Statistical method and device for asymmetric index voxel value of scanned image

Abstract

The present application relates to a statistical method and device for asymmetry index voxel values ​​of scanned images. The method comprises: obtaining a scanned image of a user, and identifying the lesion edge information of the scanned image through an image edge recognition algorithm; generating a lesion contralateral mask image of the scanned image based on the lesion edge information, and generating an asymmetry index image through an asymmetry index algorithm based on the lesion edge information and the lesion contralateral mask image; generating a regional segmentation mask image of the user based on the asymmetry index image, and statistically analyzing the asymmetry index voxel values ​​of the scanned image of the user based on the regional segmentation mask image through a voxel statistical strategy. The use of this method can improve the accuracy of identifying the asymmetry index voxel values ​​of the scanned image.

Description

Technical Field

[0001] The present application relates to the technical field of statistics of asymmetry index voxel values, and in particular to a method and device for statistics of asymmetry index voxel values ​​of scanned images. Background Art

[0002] In medical imaging technology, the quality of PET images is affected by many factors, including equipment performance, operating techniques, patient position, etc. Among them, the accuracy of the asymmetry index image of the PET image directly affects the accuracy of diagnosis, especially the asymmetry index voxel value of the user's scanned image based on the asymmetry index image, which plays a key role in the user's lesion identification. Therefore, how to improve the asymmetry index voxel value of the PET image is the current research focus.

[0003] The traditional way to identify the asymmetry index voxel value is to manually perform image symmetry processing on the image of the lesion area, then manually annotate the asymmetry index image of the lesion area with the naked eye, and finally calculate the asymmetry index voxel value based on the asymmetry index image. However, the deviation of the naked eye annotation is high and the time efficiency is slow, resulting in low accuracy in identifying the asymmetry index voxel value of the scanned image. Summary of the invention

[0004] Based on this, it is necessary to provide a statistical method, device, computer equipment, computer-readable storage medium and computer program product based on the asymmetry index voxel value of PET / MR scan images to address the above technical problems.

[0005] In a first aspect, the present application provides a statistical method for asymmetry index voxel values ​​of a scanned image, comprising:

[0006] Obtaining a scanned image of the user, and identifying the edge information of the lesion in the scanned image through an image edge recognition algorithm;

[0007] Based on the lesion edge information, a mask image on the opposite side of the scanned image is generated, and based on the lesion edge information and the mask image on the opposite side of the lesion, an asymmetry index image is generated by an asymmetry index algorithm;

[0008] Based on the asymmetry index image, a region segmentation mask image of the user is generated, and based on the region segmentation mask image, a voxel value of the asymmetry index of the scanned image of the user is counted by a voxel statistics strategy.

[0009] Optionally, identifying the lesion edge information of the scanned image by an image edge recognition algorithm includes:

[0010] Identifying the image type of the scanned image through an image recognition network, and acquiring the lesion type of the user in response to a lesion upload operation by a staff member;

[0011] Based on the image type, querying each sample image corresponding to the image type in a historical database, and based on the lesion type, screening each sample image corresponding to the scanned image in each sample image;

[0012] Based on each of the sample images, an initial image edge recognition algorithm is trained to obtain an image edge recognition algorithm, and the scanned image is input into the image edge recognition algorithm to obtain lesion edge information of the scanned image.

[0013] Optionally, generating a mask image on the opposite side of the scanned image based on the lesion edge information includes:

[0014] Based on the lesion edge information, identifying the lesion range of the lesion area of ​​the scanned image in the scanned image, and equally dividing the scanned image to obtain symmetrical scanned images;

[0015] Based on the lesion range of the lesion area in the scanned image, identifying the target symmetrical scanned image to which the lesion area belongs, and based on the target symmetrical scanned image, generating a mirror image of the target symmetrical scanned image through mirror mapping processing;

[0016] The target symmetrical scanned image and the mirror image are subjected to image stitching processing to obtain a mask image on the opposite side of the scanned image.

[0017] Optionally, generating an asymmetry index image by using an asymmetry index algorithm based on the lesion edge information and the mask image on the opposite side of the lesion includes:

[0018] Identifying each scanned voxel value of the target symmetrical scanned image and each mirror voxel value in the mirror image, and calculating the asymmetry index value of each voxel of the target symmetrical scanned image by an asymmetry index algorithm based on each scanned voxel value and each mirror voxel value;

[0019] In each voxel, each lesion voxel within the lesion range is selected, and the asymmetry index value of each lesion voxel is reset to zero to obtain a target asymmetry index value of each lesion voxel;

[0020] Based on the asymmetry index value of each non-lesion voxel of the target symmetric scan image and the target scan voxel value of each lesion voxel, the target symmetric scan image is converted into a medical format image to obtain an asymmetry index image.

[0021] Optionally, generating a region segmentation mask image of the user based on the asymmetry index image includes:

[0022] The asymmetric index image is subjected to image partition processing by a preset image partitioning strategy to obtain an asymmetric index image of each partition;

[0023] The volume value of each partition asymmetry index image is identified, and the volume value of each partition asymmetry index image is marked in each partition asymmetry index image to obtain each region segmentation mask image of the user.

[0024] Optionally, the counting of asymmetry index voxel values ​​of the scanned image of the user based on the region segmentation mask image by a voxel statistics strategy includes:

[0025] For each regional segmentation mask image, based on the asymmetry index value of each voxel of the regional segmentation mask image and the volume value of the regional segmentation mask image, a sub-asymmetry index voxel value of each voxel statistical type in the regional segmentation mask image is counted by a voxel statistical strategy;

[0026] The sub-asymmetry index voxel value of each voxel statistical type in all region segmentation mask images is used as the asymmetry index voxel value of the scanned image of the user.

[0027] In a second aspect, the present application further provides a statistical device for asymmetric index voxel values ​​of a scanned image, comprising:

[0028] An acquisition module is used to acquire a scanned image of a user and identify the edge information of a lesion in the scanned image through an image edge recognition algorithm;

[0029] A generating module, configured to generate a mask image on the opposite side of the scanned image based on the lesion edge information, and to generate an asymmetry index image based on the lesion edge information and the mask image on the opposite side of the lesion by an asymmetry index algorithm;

[0030] A statistical module is used to generate a regional segmentation mask image of the user based on the asymmetry index image, and to count the asymmetry index voxel values ​​of the scanned image of the user through a voxel statistical strategy based on the regional segmentation mask image.

[0031] Optionally, the acquisition module is specifically used to:

[0032] Identifying the image type of the scanned image through an image recognition network, and acquiring the lesion type of the user in response to a lesion upload operation by a staff member;

[0033] Based on the image type, querying each sample image corresponding to the image type in a historical database, and based on the lesion type, screening each sample image corresponding to the scanned image in each sample image;

[0034] Based on each of the sample images, an initial image edge recognition algorithm is trained to obtain an image edge recognition algorithm, and the scanned image is input into the image edge recognition algorithm to obtain lesion edge information of the scanned image.

[0035] Optionally, the generating module is specifically used for:

[0036] Based on the lesion edge information, identifying the lesion range of the lesion area of ​​the scanned image in the scanned image, and equally dividing the scanned image to obtain symmetrical scanned images;

[0037] Based on the lesion range of the lesion area in the scanned image, identifying the target symmetrical scanned image to which the lesion area belongs, and based on the target symmetrical scanned image, generating a mirror image of the target symmetrical scanned image through mirror mapping processing;

[0038] The target symmetrical scanned image and the mirror image are subjected to image stitching processing to obtain a mask image on the opposite side of the scanned image.

[0039] Optionally, the generating module is specifically used for:

[0040] Identifying each scanned voxel value of the target symmetrical scanned image and each mirror voxel value in the mirror image, and calculating the asymmetry index value of each voxel of the target symmetrical scanned image by an asymmetry index algorithm based on each scanned voxel value and each mirror voxel value;

[0041] In each voxel, each lesion voxel within the lesion range is selected, and the asymmetry index value of each lesion voxel is reset to zero to obtain a target asymmetry index value of each lesion voxel;

[0042] Based on the asymmetry index value of each non-lesion voxel of the target symmetric scan image and the target scan voxel value of each lesion voxel, the target symmetric scan image is converted into a medical format image to obtain an asymmetry index image.

[0043] Optionally, the statistics module is specifically used for:

[0044] The asymmetric index image is subjected to image partition processing by a preset image partitioning strategy to obtain an asymmetric index image of each partition;

[0045] The volume value of each partition asymmetry index image is identified, and the volume value of each partition asymmetry index image is marked in each partition asymmetry index image to obtain each region segmentation mask image of the user.

[0046] Optionally, the statistics module is specifically used for:

[0047] For each regional segmentation mask image, based on the asymmetry index value of each voxel of the regional segmentation mask image and the volume value of the regional segmentation mask image, a sub-asymmetry index voxel value of each voxel statistical type in the regional segmentation mask image is counted by a voxel statistical strategy;

[0048] The sub-asymmetry index voxel value of each voxel statistical type in all region segmentation mask images is used as the asymmetry index voxel value of the scanned image of the user.

[0049] In a third aspect, the present application provides a computer device, wherein the computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of any one of the methods in the first aspect are implemented.

[0050] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the steps of any one of the methods in the first aspect are implemented.

[0051] In a fifth aspect, the present application provides a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, the steps of any one of the methods in the first aspect are implemented.

[0052] The statistical method and device for the asymmetry index voxel value of the above-mentioned scanned image obtains the scanned image of the user, and identifies the lesion edge information of the scanned image through an image edge recognition algorithm; based on the lesion edge information, generates the lesion contralateral mask image of the scanned image, and based on the lesion edge information and the lesion contralateral mask image, generates an asymmetry index image through an asymmetry index algorithm; based on the asymmetry index image, generates the user's regional segmentation mask image, and based on the regional segmentation mask image, counts the asymmetry index voxel value of the user's scanned image through a voxel statistical strategy. This scheme performs lesion edge recognition on the scanned image, so that the lesion range (i.e., the range surrounded by the lesion edge information) in the scanned image can be located, and then the terminal generates the lesion contralateral mask image of the scanned image based on the identified lesion range, so as to perform asymmetry index analysis on the image area where the lesion exists. Then, this scheme generates an asymmetry index image through the lesion contralateral mask image and the asymmetry index algorithm, so that the asymmetry index information of the area not containing the lesion can be accurately analyzed. Then, this solution generates segmentation mask images for each region by means of region splitting, and performs voxel statistics on each region segmentation mask image to obtain the asymmetry index voxel value of the user's scanned image, thereby improving the comprehensiveness and granularity of the analysis of the asymmetry index voxel value, avoiding the problems of high deviation and slow timeliness of manual analysis, thereby comprehensively improving the accuracy of identifying the asymmetry index voxel value of the scanned image. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related technologies, the drawings required for use in the embodiments or the related technical descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0054] Figure 1 It is a flowchart of a statistical method for asymmetry index voxel value of a scanned image in one embodiment;

[0055] Figure 2 A schematic diagram of a process of obtaining a mirror image based on mirror mapping in one embodiment;

[0056] Figure 3 A schematic diagram of a voxel image of an axially symmetric target symmetric scan image in one embodiment;

[0057] Figure 4 is a schematic diagram of an asymmetric index graph in one embodiment;

[0058] Figure 5is a schematic diagram of a region segmentation mask image in one embodiment;

[0059] Figure 6 A flowchart of a statistical example of asymmetry index voxel values ​​of a scanned image in one embodiment;

[0060] Figure 7 It is a structural block diagram of a statistical device for calculating asymmetry index voxel values ​​of a scanned image in one embodiment;

[0061] Figure 8 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0063] The statistical method of the asymmetry index voxel value of the scanned image provided in the embodiment of the present application can be applied to the statistical application environment of the asymmetry index voxel value of the scanned image of positron emission tomography (PET) or magnetic resonance (MR). The method can be applied to a terminal, a server, or a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. Among them, the terminal can be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, etc. Among them, the terminal can locate the lesion range (i.e., the range surrounded by the lesion edge information) in the scanned image by performing lesion edge recognition on the scanned image, and then the terminal generates a lesion contralateral mask image of the scanned image based on the identified lesion range, so as to perform asymmetry index analysis on the image area where the lesion exists. Then, this scheme generates an asymmetry index image through the asymmetry index algorithm through the lesion contralateral mask image, so that the asymmetry index information of the area not containing the lesion can be accurately analyzed. Then, this solution generates each regional segmentation mask image through regional segmentation, and performs voxel statistics on each regional segmentation mask image to obtain the asymmetry index voxel value of the user's scanned image, thereby improving the comprehensiveness and granularity of the analysis of the asymmetry index voxel value, avoiding the problems of high deviation and slow timeliness of manual analysis, thereby comprehensively improving the accuracy of identifying the asymmetry index voxel value of the scanned image.

[0064] In an exemplary embodiment, Figure 1As shown, a statistical method for the asymmetry index voxel value of a scanned image is provided, and the method is applied to a terminal as an example for explanation, including the following steps S101 to S103. Among them:

[0065] Step S101, obtaining a scanned image of a user, and identifying the edge information of the lesion in the scanned image through an image edge recognition algorithm.

[0066] In this embodiment, the terminal responds to the image upload operation of the scanning device to obtain the scanned image of the user, and the scanned image may be, but is not limited to, a brain PET image scanned by a PET / MR device, and the PET imaging agent drugs used include, but are not limited to, 18F-FDG, etc.; or the MR device scan includes, but is not limited to, T1 images, T2 images, etc. Then, the terminal identifies the lesion edge information of the scanned image through a pre-trained image edge recognition algorithm. Among them, the lesion edge information is the regional edge information of the area where the lesion is located, for example, the region of interest (ROI) of ischemic cerebral infarction lesions. Among them, the shape of the area surrounded by the regional edge information may be, but is not limited to, a regional shape outlined in a box, a circle, an ellipse, an irregular polygon, etc. Among them, the regional shape is a three-dimensional regional shape. The specific recognition process will be described in detail later. The scanned image of the user may be, but is not limited to, a brain magnetic resonance image, a head magnetic resonance image, etc. The scanned image is a three-dimensional image.

[0067] Step S102, based on the lesion edge information, a mask image on the opposite side of the scanned image is generated, and based on the lesion edge information and the mask image on the opposite side of the lesion, an asymmetry index image is generated by an asymmetry index algorithm.

[0068] In this embodiment, the terminal generates a mask image of the contralateral side of the lesion of the scanned image based on the lesion edge information, and generates an asymmetric index image through an asymmetric index algorithm based on the lesion edge information and the mask image of the contralateral side of the lesion. Wherein, the mask image of the contralateral side of the lesion is a half-scanned image including the area where the lesion is located, and a contralateral mirror image of the half-scanned image, and the half-image is a left and right half-scanned image divided based on the central axis of the head as the midline. The specific generation process will be described in detail later, and the asymmetric index algorithm is an algorithm for calculating the asymmetric index value for each voxel in the scanned image. Wherein, a voxel is a unit volume image corresponding to the smallest volume unit that can be scanned in a scanned image of medical imaging. The specific calculation process will be described in detail later.

[0069] Step S103: generating a region segmentation mask image of the user based on the asymmetry index image, and counting the asymmetry index voxel values ​​of the user's scanned image based on the region segmentation mask image through a voxel statistics strategy.

[0070] In this embodiment, the terminal generates a user's regional segmentation mask image based on the asymmetry index image, and based on the regional segmentation mask image, counts the asymmetry index voxel value of the user's scanned image through a voxel statistical strategy. The regional segmentation mask image is an image that is masked after the user's asymmetry index image is segmented, and the regional segmentation method is the segmentation method preset in the terminal by the staff. After segmentation, the shape, size, and volume range of each region can be the same or different, and the shape, size, and volume range of each region can change with the adjustment of the staff.

[0071] Based on the above scheme, by identifying the edge of the lesion on the scanned image, the lesion range in the scanned image (i.e., the range surrounded by the lesion edge information) can be located. Then, based on the identified lesion range, the terminal generates a lesion contralateral mask image of the scanned image to perform asymmetry index analysis on the image area where the lesion exists. Then, this scheme generates an asymmetry index image through the asymmetry index algorithm using the lesion contralateral mask image, so that the asymmetry index information of the area not containing the lesion can be accurately analyzed. After that, this scheme generates each regional segmentation mask image by regional splitting, and performs voxel statistics on each regional segmentation mask image to obtain the asymmetry index voxel value of the user's scanned image, thereby improving the comprehensiveness and granularity of the analysis of the asymmetry index voxel value, avoiding the problem of high deviation and slow timeliness of manual analysis, thereby comprehensively improving the accuracy of identifying the asymmetry index voxel value of the scanned image.

[0072] Optionally, the lesion edge information of the scanned image is identified through an image edge recognition algorithm, including: identifying the image type of the scanned image through an image recognition network, and obtaining the user's lesion type in response to the staff's lesion upload operation; based on the image type, querying each sample image corresponding to the image type in a historical database, and based on the lesion type, screening each sample image corresponding to the scanned image in each sample image; based on each sample image, training an initial image edge recognition algorithm to obtain an image edge recognition algorithm, and inputting the scanned image into the image edge recognition algorithm to obtain the lesion edge information of the scanned image.

[0073] In this embodiment, the terminal identifies the image type of the scanned image through the image recognition network, and obtains the user's lesion type in response to the staff's lesion upload operation. The image type includes but is not limited to a 3D head scan image, a 3D brain scan image, etc. The lesion type can be but is not limited to a thrombus type, a cerebral infarction type, etc.

[0074] Then, the terminal searches for each sample image corresponding to the image type in the historical database based on the image type, and selects each sample image corresponding to the scanned image based on the lesion type. The historical database includes the image type and lesion type corresponding to each sample image, and each sample image corresponds to one image type and one lesion type.

[0075] Then, the terminal trains the initial image edge recognition algorithm based on each sample image to obtain the image edge recognition algorithm, and inputs the scanned image into the image edge recognition algorithm to obtain the lesion edge information of the scanned image. The image edge recognition algorithm can be, but is not limited to, an automatic segmentation algorithm based on a deep neural network, such as a Canny edge detection algorithm.

[0076] Based on the above scheme, the lesion edge information of the scanned image is identified through the image edge recognition algorithm, thereby improving the accuracy of the identified lesion edge information.

[0077] Optionally, based on the lesion edge information, a mask image on the opposite side of the scanned image is generated, including: based on the lesion edge information, identifying the lesion range of the lesion area in the scanned image, and equally dividing the scanned image to obtain symmetrical scanned images; based on the lesion range of the lesion area in the scanned image, identifying the target symmetrical scanned image to which the lesion area belongs, and based on the target symmetrical scanned image, generating a mirror image of the target symmetrical scanned image through mirror mapping processing; stitching the target symmetrical scanned image and the mirror image to obtain a mask image on the opposite side of the scanned image.

[0078] In this embodiment, the terminal identifies the lesion area of ​​the scanned image in the lesion range in the scanned image based on the lesion edge information, and performs equal segmentation processing on the scanned image to obtain symmetrical scanned images on the left and right sides, wherein the equal segmentation processing method is to perform equal segmentation on the scanned image based on the central axis of the head to obtain symmetrical scanned images on the left and right sides.

[0079] Then, the terminal identifies the target symmetrical scan image to which the lesion area belongs based on the lesion range of the lesion area in the scan image, and generates a mirror image of the target symmetrical scan image through mirror mapping based on the target symmetrical scan image. Figure 2 As shown in FIG. 1 , it is a mirror image obtained based on mirror mapping, wherein the left image is a target symmetrical scan image, and the right image is a mirror image.

[0080] Finally, the terminal performs image stitching processing on the target symmetrical scan image and the mirror image to obtain a mask image on the opposite side of the lesion of the scan image.

[0081] Based on the above scheme, by identifying the symmetrical scan image where the lesion is located, mirror mapping is performed, thereby limiting the range of calculating the asymmetry index value, reducing the amount of calculation, and improving the calculation accuracy.

[0082] Optionally, based on the lesion edge information and the mask image on the opposite side of the lesion, an asymmetry index image is generated by an asymmetry index algorithm, including: identifying each scanning voxel value of the target symmetrical scanning image and each mirror voxel value in the mirror image, and based on each scanning voxel value and each mirror voxel value, calculating the asymmetry index value of each voxel of the target symmetrical scanning image by an asymmetry index algorithm; in each voxel, screening each lesion voxel within the lesion range, and returning the asymmetry index value of each lesion voxel to zero to obtain a target asymmetry index value of each lesion voxel; based on the asymmetry index value of each non-lesion voxel of the target symmetrical scanning image and the target scanning voxel value of each lesion voxel, converting the target symmetrical scanning image into a medical format image to obtain an asymmetry index image.

[0083] In this embodiment, Figure 3 As shown, the terminal identifies each scanned voxel value of the target symmetrical scanned image and each mirror voxel value in the mirror image, and calculates the asymmetric index value of each voxel of the target symmetrical scanned image through an asymmetric index algorithm based on each scanned voxel value and each mirror voxel value. The calculation formula of the asymmetric index algorithm (Asymmetric Index, AI) is:

[0084] Formula 1:

[0085] Formula 2:

[0086] Then, the terminal selects each lesion voxel in the lesion range from each voxel, and returns the asymmetry index value of each lesion voxel to zero, thereby obtaining the target asymmetry index value of each lesion voxel. Figure 4 As shown, the area without gray mark is the lesion area, and each voxel within the lesion range corresponding to the lesion area is each lesion voxel.

[0087] Finally, based on the asymmetry index values ​​of each non-lesion voxel in the target symmetric scan image and the target scan voxel value of each lesion voxel, the terminal converts the target symmetric scan image into a medical format image to obtain an asymmetry index image. Specifically, the terminal saves the image corresponding to the calculated affected side (i.e., the target symmetric scan image) as a 3D medical image format NIFITY (Neuroimaging Informatics Technology Initiative) image or NRRD (Nearly Raw Raster Data) image to obtain an asymmetry index image.

[0088] Based on the above scheme, the asymmetry index algorithm designed in this scheme is used to calculate the asymmetry index value of each voxel of the target asymmetry index image, thereby improving the accuracy of the calculation. