourth region H4 with image boundaries in the periphery and missing pixels in the periphery.

For example, the interpolation processing device 6 classifies the attention pixels in the second classification C2 into three stages d1, d2, and d3, and interpolates the attention pixels in the first stage d1 using the weighted data G1, An interpolation process is performed on the attention pixel in the third stage d3 using the weighting data G3. By using the weighting data G1 for the attention pixel in the first stage d1, it is possible to perform an interpolation process in which the weight for the target pixel is set larger than the weights of the surrounding pixels and the contribution ratio to the target pixel is increased. By using the weighting data G3 for the missing pixel in the third stage d3, it is possible to perform an interpolation process in which the weight for the target pixel is set smaller than the weights of the surrounding pixels and the contribution ratio to the target pixel is reduced. Thereby, interpolation processing can be performed by performing weighting in accordance with the degree of performance of the pixel type semiconductor element 1a within the second classification C2.
The interpolation processing as described above can improve the image quality without changing the pixel type semiconductor element 1a itself or the hardware system in the semiconductor detector 1. Further, the calculation time required for the interpolation processing is not different from the conventional one.

In the above-described embodiment, for example, the semiconductor detector 1 used in the gamma-ray imaging apparatus used for the medical nuclear medicine examination has been described. However, the present invention is not limited to this and can be applied to a radiation detector such as an X-ray.
The interpolation process may use nonlinear interpolation such as linear interpolation and three-dimensional spline interpolation.
Note that the present invention is not limited to the above-described embodiment as it is, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. In addition, various inventions can be formed by appropriately combining a plurality of components disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined.

  1: Semiconductor detector, 1a: Pixel type semiconductor element, 2: Surface source, 3: Pixel energy spectrum analyzer, 4: Classification device, 5: Memory, 6: Interpolation processing device, 7: Weighted data memory, 8 : Data collection device, 9: Uniformity correction device, 10: Data table, 10a: column.

Claims (14)

In a radiation imaging apparatus including a semiconductor detector that forms each pixel by a plurality of semiconductor elements,
An analysis unit for analyzing the shape of each energy signal as energy spectrum data indicating the energy intensity distribution of radiation output from each of the semiconductor elements;
A classification unit for classifying the performance of each semiconductor element into three based on the analysis result of the energy spectrum data by the analysis unit;
A correction unit that interpolates each pixel value of the two classifications classified as low performance of each of the semiconductor elements among the three classification results;
An interpolating device for a semiconductor detector, comprising: The analysis unit analyzes a shape of each energy signal output from each semiconductor element, and analyzes a peak position of each energy signal, a count value of the peak position, and a half value width. 2. The semiconductor detector interpolating device according to claim 1.   The classification unit includes a first classification in which each characteristic of the plurality of semiconductor elements is normal according to at least the performance of each semiconductor element based on the analysis result of the energy spectrum data, and pixels around the pixel 2. The classification according to claim 1, wherein the second classification that enables image acquisition by performing interpolation processing using values and the third classification in which correction of the pixel value is impossible and abnormal are performed. The semiconductor detector interpolating device described. The classification unit classifies each semiconductor element into the first classification when the energy signal is normally output from each semiconductor element,
Compared with the first classification, the energy resolution of each semiconductor element and the intensity of the energy signal vary greatly, and an image can be acquired by performing interpolation processing using a plurality of pixel values around the pixel. And classifying each of the semiconductor elements into the second classification,
Classifying each semiconductor element to which the energy signal is not output from each semiconductor element or the energy signal of which an abnormal level is always output is continuously output into the third classification;
4. The semiconductor detector interpolating device according to claim 3, wherein: The first classification is set to a pixel value range across the average value of the pixel values of a predetermined ratio of all the pixels of the semiconductor detector as an intermediate value,
The second classification is set to each pixel value range whose pixel value is larger on the positive and negative side than the pixel value range of the first classification,
The third classification is set to each pixel value range in which the pixel value is larger on the positive and negative side than the pixel value range of the second classification.
5. The semiconductor detector interpolating device according to claim 4, wherein the interpolating device is a semiconductor detector.   The correction unit is disposed around the pixel value of the pixel of interest and the pixel of interest among the pixel values classified into at least the second classification and the third classification among the classification results of the classification unit. The pixel of interest is interpolated by performing at least an average value process with the respective pixel values of a plurality of neighboring pixels, and the pixel of interest is classified into the third classification during the interpolation process. 4. The semiconductor detector interpolating apparatus according to claim 3, wherein the pixel of interest is excluded from the average value processing.   The said classification | category part classifies the said semiconductor element classified into the said 2nd classification into a several step according to the grade of the performance of each said semiconductor element within the said 2nd classification. 2. The semiconductor detector interpolating device according to 1.   2. The semiconductor detector interpolation according to claim 1, further comprising a memory for storing each of the semiconductor elements classified into at least the second classification and the third classification by the classification unit in a data table. apparatus.   9. The semiconductor detector interpolation device according to claim 8, wherein the data table is formed from coordinates of the semiconductor element and a classification result of the at least the second classification or the third classification.   2. The semiconductor detector interpolation device according to claim 1, wherein the correction unit performs weighting on the pixel value of the target pixel and the pixel values of the plurality of peripheral pixels.   The correction unit is arranged around the pixel value of the pixel of interest and the pixel of interest among the pixel values corresponding to the semiconductor elements classified into the plurality of stages in the second classification. 7. The semiconductor detector according to claim 6, wherein the pixel values of the plurality of surrounding pixels are weighted, and at least an average value process is performed on the weighted pixel values. Interpolator.   12. The semiconductor detector interpolating apparatus according to claim 11, wherein the correction unit varies the weighting value for each of the plurality of stages. In a semiconductor detector interpolation method provided in a radiation imaging apparatus and forming each pixel by a plurality of semiconductor elements,
Analyzing the shape of each energy signal as energy spectrum data indicating the energy intensity distribution of the radiation output from each semiconductor element,
Based on the analysis result of the energy spectrum data, the performance of each semiconductor element is classified into three,
Interpolating each pixel value of the two classifications classified as low performance of each semiconductor element among the three classification results,
An interpolating method for a semiconductor detector. The analysis includes analyzing the shape of each energy signal output from each semiconductor element, and analyzing a peak position of each energy signal, a count value of the peak position, and a half value width. Item 14. A semiconductor detector interpolation method according to Item 13.

Images