Automatic identification method and device for respiratory gating region of interest and computer medium

By analyzing the three-dimensional coordinates and fluctuations of PET point cloud data, the respiratory points are automatically identified, which solves the problem of high respiratory requirements for patients in the existing technology, and realizes efficient and low-cost automatic identification of breathing gating areas of interest, which is suitable for CT treatment.

CN119091128AActive Publication Date: 2024-12-06KLARITY MEDICAL & EQUIP GZ
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202411572115.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-11-06
Publication Date
2024-12-06
Estimated Expiration
2044-11-06

AI Technical Summary

Technical Problem

The existing respiratory gating technology has high requirements for patients' breathing, and it is necessary to determine the highest and lowest point of breathing amplitude according to doctor's instructions, which affects the patient's comfort and imaging quality.

Method used

By acquiring multiple PET point cloud data collected continuously, analyzing the maximum values ​​and fluctuations of three-dimensional coordinates, and automatically identifying breathing points using projection and morphological operations to reduce the requirements for the patient's breathing ability.

Benefits of technology

It realizes the detection of breathing points without the need for professional respiratory training in patients, reduces calculation costs and improves efficiency, simplifies the algorithm, and is suitable for automatic recognition of respiratory gating areas of interest in CT treatment.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119091128A_ABST
    Figure CN119091128A_ABST
Patent Text Reader

Abstract

The invention discloses an automatic identification method and device for a respiratory gating region of interest, and a computer medium, point cloud converted from a binocular camera is subjected to simple projection and connected domain calculation through continuous frame point cloud, so that respiratory point detection can be realized, the respiratory amplitude of a patient can be suppressed through a visual respiratory amplitude curve subsequently, and the automatic identification method and device for the respiratory gating region of interest can be realized. Therefore, ct treatment is facilitated. The method is low in cost, simple in algorithm principle, low in calculation cost and high in efficiency.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of respiratory gating technology, and more specifically, to a method, a device and a computer medium for automatically identifying a respiratory gating region of interest. Background Art

[0002] When using some medical imaging equipment such as magnetic resonance imaging (MRI) systems and CT (computed tomography) equipment, respiratory artifacts are usually generated during abdominal imaging due to the patient's respiratory movement, which has a significant impact on the clarity of the image, thereby causing a decrease in imaging quality and affecting the diagnosis of lesions. In current medical imaging equipment, in order to solve the problem of abdominal imaging artifacts, respiratory gating technology is usually used to avoid and compensate.

[0003] Existing respiratory gating technology has relatively high requirements on patients. It is necessary to analyze and calculate the respiratory gating signal in the acquired PET or CT data stream. Patients need to follow the doctor's instructions to breathe to determine the highest and lowest points of the breathing amplitude, which places high demands on patients. Summary of the invention

[0004] One of the purposes of the present invention is to provide a method for automatically identifying a respiratory gating region of interest, so as to solve the technical problem that the existing respiratory gating technology has relatively high requirements on the patient's breathing; the second purpose of the present invention is to provide a device for automatically identifying a respiratory gating region of interest; the third purpose of the present invention is to provide a computer medium.

[0005] In order to solve the above technical problems, the technical solution of the present invention is as follows: A first aspect of the present invention provides a method for automatically identifying a respiratory gating region of interest, comprising the following steps: acquiring a plurality of continuously acquired PET point cloud data, wherein the number of the PET point cloud data is not less than a preset number; According to the plurality of PET point cloud data, respectively obtain the maximum values ​​of the three-dimensional coordinates in the PET point cloud data; According to the maximum values ​​of the X-axis and the Y-axis in the maximum values ​​of the three-dimensional coordinates, a plurality of full-empty matrices and an all-0 matrix are created, and one PET point cloud data corresponds to one full-empty matrix; Project each point of each PET point metadata to the corresponding full-space matrix, and the projection value is the value of the point on the Z axis, to obtain multiple first projection images corresponding to the multiple PET point metadata; Projecting the plurality of PET point cloud data onto the all-0 matrix, where the projection value is the number of repetitions of the points, to obtain a second projection image; According to the plurality of first projection images, a third projection image and a fourth projection image are obtained, wherein the value of each point in the third projection image is the maximum value of the corresponding points in the plurality of first projection images, and the value of each point in the fourth projection image is the minimum value of the corresponding points in the plurality of first projection images; Subtracting the fourth projection image from the third projection image to obtain a fifth projection image; Performing zero-setting processing on the second projection image to obtain a first mask image; Obtaining a second mask image according to the mask image and the fifth projection image; Obtaining a center point of the second mask image according to the second mask image; The center point of the second mask image is mapped to the PET point cloud data to obtain the breathing point.

