Processing Method for Discontinuous Boundary Points in the Crystal Bar Position Mapping Table Generation Algorithm

Abstract

The present invention provides a method for processing discontinuous boundary points in the crystal bar position mapping table generation algorithm, including: determination of discontinuous boundary points and their position determination, initiation of boundary extension, continuous loop of boundary extension, and completion of boundary extension. In the process of the crystal bar and gamma photon interaction position mapping table generation algorithm, the problem of discontinuous boundary points occurring in the boundary position division of each crystal bar is well solved, the boundary position division effect of the crystal bar is improved, and the accuracy of the final result is ensured.

Description

Technical Field

[0001] The present invention relates to a method for processing discontinuous boundary points in an algorithm for generating a crystal bar position mapping table, and more particularly to a method for processing discontinuous boundary points in an algorithm for generating a mapping table of the interaction position between a crystal bar and a gamma photon. Background Art

[0002] Positron emission tomography (PET) is a novel three-dimensional nuclear medicine imaging technique, which belongs to a type of emission tomography technique. It uses positron-emitting radionuclides to label human metabolites such as glucose as imaging agents, and reflects the metabolic changes by the uptake of the imaging agent by lesions, providing biological metabolic information of diseases. It can diagnose and guide the treatment of various diseases earlier, more sensitively and accurately, and has important application values in biology, neuroscience, oncology, pharmacokinetics, clinical diagnosis and treatment evaluation, etc. Its basic principle is to inject a trace amount of positron-emitting radionuclide tracer into the human body. The positron annihilates with the negative electron in the human body to generate a pair of gamma photons emitted back to back. An external detection device is used to detect this pair of photons to detect the distribution of the radioactive drug in the body. Currently, the commonly used detector system is realized by coupling a crystal array composed of crystal bars with a photomultiplier tube. When the gamma photons generated by the annihilation of positrons and electrons enter the crystal bar and react with the crystal, fluorescence will be generated when the crystal atoms de-excite. The fluorescence propagates along the crystal array and is then converted into an electrical signal by the photomultiplier tube. This electrical signal carries the position information of the incident gamma photon. By using a positioning algorithm, the position information of the crystal bar where the incident gamma photon reacts is obtained, so as to determine the position of the coincidence line of the annihilation event. The accuracy of the position information will directly affect the spatial resolution of the detector, and further affect the imaging quality of PET and the accuracy of image quantitative analysis. Therefore, the accurate positioning of the crystal bar where the gamma photon undergoes energy deposition is crucial.

[0003] However, in reality, due to differences in the arrangement of scintillation crystals, the cutting method of the light-splitting light guide, the signal processing ability of the photomultiplier tube, and the front-end electronic signal acquisition circuit, etc., the position information calculated by the positioning algorithm is prone to distortion. Therefore, it is necessary to correct it through a corresponding position mapping algorithm.

[0004] In addition, in the algorithm for generating a mapping table of the interaction position between a crystal bar and a gamma photon involved in the present invention, after obtaining a boundary position distribution map by dividing the boundary positions of each crystal bar, discontinuous boundary points often appear in the boundary position distribution map. Although the probability of such appearance is relatively small, once it appears, it will greatly affect the boundary position division effect of the crystal bar and cause a large error in the final result. Summary of the Invention

[0005] Aiming at the deficiencies of the prior art, the present invention aims to provide a method for processing discontinuous boundary points in the crystal bar position mapping table generation algorithm, including: determining and locating discontinuous boundary points. After dividing the boundary positions of each crystal bar to obtain a boundary position distribution map, it is determined whether there are discontinuous boundary points in the boundary position distribution map according to the number and positions of adjacent boundary points at the positions of adjacent pixels near each pixel point, and when there are discontinuous boundary points, the position of the pixel point that meets the existence condition of the discontinuous boundary point is determined as the position of the discontinuous boundary point; starting boundary extension. According to the trend from the second adjacent boundary point upstream through the first adjacent boundary point upstream to the discontinuous boundary point, corresponding boundary extension is made at the positive direction position of the downstream adjacent boundary point position of the above discontinuous boundary point to obtain the downstream adjacent boundary point; continuously looping the boundary extension. The loop repeats the determination of the discontinuous boundary point to determine whether the downstream adjacent boundary point obtained each time is a new discontinuous boundary point. If so, the loop repeats the above process of starting the boundary extension to make a new corresponding boundary extension for the new discontinuous boundary point; and completing the boundary extension until it is determined that the newly obtained downstream adjacent boundary point is not a new discontinuous boundary point when repeating the above determination of the discontinuous boundary point.

[0006] According to the above method for processing discontinuous boundary points in the crystal bar position mapping table generation algorithm of the present invention, further, the determination of discontinuous boundary points adopts the eight-connected discrimination method, and the positions of adjacent pixels near each pixel point include the positions of eight pixels adjacent to the upper, lower, left, right, upper left, upper right, lower left, and lower right of each pixel point.

[0007] According to the above method for processing discontinuous boundary points in the crystal bar position mapping table generation algorithm of the present invention, further, when there is one adjacent boundary point or two adjacent adjacent boundary points at the positions of adjacent pixels near each pixel point, it is determined that there are discontinuous boundary points in the boundary position distribution map.

[0008] According to the above method for processing discontinuous boundary points in the crystal bar position mapping table generation algorithm of the present invention, further, the positive direction positions of the downstream adjacent boundary point positions of the discontinuous boundary points include: the positions of the upper, lower, left, and right pixels that are not adjacent to the upstream adjacent boundary point position near the discontinuous boundary point.

[0009] According to the above method for processing discontinuous boundary points in the crystal bar position mapping table generation algorithm of the present invention, further, the trend from the second adjacent boundary point upstream through the first adjacent boundary point upstream to the discontinuous boundary point includes a positive direction trend, a clockwise direction trend, and a counterclockwise direction trend.

[0010] According to the method for processing discontinuous boundary points in the algorithm for generating the crystal bar position mapping table described above in the present invention, further, when the trend from the second adjacent boundary point upstream through the first adjacent boundary point upstream to the discontinuous boundary point is a positive direction trend, a corresponding boundary extension is made at the positive direction position of the position of the downstream adjacent boundary point of the discontinuous boundary point, and the obtained downstream adjacent boundary point is located on the positive direction trend line.

[0011] According to the method for processing discontinuous boundary points in the algorithm for generating the crystal bar position mapping table described above in the present invention, further, when the trend from the second adjacent boundary point upstream through the first adjacent boundary point upstream to the discontinuous boundary point is a clockwise direction trend or a counterclockwise direction trend, a corresponding boundary extension is made at the positive direction position of the position of the downstream adjacent boundary point of the discontinuous boundary point, and the trend from the first adjacent boundary point upstream closest to the discontinuous boundary point through the discontinuous boundary point to the obtained downstream adjacent boundary point is opposite to the trend from the second adjacent boundary point upstream through the first adjacent boundary point upstream to the discontinuous boundary point.

[0012] According to the method for processing discontinuous boundary points in the algorithm for generating the crystal bar position mapping table described above in the present invention, further, when determining whether the newly obtained downstream adjacent boundary point is a discontinuous boundary point, when there are two non-adjacent adjacent boundary points or more than two adjacent boundary points at the positions of the pixel points near each pixel point, it is determined that the newly obtained downstream adjacent boundary point is not a new discontinuous boundary point and the boundary extension is completed. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The above description is only an overview of the technical solution of the present invention. In order to be able to understand the technical means of the present invention more clearly, it can be implemented according to the content of the specification. And in order to make the above technical solution and other features and advantages of the present invention more obvious and understandable, the following preferred embodiments are specifically given and described in detail in conjunction with the drawings as follows.

[0014] Figure 1 is a schematic structural diagram of a system for generating an algorithm for the mapping table of the interaction position between a crystal bar and a gamma photon involved in the present invention;

[0015] Figure 2 is a schematic structural diagram of the central position estimation module of the crystal bar of the system for generating an algorithm for the mapping table of the interaction position between a crystal bar and a gamma photon involved in the present invention. In addition, the connection relationship between the median filter processor, the mean filter processor, the central position estimation module of the crystal bar, and the distance calculation module is shown in the figure, and the rest of the system is Figure 1 similar and omitted;

[0016] Figure 3It is a schematic structural diagram of the crystal bar boundary determination module of the system for generating the mapping table of the interaction position between the crystal bar and gamma photons in the present invention. Additionally, the connection relationships among the distance calculation module, the crystal bar boundary determination module, and the region identification module are shown in the figure. The remaining parts of the system are similar to Figure 1 and are omitted;

[0017] Figure 4 It is a flowchart of the algorithm for generating the mapping table of the interaction position between the crystal bar and gamma photons in the present invention;

[0018] Figure 5 It is a two-dimensional position spectrum diagram obtained statistically in the algorithm for generating the mapping table of the interaction position between the crystal bar and gamma photons in the present invention, that is, a scatter plot;

[0019] Figure 6 Figure 1 It is the peak search result of the scatter plot along the X direction in the algorithm for generating the mapping table of the interaction position between the crystal bar and gamma photons in the present invention;

[0020] Figure 7 It is the peak search result of the scatter plot along the Y direction in the algorithm for generating the mapping table of the interaction position between the crystal bar and gamma photons in the present invention;

[0021] Figure 8 It is the central position diagram of the crystal bar estimated in the algorithm for generating the mapping table of the interaction position between the crystal bar and gamma photons in the present invention, that is, a binary position spectrum diagram;

[0022] Figure 9 It is the distance distribution diagram in the algorithm for generating the mapping table of the interaction position between the crystal bar and gamma photons in the present invention, only showing the Figure 8 local distance distribution diagram corresponding to the E1 region shown in

[0023] Figure 10 It is the result display diagram of the boundary division in the algorithm for generating the mapping table of the interaction position between the crystal bar and gamma photons in the present invention;

[0024] Figure 11 It is Figure 10 the crystal efficiency display diagram of each partition after the boundary division in

[0025] Figure 12 It is the position mapping table of the algorithm in the algorithm for generating the mapping table of the interaction position between the crystal bar and gamma photons in the present invention, only showing the Figure 10 partial position mapping table corresponding to the E2 region shown in

[0026] Figure 13 It is a flowchart of the method for processing discontinuous boundary points in the algorithm for generating the crystal bar position mapping table in the present invention.

[0027] Figure 14 This is an embodiment of the boundary extension for the processing method of discontinuous boundary points in the crystal bar position mapping table generation algorithm disclosed by the present invention;

[0028] Figure 15 This is another embodiment of the boundary extension for the processing method of discontinuous boundary points in the crystal bar position mapping table generation algorithm disclosed by the present invention;

[0029] Figure 16 This is another embodiment of the boundary extension for the processing method of discontinuous boundary points in the crystal bar position mapping table generation algorithm disclosed by the present invention;

[0030] Figure 17 This is an embodiment in which discontinuous boundary points appear in the boundary position distribution diagram of the algorithm for generating the mapping table of the interaction position between the crystal bar and gamma photons involved in the present invention;

[0031] Figure 18 It is a schematic diagram of the result of processing the embodiment shown in Figure 17 by using the processing method of discontinuous boundary points in the crystal bar position mapping table generation algorithm disclosed by the present invention. Specific Embodiments

[0032] Next, in combination with the accompanying drawings and specific embodiments, the present invention will be further described:

[0033] As Figure 1 shown, the generation of the position mapping table mentioned in the present invention is executed by a system for generating a mapping table of the interaction position between a crystal bar and gamma photons, which includes an array crystal 01, a data statistical detector 02, a median filtering processor 03, mean filtering processors 04 and 05, a central position estimation module 06 of the crystal bar, a distance calculation module 07, a crystal bar boundary determination module 08, a region identification module 09, a boundary attribution module 10, a region sequence encoder 11, a position mapping table generation module 12, and a position mapping table output module 13. The array crystal 01 has a plurality of crystal bars (not shown in the figure). The data statistical detector 02 is coupled to the array crystal 01. The data statistical detector 02, the median filtering processor 03, the mean filtering processors 04 and 05, the central position estimation module 06 of the crystal bar, the distance calculation module 07, the crystal bar boundary determination module 08, the region identification module 09, the boundary attribution module 10, the region sequence encoder 11, the position mapping table generation module 12, and the position mapping table output module 13 are connected in sequence.

[0034] According to the system of the algorithm for generating the mapping table of the interaction position between the crystal bar and gamma photons described above in the present invention, further, the data statistical detector 02 is a PET detector. In an embodiment of the present invention, a PET detector coupled with an 8*8 BGO array crystal and a Hamamatsu PMT is used for data acquisition, and the time position distribution of the incident photons detected by the statistical detector 02 is statistically analyzed, and the position distribution map is displayed in an image of 64*64 pixels. As Figure 5 shown, the total number of incident photon times is 736124.

[0035] In addition, according to the system of the algorithm for generating the mapping table of the interaction position between the crystal bar and gamma photons described above in the present invention, further, the median filtering processor 03 and the mean filtering processors 04 and 05 are all Gaussian filtering processors. The median filtering processor 03 is a median filtering processor with a 3*3 two-dimensional template. The mean filtering processors 04 and 05 include an X-direction mean filtering processor 04 and a Y-direction mean filtering processor 05. The X-direction mean filtering processor 04 and the Y-direction mean filtering processor 05 are both connected to the median filtering processor 03 and the central position estimation module 06 of the crystal bar. The X-direction mean filtering processor 04 is a mean filtering processor with a 1*3 template, and the Y-direction mean filtering processor 05 is a mean filtering processor with a 3*1 template.

[0036] As Figure 2 shown, according to the system of the algorithm for generating the mapping table of the interaction position between the crystal bar and gamma photons described above in the present invention, further, the central position estimation module 06 of the crystal bar has an "AND operation" processor 14 and a binarization processor 15. The "AND operation" processor 14 and the binarization processor 15 are integrally connected in series to the central position estimation module 06 of the crystal bar.

[0037] According to the system of the algorithm for generating the mapping table of the interaction position between the crystal bar and gamma photons described above in the present invention, further, the distance calculation module 07 is an Euclidean distance calculator.

[0038] As Figure 3 shown, according to the system of the algorithm for generating the mapping table of the interaction position between the crystal bar and gamma photons described above in the present invention, further, the crystal bar boundary determination module 08 includes an X-direction one-dimensional peak searching unit 16, a Y-direction one-dimensional peak searching unit 17 and an "OR operation" processor 18. The X-direction one-dimensional peak searching unit 16, the Y-direction one-dimensional peak searching unit 17 and the "OR operation" processor 18 are integrated into the crystal bar boundary determination module 08. Both ends of the X-direction one-dimensional peak searching unit 16 and the Y-direction one-dimensional peak searching unit 17 are respectively connected to the distance calculation module 07 and the "OR operation" processor 18. The other end of the "OR operation" processor 18 is connected to the region identification module 09, and the encoder of the region identification module 09 is a random encoder.

[0039] As Figure 4As shown in the figure, the algorithm for generating the position mapping table related to the crystal bar and the gamma photon interaction position mapping table mentioned in the present invention includes the following steps:

[0040] Step 1: The data statistical detector 02 performs two-dimensional statistics on the position information of the incident photon events to obtain a two-dimensional position spectrum, that is, a scatter plot. As Figure 5 shown, the brightness of each pixel point in the two-dimensional position spectrum represents the number of events at that position.

[0041] Step 2: The median filter processor 03 performs median filtering on the obtained two-dimensional position spectrum using a 3*3 two-dimensional template.

[0042] Step 3: The X-direction mean filter processor 04 and the Y-direction mean filter processor 05 respectively perform mean filtering on the two-dimensional position spectrum along the X direction and the Y direction using 1*3 and 3*1 templates.

[0043] Among them, in the above technical solution, steps 2 and 3 are to remove the singular points in the two-dimensional position spectrum, reduce the influence of statistical fluctuations on the original data, and avoid the situation of over-division in the subsequent boundary division algorithm. In step 3, mean filtering of 1*3 and 3*1 templates is respectively performed on the two-dimensional position spectrum along the X direction and the Y direction to obtain the two-dimensional position spectrum 2X and the two-dimensional position spectrum 2Y. Then, the average of the same pixel positions of these two position spectra is taken to obtain the processed two-dimensional position spectrum, that is, the value of the pixel (X, Y) in the processed two-dimensional position spectrum = (two-dimensional position spectrum 2X(X, Y) + two-dimensional position spectrum 2Y(X, Y)) / 2.

[0044] Step 4: For the processed two-dimensional position spectrum, the central position estimation module 06 estimates the central position of each crystal bar using the local maximum algorithm for the two-dimensional position spectrum, and the binarization processor 15 performs binarization processing on the data. The pixel value of the brightest point is set to 1, and the pixel values of the rest are set to 0 to obtain a binarized position map, as Figure 8 shown.

[0045] Among them, when the central position estimation module 06 performs one-dimensional peak searching on the processed two-dimensional position spectrum, we use the derivative peak searching method. One-dimensional peak searching is performed on the processed two-dimensional position spectrum along the X direction and the Y direction respectively. The pixel point value at the peak position is set to 1, and the pixel points at non-peak positions are set to 0. See Figure 6 and Figure 7 , Figure 6 is the peak searching result along the X direction; Figure 7This is the peak search result along the Y direction. Then, the peak search results obtained along the X direction and the Y direction are subjected to an "AND operation" by the "AND operation" processor 14 at the same pixel position. If the pixel is a peak point both along the X direction and the Y direction, then the binarization processor 15 sets the value of this pixel to 1, otherwise it is set to 0. The binarization process is completed. Finally, the center position of each crystal bar is estimated, as shown in Figure 8 shown.

[0046] Step Five: The distance calculation module 07 calculates the distance from each pixel to the nearest bright point (pixel with a value of 1), obtaining a distance distribution map, as shown in Figure 9 shown.

[0047] Among them, the Euclidean distance from each pixel to the nearest bright point is calculated to obtain a distance distribution map. The calculated results are shown in Figure 9 shown. Only a part of it is shown. In the distance distribution map, the crystal bar boundary determination module 08 determines the boundary position corresponding to the crystal bar we need according to the local maximum.

[0048] Step Six: The one-dimensional peak search unit 16 in the X direction and the one-dimensional peak search unit 17 in the Y direction respectively perform local maximum search on the distance distribution map along the X direction and the Y direction to determine the center position of the crystal bar and divide the boundaries of each crystal bar.

[0049] Among them, the local maximum algorithm is also used in this step six. The specific steps are as follows:

[0050] The one-dimensional peak search unit 16 in the X direction and the one-dimensional peak search unit 17 in the Y direction perform one-dimensional peak search on the distance distribution map along the X direction and the Y direction respectively using the derivative peak search method;

[0051] The "OR operation" processor 18 performs an "OR operation" on the peak search results in the X direction and the Y direction at the same pixel point. The maximum value corresponds to the boundary position of each crystal bar, obtaining a boundary position distribution map, as shown in Figure 10 shown; at the same time, the crystal efficiency of each partition is calculated, as shown in Figure 11 shown.

[0052] Step Seven: Using the region identification algorithm, each region is corresponded to the crystal bar number one by one to complete the mapping relationship between the scatter plot and the crystal bar.

[0053] The specific steps are as follows:

[0054] The region identification module 09 performs region identification on the boundary position distribution map. At this time, the formed region identification is not encoded in sequence. It only calculates which pixel points belong to the same region. The pixel points belonging to the same region use the same encoding, and the encodings between each region are ensured not to repeat;

[0055] The boundary position obtained in Step 6 above occupies one pixel, and it is necessary to determine which region this pixel belongs to. The boundary attribution module 10 makes a boundary attribution determination according to the characteristics of the one-dimensional peak-seeking algorithm in Step 6 and the following rules: The boundary points in the X direction belong to the region on the right, and the boundary points in the Y direction belong to the region below;

[0056] The region sequence encoder 11 adopts a sequence coding algorithm to make each region correspond to the crystal table coding sequence, and completes the mapping relationship between the scatter plot and the crystal bars.

[0057] And Step 8: The position mapping table generation module 12 adopts a boundary extraction position mapping algorithm to generate a position mapping table. As Figure 12 shown, only a partial position mapping table corresponding to Region B shown in Figure 10 is shown. Among them, it is required that the generated position mapping table should match the subsequent software interface, and the generated position mapping table is output from the system through the position mapping table output module 13.

[0058] As Figure 13 shown, the present invention provides a method for processing discontinuous boundary points in the crystal bar position mapping table generation algorithm, including: determination of discontinuous boundary points and determination of their positions, initiation of boundary extension, continuous loop of boundary extension, and completion of boundary extension.

[0059] Among them, the determination of discontinuous boundary points includes: after obtaining a boundary position distribution map by dividing the boundary positions of each crystal bar, it is determined whether there are discontinuous boundary points in the boundary position distribution map according to the number and positions of adjacent boundary points at the positions of adjacent pixels near each pixel.

[0060] Further, the determination of discontinuous boundary points adopts an eight-connected discrimination method. The positions of adjacent pixels near each pixel include the positions of eight pixels adjacent to each pixel in the directly above, directly below, directly left, directly right, upper left, upper right, lower left, and lower right directions. As Figure 14 - 16 shown, the positions of pixels in the boundary position distribution map are represented by squares, and it is clearly shown that a certain pixel and the positions of eight pixels adjacent to it in the directly above, directly below, directly left, directly right, upper left, upper right, lower left, and lower right directions.

[0061] Further, when there is one adjacent boundary point or two adjacent adjacent boundary points at the positions of adjacent pixels near each pixel, it is determined that there are discontinuous boundary points in the boundary position distribution map, and when there are discontinuous boundary points, the position of the pixel that satisfies the condition for the existence of discontinuous boundary points is determined as the position of the discontinuous boundary point; as Figure 14 - 16 shown, square A represents the position of a discontinuous boundary point, where Figure 14 and 15There is an adjacent boundary point grid B at the position of adjacent pixels near the discontinuous boundary point grid A, and grid C represents the upstream adjacent boundary point of the adjacent boundary point grid B. Figure 16 There are two adjacent adjacent boundary point grids B and C at the positions of adjacent pixels near the discontinuous boundary point grid A. In the following text, with respect to the discontinuous boundary point represented by grid A, Figure 14 - 16 the adjacent boundary point represented by grid B in it is named the upstream first adjacent boundary point, and the adjacent boundary point represented by grid B is named the upstream second adjacent boundary point.

[0062] The start of boundary extension includes: according to the trend from the upstream second adjacent boundary point (represented by grid C) through the upstream first adjacent boundary point (represented by grid B) to the discontinuous boundary point ( Figure 14 - 16 the trend line of this trend is represented by a single-line arrow in it), make corresponding boundary extension at the positive direction position of the downstream adjacent boundary point position of the above-mentioned discontinuous boundary point (represented by grid A), so as to obtain the downstream adjacent boundary point (represented by grid D).

[0063] Furthermore, the positive direction position of the downstream adjacent boundary point position of the discontinuous boundary point (represented by grid A) includes: the positive up, positive down, positive left, and positive right pixel point positions near the discontinuous boundary point (represented by grid A) that are not adjacent to the upstream adjacent boundary point positions ( Figure 14 and 15 the position represented by grid B in it, Figure 16 the positions represented by grids B and C in it), and the above-mentioned "not adjacent to the upstream adjacent boundary point position" specifically means not having a common side with the upstream adjacent boundary point position.

[0064] Furthermore, the trend from the upstream second adjacent boundary point (represented by grid C) through the upstream first adjacent boundary point (represented by grid B) to the discontinuous boundary point (represented by grid A) ( Figure 14 and 16 the single-line arrow in it represents its trend line) includes a positive direction trend, a clockwise direction trend, and a counterclockwise direction trend.

[0065] As described above, Figure 14 and Figure 15 show that there is an adjacent boundary point (represented by grid B) at the position of adjacent pixels near the discontinuous boundary point (represented by grid A).

[0066] Furthermore, when the trend from the upstream second adjacent boundary point (represented by grid C) through the upstream first adjacent boundary point (represented by grid B) to the discontinuous boundary point (represented by grid A) is a positive direction trend, make corresponding boundary extension at the positive direction position of the downstream adjacent boundary point position of the discontinuous boundary point, and the obtained downstream adjacent boundary point is located on the positive direction trend line, as Figure 14As shown in (a)-(d), where the discontinuous boundary points (represented by square A) are respectively located directly to the right, directly to the left, directly below, and directly above the upstream first adjacent boundary point (represented by square B). According to the above description, the downstream adjacent boundary points obtained by boundary extension (represented by square D) are respectively located directly to the right, directly to the left, directly below, and directly above the discontinuous boundary points (represented by square A).

[0067] In addition, as Figure 15 shown in (a)-(h), when the trend from the upstream second adjacent boundary point (represented by square C) via the upstream first adjacent boundary point (represented by square B) to the discontinuous boundary point (represented by square A) is a clockwise trend or a counterclockwise trend, a corresponding boundary extension is made at the positive direction position of the downstream adjacent boundary point position of the discontinuous boundary point (represented by square A). The trend from the upstream first adjacent boundary point (represented by square B) closest to the discontinuous boundary point (represented by square A) via the discontinuous boundary point (represented by square A) to the obtained downstream adjacent boundary point (represented by square D) can be opposite to the trend from the upstream second adjacent boundary point (represented by square C) via the upstream first adjacent boundary point (represented by square B) to the discontinuous boundary point (represented by square A). That is, when the trend from the upstream second adjacent boundary point (represented by square C) via the upstream first adjacent boundary point (represented by square B) to the discontinuous boundary point (represented by square A) is clockwise, Figure 15 as shown in (a), (d), (f), and (g), the trend from the upstream first adjacent boundary point (represented by square B) closest to the discontinuous boundary point (represented by square A) via the discontinuous boundary point (represented by square A) to the obtained downstream adjacent boundary point (represented by square D) is counterclockwise; when the trend from the upstream second adjacent boundary point (represented by square C) via the upstream first adjacent boundary point (represented by square B) to the discontinuous boundary point (represented by square A) is counterclockwise, Figure 15 as shown in (b), (c), (e), and (h), the trend from the upstream first adjacent boundary point (represented by square B) closest to the discontinuous boundary point (represented by square A) via the discontinuous boundary point (represented by square A) to the obtained downstream adjacent boundary point (represented by square D) is clockwise. In this way, compared with the method where the trend from the upstream first adjacent boundary point (represented by square B) closest to the discontinuous boundary point (represented by square A) via the discontinuous boundary point (represented by square A) to the obtained downstream adjacent boundary point (represented by square D) is the same as the trend from the upstream second adjacent boundary point (represented by square C) via the upstream first adjacent boundary point (represented by square B) to the discontinuous boundary point (represented by square A), it can reduce the error caused by boundary extension.

[0068] As described above, Figure 16 it shows that there are two adjacent adjacent boundary points (represented by squares B and C) at the positions of adjacent pixel points near the discontinuous boundary point (represented by square A).

[0069] According to the method for processing discontinuous boundary points in the above-mentioned algorithm for generating the crystal bar position mapping table of the present invention, further, as Figure 16 (a)-(h) show that when the trend from the second adjacent boundary point upstream (represented by square C) through the first adjacent boundary point upstream (represented by square B) to the discontinuous boundary point (represented by square A) is a clockwise trend or a counterclockwise trend, a corresponding boundary extension is made at the positive direction position of the downstream adjacent boundary point position of the discontinuous boundary point (represented by square A). The trend from the first adjacent boundary point upstream (represented by square B) closest to the discontinuous boundary point (represented by square A) through the discontinuous boundary point (represented by square A) to the obtained downstream adjacent boundary point (represented by square D) can be opposite to the trend from the second adjacent boundary point upstream (represented by square C) through the first adjacent boundary point upstream (represented by square B) to the discontinuous boundary point (represented by square A), that is, when the trend from the second adjacent boundary point upstream (represented by square C) through the first adjacent boundary point upstream (represented by square B) to the discontinuous boundary point (represented by square A) is clockwise, Figure 15 (b), (c), (e) and (h) show that the trend from the first adjacent boundary point upstream (represented by square B) closest to the discontinuous boundary point (represented by square A) through the discontinuous boundary point (represented by square A) to the obtained downstream adjacent boundary point (represented by square D) is counterclockwise; when the trend from the second adjacent boundary point upstream (represented by square C) through the first adjacent boundary point upstream (represented by square B) to the discontinuous boundary point (represented by square A) is counterclockwise, Figure 15 (a), (d), (f) and (g) show that the trend from the first adjacent boundary point upstream (represented by square B) closest to the discontinuous boundary point (represented by square A) through the discontinuous boundary point (represented by square A) to the obtained downstream adjacent boundary point (represented by square D) is clockwise.

[0070] The continuous loop of boundary extension, and the above-mentioned determination of discontinuous boundary points is cyclically repeated to determine whether the obtained downstream adjacent boundary point each time is a new discontinuous boundary point. If so, the above-mentioned starting process of boundary extension is cyclically repeated to make a new corresponding boundary extension for the new discontinuous boundary point. Similarly, compared with the method in which the trend from the first adjacent boundary point upstream (represented by square B) closest to the discontinuous boundary point (represented by square A) through the discontinuous boundary point (represented by square A) to the obtained downstream adjacent boundary point (represented by square D) is the same as the trend from the second adjacent boundary point upstream (represented by square C) through the first adjacent boundary point upstream (represented by square B) to the discontinuous boundary point (represented by square A), it can reduce the error caused by boundary extension.

[0071] The completion of boundary extension is until it is determined that the newly obtained downstream adjacent boundary point is not a new discontinuous boundary point when repeating the above-mentioned determination of discontinuous boundary points.

[0072] Further, when determining the discontinuous boundary points for the newly obtained downstream adjacent boundary points, when there are two non-adjacent adjacent boundary points or more than two adjacent boundary points at the positions of the pixel points near each pixel point, it is determined that the newly obtained downstream adjacent boundary points are not new discontinuous boundary points and the completion of the boundary extension.

[0073] Figure 17 This is an embodiment in which discontinuous boundary points appear in the boundary position distribution diagram in the algorithm of the mapping table generation algorithm for the interaction position between the crystal bar and the gamma photon involved in the present invention; Figure 18 It is the result schematic diagram of processing the embodiment shown in Figure 17 by using the method for processing discontinuous boundary points in the crystal bar position mapping table generation algorithm disclosed in the present invention. As shown in Figure 17 and 18 The present invention provides a method for processing discontinuous boundary points in the crystal bar position mapping table generation algorithm. During the process of the mapping table generation algorithm for the interaction position between the crystal bar and the gamma photon, the problem that discontinuous boundary points may appear in the boundary position division of each crystal bar is excellently solved, the boundary position division effect of the crystal bar is improved, and the accuracy of the final result is ensured.

[0074] For those skilled in the art, various corresponding changes and deformations can be made according to the above-described technical solutions and concepts, and all such changes and deformations should fall within the protection scope of the claims of the present invention.

Claims

1. A method for processing discontinuous boundary points in the crystal bar position mapping table generation algorithm, characterized in that: The method includes: determining discontinuous boundary points and their positions. After dividing the boundary positions of each crystal strip to obtain a boundary position distribution map, it is determined whether there are discontinuous boundary points in the boundary position distribution map according to the number and positions of adjacent boundary points at the positions of adjacent pixels near each pixel, and when there are discontinuous boundary points, the position of the pixel that meets the condition for the existence of discontinuous boundary points is determined as the position of the discontinuous boundary point; Starting boundary extension. According to the trend from the second upstream adjacent boundary point, through the first upstream adjacent boundary point, to the discontinuous boundary point, corresponding boundary extension is made at the positive direction position of the downstream adjacent boundary point position of the above-mentioned discontinuous boundary point, so as to obtain the downstream adjacent boundary point; Continuously looping boundary extension. The loop repeats the determination of the discontinuous boundary point to determine whether each obtained downstream adjacent boundary point is a new discontinuous boundary point. If so, the loop repeats the above process of starting boundary extension to make a new corresponding boundary extension for the new discontinuous boundary point; and Completing boundary extension until it is determined that the newly obtained downstream adjacent boundary point is not a new discontinuous boundary point when repeating the determination of the discontinuous boundary point; Among them, the positive direction positions of the downstream adjacent boundary point positions of the discontinuous boundary point include: the pixel positions of directly above, directly below, directly left, and directly right that are not adjacent to the upstream adjacent boundary point position near the discontinuous boundary point; The trend from the second upstream adjacent boundary point, through the first upstream adjacent boundary point, to the discontinuous boundary point includes a positive direction trend, a clockwise direction trend, and a counterclockwise direction trend; When the trend from the second upstream adjacent boundary point, through the first upstream adjacent boundary point, to the discontinuous boundary point is a positive direction trend, corresponding boundary extension is made at the positive direction position of the downstream adjacent boundary point position of the discontinuous boundary point, and the obtained downstream adjacent boundary point is located on the positive direction trend line; When the trend from the second upstream adjacent boundary point, through the first upstream adjacent boundary point, to the discontinuous boundary point is a clockwise direction trend or a counterclockwise direction trend, corresponding boundary extension is made at the positive direction position of the downstream adjacent boundary point position of the discontinuous boundary point, and the trend from the first upstream adjacent boundary point closest to the discontinuous boundary point, through the discontinuous boundary point, to the obtained downstream adjacent boundary point is opposite to the trend from the second upstream adjacent boundary point, through the first upstream adjacent boundary point, to the discontinuous boundary point.

2. The method for processing discontinuous boundary points in the crystal bar position mapping table generation algorithm according to claim 1, characterized in that: The determination of the discontinuous boundary point adopts an eight-connected discrimination method, and the adjacent pixel positions near each pixel include the eight pixel positions of directly above, directly below, directly left, directly right, upper left, upper right, lower left, and lower right adjacent to each pixel; 3. The method for processing discontinuous boundary points in the crystal bar position mapping table generation algorithm according to claim 2, characterized in that: When there is one adjacent boundary point or two adjacent adjacent boundary points at the adjacent pixel positions near each pixel, it is determined that there are discontinuous boundary points in the boundary position distribution map; 4. The method for processing discontinuous boundary points in the crystal bar position mapping table generation algorithm according to claim 1, characterized in that: When determining the discontinuous boundary point for the newly obtained downstream adjacent boundary point, when there are two non-adjacent adjacent boundary points or more than two adjacent boundary points at the pixel positions near each pixel, it is determined that the newly obtained downstream adjacent boundary point is not a new discontinuous boundary point and the boundary extension is completed.

Images