[0006] In the above-mentioned technical means, analysis is performed based on multi-frame point clouds, and the area with the largest fluctuation of the multi-frame point clouds is determined as the breathing point. The breathing point detection can be achieved through simple projection and connected domain calculation of continuous frame point clouds. The patient does not need to breathe according to the doctor's instructions to determine the highest and lowest points of the breathing amplitude, thereby reducing the requirements for the patient's breathing ability.

[0007] Furthermore, before obtaining the maximum values ​​of the three-dimensional coordinates in the PET point cloud data respectively according to the plurality of PET point cloud data, the method further includes: Stack the multiple PET point cloud data to obtain a matrix all_pts; Calculate the modulus of each point in the matrix all_pts to obtain the point with the smallest modulus; Subtract the coordinate value of the point with the smallest modulus from the coordinate value of each point in the matrix all_pts.

[0008] Further, according to the plurality of PET point cloud data, the maximum values ​​of the three-dimensional coordinates in the PET point cloud data are respectively obtained, including: After subtracting the coordinate value of the point with the smallest modulus from the coordinate value of each point in the matrix all_pts, the modulus of each point is calculated to obtain the point with the largest modulus; The three-dimensional coordinates of the point with the largest modulus are taken as the maximum value of the three-dimensional coordinates in the PET point cloud data.

[0009] Furthermore, according to the maximum values ​​of the X-axis and the Y-axis in the maximum values ​​of the three-dimensional coordinates, multiple full-empty matrices and one full-0 matrix are created, including: The size of each of the all-empty matrix and the all-0 matrix is ​​(max_y+1, max_x+1), where max_x and max_y are the maximum values ​​of the X-axis and Y-axis in the maximum values ​​of the three-dimensional coordinates, respectively.

[0010] Furthermore, projecting each point of each PET point metadata to the corresponding full-space matrix and projecting the plurality of PET point cloud data to the full-0 matrix both include the following steps: The coordinate value of each point in each PET point metadata is subtracted from the coordinate value of the point with the smallest modulus.

[0011] Furthermore, the projection value is the number of repetitions of the point, including: When there are points with the same three-dimensional coordinates in different PET point metadata, the number of repetitions of the point is the number of repetitions of the point.

[0012] Further, performing zeroing processing on the second projection image to obtain a first mask image includes: Points in the second projection image whose values ​​are less than a preset first threshold and greater than a preset second threshold are set to zero to obtain a first mask image, wherein the preset first threshold is less than the preset second threshold.

[0013] Further, obtaining a second mask image according to the first mask image and the fifth projection image includes: Acquire an effective breathing amplitude area from the fifth projection image; An intersection area of ​​the first mask image and the effective breathing amplitude area is used as a second mask image.

[0014] Furthermore, in the second mask image, points outside the intersection area of ​​the first mask image and the effective breathing amplitude area are set to 0.

[0015] Furthermore, acquiring the effective breathing amplitude area from the fifth projection image includes: The fifth projection image is sorted from small to large according to the point values, and points corresponding to values ​​greater than a preset percentile are taken as effective breathing amplitude areas.

[0016] Further, obtaining the center point of the second mask image according to the second mask image includes: Performing several morphological operations on the second mask image to obtain a connected domain of the second mask image, and calculating a center point of the connected domain to obtain a center point of the second mask image.

[0017] Furthermore, several morphological operations are performed on the second mask image, including: The second mask image is sequentially subjected to a 3*3 calculation kernel, a closing operation, a 5*5 calculation kernel, a 9*3 calculation kernel, a 3*9 calculation kernel, an opening operation, and a connected domain calculation to obtain a connected domain of the second mask image.

[0018] A second aspect of the present invention provides an automatic identification device for a respiratory gated region of interest, comprising: A point cloud data module, wherein the point cloud data module acquires a plurality of continuously collected PET point cloud data, wherein the number of the PET point cloud data is not less than a preset number of copies; a first calculation module, wherein the first calculation module obtains the maximum value of the three-dimensional coordinates in the PET point cloud data according to the plurality of PET point cloud data; A matrix creation module, wherein the matrix creation module creates a plurality of full-empty matrices and an all-0 matrix according to the maximum values ​​of the X-axis and the Y-axis in the maximum values ​​of the three-dimensional coordinates, and one PET point cloud data corresponds to one full-empty matrix; A first projection module, which projects each point of each PET point metadata to a corresponding full-space matrix, where the projection value is the value of the point on the Z axis, to obtain multiple first projection images corresponding to the multiple PET point metadata; A second projection module, wherein the second projection module projects the plurality of PET point cloud data onto the all-0 matrix, and the projection value is the number of repetitions of the point, to obtain a second projection image; a second calculation module, wherein the second calculation module obtains a third projection image and a fourth projection image according to the plurality of first projection images, wherein the value of each point in the third projection image is the maximum value of the corresponding points in the plurality of first projection images, and the value of each point in the fourth projection image is the minimum value of the corresponding points in the plurality of first projection images; a third calculation module, wherein the third calculation module subtracts the fourth projection image from the third projection image to obtain a fifth projection image; a zeroing module, wherein the zeroing module performs zeroing processing on the second projection image to obtain a first mask image; a fourth calculation module, wherein the fourth calculation module obtains a second mask image according to the mask image and the fifth projection image; a fifth calculation module, wherein the fifth calculation module obtains a center point of the second mask image according to the second mask image; A mapping module maps the center point of the second mask image to the PET point cloud data to obtain a breathing point.

[0019] Furthermore, the first computing module stacks the multiple PET point cloud data to obtain a matrix all_pts, calculates the modulus of each point in the matrix all_pts to obtain a point with the smallest modulus, and subtracts the coordinate value of the point with the smallest modulus from the coordinate value of each point in the matrix all_pts.

[0020] A third aspect of the present invention provides a computer medium storing a computer program, and when the computer program is executed by a processor, the method for automatically identifying a respiratory gated region of interest is performed.

[0021] Compared with the prior art, the technical solution of the present invention has the following beneficial effects: The present invention provides an automatic identification method for a respiratory gating region of interest. The point cloud converted by the binocular camera can be used to detect respiratory points by simple projection and connected domain calculation of continuous frame point clouds, so as to enable the patient to suppress the respiratory amplitude by visualizing the respiratory amplitude curve in the future, so as to facilitate CT treatment. The present invention has low cost, simple algorithm principle, low calculation cost and high efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 A schematic flow chart of a method for automatically identifying a respiratory gating region of interest provided by an embodiment of the present invention; Figure 2 A schematic diagram of a second projection image provided by an embodiment of the present invention; Figure 3 A schematic diagram of a fifth projection image provided by an embodiment of the present invention; Figure 4 A schematic diagram of a first mask image provided by an embodiment of the present invention; Figure 5 A schematic diagram of a second mask image provided by an embodiment of the present invention; Figure 6 A schematic diagram of a process of performing a morphological operation on a second mask image provided by an embodiment of the present invention; Figure 7 A schematic diagram of mapping the center point of the second mask image to PET point cloud data provided by an embodiment of the present invention; Figure 8 A schematic diagram of an automatic identification system for a respiratory-gated region of interest provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0023] The accompanying drawings are only used for illustrative purposes and are not to be construed as limiting the present application; In order to better illustrate the present embodiment, some parts in the drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product; It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0024] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0025] Example This embodiment provides a method for automatically identifying a respiratory gating region of interest. Figure 1 As shown, the following steps are included: obtaining a plurality of continuously collected PET point cloud data, wherein the number of the PET point cloud data is not less than a preset number; According to the plurality of PET point cloud data, respectively obtain the maximum values ​​of the three-dimensional coordinates in the PET point cloud data; According to the maximum values ​​of the X-axis and the Y-axis in the maximum values ​​of the three-dimensional coordinates, a plurality of full-empty matrices and an all-0 matrix are created, and one PET point cloud data corresponds to one full-empty matrix; Project each point of each PET point metadata to the corresponding full-space matrix, and the projection value is the value of the point on the Z axis, to obtain multiple first projection images corresponding to the multiple PET point metadata; Projecting the plurality of PET point cloud data onto the all-0 matrix, where the projection value is the number of repetitions of the points, to obtain a second projection image; According to the plurality of first projection images, a third projection image and a fourth projection image are obtained, wherein the value of each point in the third projection image is the maximum value of the corresponding points in the plurality of first projection images, and the value of each point in the fourth projection image is the minimum value of the corresponding points in the plurality of first projection images; Subtracting the fourth projection image from the third projection image to obtain a fifth projection image; Performing zero-setting processing on the second projection image to obtain a first mask image; Obtaining a second mask image according to the mask image and the fifth projection image; Obtaining a center point of the second mask image according to the second mask image; The center point of the second mask image is mapped to the PET point cloud data to obtain the breathing point.

[0026] In this embodiment, by processing multiple continuously collected PET point cloud data, the area with the largest fluctuation in the multiple continuously collected PET point cloud data is obtained, and the area with the largest fluctuation is used as the breathing point, thereby achieving simple detection of the breathing point without the need for patients to undergo professional breathing training.

[0027] Example 2 This embodiment provides a method for automatically identifying a respiratory gating region of interest. Figure 1 As shown, the following steps are included: obtaining a plurality of continuously collected PET point cloud data, wherein the number of the PET point cloud data is not less than a preset number; According to the plurality of PET point cloud data, respectively obtain the maximum values ​​of the three-dimensional coordinates in the PET point cloud data; According to the maximum values ​​of the X-axis and the Y-axis in the maximum values ​​of the three-dimensional coordinates, a plurality of full-empty matrices and an all-0 matrix are created, and one PET point cloud data corresponds to one full-empty matrix; Project each point of each PET point metadata to the corresponding full-space matrix, and the projection value is the value of the point on the Z axis, to obtain multiple first projection images corresponding to the multiple PET point metadata; Projecting the plurality of PET point cloud data onto the all-0 matrix, where the projection value is the number of repetitions of the points, to obtain a second projection image; According to the plurality of first projection images, a third projection image and a fourth projection image are obtained, wherein the value of each point in the third projection image is the maximum value of the corresponding points in the plurality of first projection images, and the value of each point in the fourth projection image is the minimum value of the corresponding points in the plurality of first projection images; Subtracting the fourth projection image from the third projection image to obtain a fifth projection image; Performing zero-setting processing on the second projection image to obtain a first mask image; Obtaining a second mask image according to the mask image and the fifth projection image; Obtaining a center point of the second mask image according to the second mask image; The center point of the second mask image is mapped to the PET point cloud data to obtain the breathing point.

[0028] In this embodiment, by processing multiple continuously collected PET point cloud data, the area with the largest fluctuation in the multiple continuously collected PET point cloud data is obtained, and the area with the largest fluctuation is used as the breathing point, thereby achieving simple detection of the breathing point without the need for patients to undergo professional breathing training.

[0029] In a further embodiment, the preset number of copies is not less than 10 copies. If one cycle is 10 seconds, the multiple copies of PET point cloud data collected continuously are continuous point clouds in the cycle.

[0030] In this embodiment, the purpose of continuously acquiring multiple copies of PET point cloud data is to analyze the area with the largest fluctuation as the breathing point through continuous multi-frame point clouds.

[0031] In a further embodiment, before respectively obtaining the maximum value of the three-dimensional coordinates in the PET point cloud data according to the plurality of PET point cloud data, the method further includes: Stack the multiple PET point cloud data to obtain a matrix all_pts; Calculate the modulus of each point in the matrix all_pts to obtain the point with the smallest modulus; Subtract the coordinate value of the point with the smallest modulus from the coordinate value of each point in the matrix all_pts.

[0032] In this embodiment, before solving the maximum value of the three-dimensional coordinates in the PET point cloud data, the minimum value of the three-dimensional coordinates in the PET point cloud data is first solved, and then the coordinates of all points in the PET point cloud data are subtracted from the coordinates of the point with the minimum value, so as to translate the PET point cloud data to the center of the coordinate system. In this way, it can be ensured that the size of the full control matrix and the full zero matrix subsequently created according to the maximum value of the three-dimensional coordinates in the PET point cloud data can accommodate all the PET point cloud data, and some data will not be missing due to the offset of the coordinates.

[0033] In a further embodiment, obtaining the maximum values ​​of the three-dimensional coordinates in the PET point cloud data respectively according to the plurality of PET point cloud data includes: After subtracting the coordinate value of the point with the smallest modulus from the coordinate value of each point in the matrix all_pts, the modulus of each point is calculated to obtain the point with the largest modulus; The three-dimensional coordinates of the point with the largest modulus are taken as the maximum value of the three-dimensional coordinates in the PET point cloud data.

[0034] In a specific embodiment, the matrix of one point cloud is (500, 3), so ten point clouds are stacked up as all_pts (5000, 3), the minimum value min_value of the modulus of each point in the all_pts point cloud is calculated, and after all_pts is subtracted from min_value, the maximum value max_value of the modulus of each point in all_pts is calculated, and max_x, max_y, and max_z are obtained based on the three-dimensional coordinates of the point corresponding to max_value, which are used as the size limit of the subsequent projection image.

[0035] In a further embodiment, a plurality of full-empty matrices and a full-0 matrix are created according to the maximum values ​​of the X-axis and the Y-axis among the maximum values ​​of the three-dimensional coordinates, including: The size of each of the all-empty matrix and the all-0 matrix is ​​(max_y+1, max_x+1), where max_x and max_y are the maximum values ​​of the X-axis and Y-axis in the maximum values ​​of the three-dimensional coordinates, respectively.

[0036] In a further embodiment, projecting each point of each PET point metadata to the corresponding full-space matrix and projecting the plurality of PET point cloud data to the full-zero matrix both include the following steps: The coordinate value of each point in each PET point metadata is subtracted from the coordinate value of the point with the smallest modulus.

[0037] In this embodiment, when solving the maximum value of the three-dimensional coordinates in the PET point cloud data, the minimum value of the three-dimensional coordinates in the PET point cloud data is first solved, and then the coordinates of the point with the minimum value are subtracted from the coordinates of all the points in the PET point cloud data, so as to achieve the translation of the PET point cloud data to the center of the coordinate system. When the PET point cloud data is subsequently projected to an all-0 matrix, it is also necessary to subtract the coordinates of the point with the minimum value from the coordinates of all the points in the PET point cloud data, so as to achieve a one-to-one correspondence between the coordinate values ​​and pixel positions of the PET point cloud data and the all-0 matrix.

[0038] In a further embodiment, the projection value is the number of repetitions of a point, including: When there are points with the same three-dimensional coordinates in different PET point metadata, it is the number of repetitions of the point.

[0039] In a specific embodiment, the second projection image is as follows: Figure 2 shown.

[0040] In a further embodiment, the second projection image is subjected to zeroing processing to obtain a first mask image, such as Figure 4 As shown, including: Points in the second projection image whose values ​​are less than a preset first threshold and greater than a preset second threshold are set to zero to obtain a first mask image, wherein the preset first threshold is less than the preset second threshold.

[0041] In a specific embodiment, the preset first threshold and the preset second threshold are calculated based on the percentage of the number of selected PET point metadata. If the number of selected PET point metadata is 10, then in an optional embodiment, the preset first threshold is 5 and the preset second threshold is 10. After the zeroing process of this embodiment, the noise of the second projection image is removed.

[0042] In a further embodiment, a second mask image is obtained according to the first mask image and the fifth projection image, such as Figure 5 As shown, including: Acquire an effective breathing amplitude area from the fifth projection image; An intersection area of ​​the first mask image and the effective breathing amplitude area is used as a second mask image.

[0043] In a further embodiment, in the second mask image, points outside the intersection area of ​​the first mask image and the effective breathing amplitude area are set to 0.

[0044] In a further embodiment, acquiring the effective breathing amplitude area from the fifth projection image includes: The fifth projection image is sorted from small to large according to the point values, and points corresponding to values ​​greater than a preset percentile are taken as effective breathing amplitude areas.

[0045] The fifth projection image is obtained by subtracting the fourth projection image from the third projection image, such as Figure 3 As shown, specifically: range_data = max_data - min_data In the formula, range_data is the fifth projection image, max_data is the third projection image, min_data is the fourth projection image, the third projection image is the maximum z value corresponding to the same point in multiple PET point metadata, and the third projection image is the minimum z value corresponding to the same point in multiple PET point metadata. Therefore, the fifth projection image represents the z value variation of the same point in multiple PET point metadata. The values ​​of the fifth projection image are sorted from small to large, and the points corresponding to the values ​​greater than the preset percentile correspond to the part with the largest z value variation. In this embodiment, it is believed that the part with the largest z value variation corresponds to the effective breathing area with a high probability, and the effective breathing area is further locked by the mask image.

[0046] In a further embodiment, obtaining the center point of the second mask image according to the second mask image includes: Performing several morphological operations on the second mask image to obtain a connected domain of the second mask image, and calculating a center point of the connected domain to obtain a center point of the second mask image.

[0047] In a further embodiment, several morphological operations are performed on the second mask image, such as Figure 6 As shown, including: The second mask image is sequentially subjected to a 3*3 calculation kernel, a closing operation, a 5*5 calculation kernel, a 9*3 calculation kernel, a 3*9 calculation kernel, an opening operation, and a connected domain calculation to obtain a connected domain of the second mask image.

[0048] Finally, the center point of 2D is mapped to the point cloud, such as Figure 7 As shown, the breathing point is obtained.

[0049] Example 3 The embodiment of the present invention also provides an automatic identification device for a respiratory gating region of interest, such as Figure 8 As shown, including: A point cloud data module, wherein the point cloud data module acquires a plurality of continuously collected PET point cloud data, wherein the number of the PET point cloud data is not less than a preset number of copies; a first calculation module, wherein the first calculation module obtains the maximum value of the three-dimensional coordinates in the PET point cloud data according to the plurality of PET point cloud data; A matrix creation module, wherein the matrix creation module creates a plurality of full-empty matrices and an all-0 matrix according to the maximum values ​​of the X-axis and the Y-axis in the maximum values ​​of the three-dimensional coordinates, and one PET point cloud data corresponds to one full-empty matrix; A first projection module, which projects each point of each PET point metadata to a corresponding full-space matrix, where the projection value is the value of the point on the Z axis, to obtain multiple first projection images corresponding to the multiple PET point metadata; A second projection module, wherein the second projection module projects the plurality of PET point cloud data onto the all-0 matrix, and the projection value is the number of repetitions of the point, to obtain a second projection image; a second calculation module, wherein the second calculation module obtains a third projection image and a fourth projection image according to the plurality of first projection images, wherein the value of each point in the third projection image is the maximum value of the corresponding points in the plurality of first projection images, and the value of each point in the fourth projection image is the minimum value of the corresponding points in the plurality of first projection images; a third calculation module, wherein the third calculation module subtracts the fourth projection image from the third projection image to obtain a fifth projection image; a zeroing module, wherein the zeroing module performs zeroing processing on the second projection image to obtain a first mask image; a fourth calculation module, wherein the fourth calculation module obtains a second mask image according to the mask image and the fifth projection image; a fifth calculation module, wherein the fifth calculation module obtains a center point of the second mask image according to the second mask image; A mapping module maps the center point of the second mask image to the PET point cloud data to obtain a breathing point.

[0050] In a further embodiment, the first computing module further stacks the multiple PET point cloud data to obtain a matrix all_pts, the first computing module calculates the modulus of each point in the matrix all_pts to obtain the point with the smallest modulus, and the first computing module subtracts the coordinate value of the point with the smallest modulus from the coordinate value of each point in the matrix all_pts.

[0051] In a further embodiment, the first calculation module obtains the maximum value of the three-dimensional coordinates in the PET point cloud data according to the plurality of PET point cloud data, including: After subtracting the coordinate value of the point with the smallest modulus from the coordinate value of each point in the matrix all_pts, the modulus of each point is calculated to obtain the point with the largest modulus; The three-dimensional coordinates of the point with the largest modulus are taken as the maximum value of the three-dimensional coordinates in the PET point cloud data.

[0052] In a further embodiment, the matrix creation module creates a plurality of full empty matrices and a full zero matrix according to the maximum values ​​of the X-axis and the Y-axis in the maximum values ​​of the three-dimensional coordinates, including: The size of each of the all-empty matrix and the all-0 matrix is ​​(max_y+1, max_x+1), where max_x and max_y are the maximum values ​​of the X-axis and Y-axis in the maximum values ​​of the three-dimensional coordinates, respectively.

[0053] In a further embodiment, the first projection module projects each point of each PET point metadata to the corresponding full-space matrix, and before projecting the plurality of PET point cloud data to the full-0 matrix, both include the following steps: The coordinate value of each point in each PET point metadata is subtracted from the coordinate value of the point with the smallest modulus.

[0054] In a further embodiment, the zeroing module performs zeroing processing on the second projection image to obtain a first mask image, including: Points in the second projection image whose values ​​are less than a preset first threshold and greater than a preset second threshold are set to zero to obtain a first mask image, wherein the preset first threshold is less than the preset second threshold.

[0055] In a further embodiment, the fourth calculation module obtains the second mask image according to the first mask image and the fifth projection image, comprising: Acquire an effective breathing amplitude area from the fifth projection image; An intersection area of ​​the first mask image and the effective breathing amplitude area is used as a second mask image.

[0056] In a further embodiment, the fourth calculation module sets the points outside the intersection area of ​​the first mask image and the effective breathing amplitude area in the second mask image to 0.

[0057] In a further embodiment, the fourth calculation module obtains the effective breathing amplitude area from the fifth projection image, including: The fifth projection image is sorted from small to large according to the point values, and points corresponding to values ​​greater than a preset percentile are taken as effective breathing amplitude areas.

[0058] In a further embodiment, the fourth calculation module obtains the center point of the second mask image according to the second mask image, including: Performing several morphological operations on the second mask image to obtain a connected domain of the second mask image, and calculating a center point of the connected domain to obtain a center point of the second mask image.

[0059] In a further embodiment, the fourth computing module performs several morphological operations on the second mask image, including: The second mask image is sequentially subjected to a 3*3 calculation kernel, a closing operation, a 5*5 calculation kernel, a 9*3 calculation kernel, a 3*9 calculation kernel, an opening operation, and a connected domain calculation to obtain a connected domain of the second mask image.

[0060] Example 4 An embodiment of the present invention further provides a computer medium, wherein the computer medium stores a computer program, and when the computer program is executed by a processor, the method for automatically identifying a respiratory gating region of interest is completed.

[0061] The same or similar reference numerals correspond to the same or similar components; The terms used to describe the positional relationship in the drawings are only used for illustrative purposes and should not be construed as limiting the present application; Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A method for automatically identifying a respiratory gated region of interest, characterized in that: The following steps are involved: Acquire a plurality of continuously collected PET point cloud data, wherein the number of the PET point cloud data is not less than a preset number of copies; According to the plurality of PET point cloud data, respectively obtain the maximum values ​​of the three-dimensional coordinates in the PET point cloud data; According to the maximum values ​​of the X-axis and the Y-axis in the maximum values ​​of the three-dimensional coordinates, a plurality of full-empty matrices and an all-0 matrix are created, and one PET point cloud data corresponds to one full-empty matrix; Project each point of each PET point metadata to the corresponding full-space matrix, and the projection value is the value of the point on the Z axis, to obtain multiple first projection images corresponding to the multiple PET point metadata; Projecting the plurality of PET point cloud data onto the all-0 matrix, where the projection value is the number of repetitions of the points, to obtain a second projection image; According to the plurality of first projection images, a third projection image and a fourth projection image are obtained, wherein the value of each point in the third projection image is the maximum value of the corresponding points in the plurality of first projection images, and the value of each point in the fourth projection image is the minimum value of the corresponding points in the plurality of first projection images; Subtracting the fourth projection image from the third projection image to obtain a fifth projection image; Performing zero-setting processing on the second projection image to obtain a first mask image; Obtaining a second mask image according to the first mask image and the fifth projection image; Obtaining a center point of the second mask image according to the second mask image; The center point of the second mask image is mapped to the PET point cloud data to obtain the breathing point.

2. The method for automatically identifying a respiratory gating region of interest according to claim 1, characterized in that: Before respectively obtaining the maximum values ​​of the three-dimensional coordinates in the PET point cloud data according to the plurality of PET point cloud data, the method further includes: Stack the multiple PET point cloud data to obtain a matrix all_pts; Calculate the modulus of each point in the matrix all_pts to obtain the point with the smallest modulus; Subtract the coordinate value of the point with the smallest modulus from the coordinate value of each point in the matrix all_pts.

3. The automatic identification method of respiratory gating region of interest according to claim 2, characterized in that: According to the plurality of PET point cloud data, the maximum values ​​of the three-dimensional coordinates in the PET point cloud data are respectively obtained, including: After subtracting the coordinate value of the point with the smallest modulus from the coordinate value of each point in the matrix all_pts, the modulus of each point is calculated to obtain the point with the largest modulus; The three-dimensional coordinates of the point with the largest modulus are taken as the maximum value of the three-dimensional coordinates in the PET point cloud data.

4. The automatic identification method of respiratory gated region of interest according to claim 3, characterized in that: According to the maximum values ​​of the X-axis and the Y-axis in the maximum values ​​of the three-dimensional coordinates, multiple full-empty matrices and one full-0 matrix are created, including: The size of each of the all-empty matrix and the all-0 matrix is ​​(max_y+1, max_x+1), where max_x and max_y are the maximum values ​​of the X-axis and Y-axis in the maximum values ​​of the three-dimensional coordinates, respectively.

5. The automatic identification method of respiratory gated region of interest according to claim 4, characterized in that: Projecting each point of each PET point metadata to the corresponding full-space matrix and projecting the multiple PET point cloud data to the full-0 matrix both include the following steps: The coordinate value of each point in each PET point metadata is subtracted from the coordinate value of the point with the smallest modulus.

6. The automatic identification method of respiratory gated region of interest according to claim 4, characterized in that: The projection value is the number of repetitions of the point, including: When there are points with the same three-dimensional coordinates in different PET point metadata, the number of repetitions of the point is the number of repetitions of the point.

7. The automatic identification method of respiratory gated region of interest according to claim 6, characterized in that: Performing zero processing on the second projection image to obtain a first mask image includes: Points in the second projection image whose values ​​are less than a preset first threshold and greater than a preset second threshold are set to zero to obtain a first mask image, wherein the preset first threshold is less than the preset second threshold.

8. The method for automatically identifying a respiratory gated region of interest according to claim 7, characterized in that: Obtaining a second mask image according to the first mask image and the fifth projection image, comprising: Acquire an effective breathing amplitude area from the fifth projection image; An intersection area of ​​the first mask image and the effective breathing amplitude area is used as a second mask image.

9. The method for automatically identifying a respiratory gated region of interest according to claim 8, characterized in that: In the second mask image, points outside the intersection area of ​​the first mask image and the effective breathing amplitude area are set to 0.

10. The automatic identification method of respiratory gated region of interest according to claim 9, characterized in that: The acquiring the effective breathing amplitude area from the fifth projection image comprises: The fifth projection image is sorted from small to large according to the point values, and points corresponding to values ​​greater than a preset percentile are taken as effective breathing amplitude areas.

11. The method for automatically identifying a respiratory gated region of interest according to claim 10, characterized in that: Obtaining a center point of the second mask image according to the second mask image includes: Performing several morphological operations on the second mask image to obtain a connected domain of the second mask image, and calculating a center point of the connected domain to obtain a center point of the second mask image.

12. The automatic identification method of respiratory gated region of interest according to claim 11, characterized in that: Performing several morphological operations on the second mask image, including: The second mask image is sequentially subjected to a 3*3 calculation kernel, a closing operation, a 5*5 calculation kernel, a 9*3 calculation kernel, a 3*9 calculation kernel, an opening operation, and a connected domain calculation to obtain a connected domain of the second mask image.

13. An automatic identification device for a respiratory gated region of interest, characterized in that: include: A point cloud data module, wherein the point cloud data module acquires a plurality of continuously collected PET point cloud data, wherein the number of the PET point cloud data is not less than a preset number of copies; a first calculation module, wherein the first calculation module obtains the maximum value of the three-dimensional coordinates in the PET point cloud data according to the plurality of PET point cloud data; A matrix creation module, wherein the matrix creation module creates a plurality of full-empty matrices and an all-0 matrix according to the maximum values ​​of the X-axis and the Y-axis in the maximum values ​​of the three-dimensional coordinates, and one PET point cloud data corresponds to one full-empty matrix; A first projection module, which projects each point of each PET point metadata to a corresponding full-space matrix, where the projection value is the value of the point on the Z axis, to obtain multiple first projection images corresponding to the multiple PET point metadata; A second projection module, wherein the second projection module projects the plurality of PET point cloud data onto the all-0 matrix, and the projection value is the number of repetitions of the point, to obtain a second projection image; a second calculation module, wherein the second calculation module obtains a third projection image and a fourth projection image according to the plurality of first projection images, wherein the value of each point in the third projection image is the maximum value of the corresponding points in the plurality of first projection images, and the value of each point in the fourth projection image is the minimum value of the corresponding points in the plurality of first projection images; a third calculation module, wherein the third calculation module subtracts the fourth projection image from the third projection image to obtain a fifth projection image; a zeroing module, wherein the zeroing module performs zeroing processing on the second projection image to obtain a first mask image; a fourth calculation module, wherein the fourth calculation module obtains a second mask image according to the mask image and the fifth projection image; a fifth calculation module, wherein the fifth calculation module obtains a center point of the second mask image according to the second mask image; A mapping module maps the center point of the second mask image to the PET point cloud data to obtain a breathing point.

14. The automatic identification device of respiratory gating region of interest according to claim 13, characterized in that: The first calculation module also stacks the multiple PET point cloud data to obtain a matrix all_pts, calculates the modulus of each point in the matrix all_pts to obtain the point with the smallest modulus, and subtracts the coordinate value of the point with the smallest modulus from the coordinate value of each point in the matrix all_pts.

15. A computer medium, characterized in that The computer medium stores a computer program, and when the computer program is executed by a processor, the method for automatically identifying a respiratory gated region of interest according to any one of claims 1 to 12 is completed.

Citation Information

Patent Citations

  • Method and system for positioning region of interest in real time based on PET data

    CN111544023A

  • Diaphragm prediction system and method based on PCA and TSSM models

    CN112053330A

  • Method and system for extracting three-dimensional information of radiotherapy CT (Computed Tomography) image

    CN118552645A

  • Non-contact diagnostic system for respiratory disease prognosis through thermal-depth spatiotemporal behavioral modeling

    WO2024145683A2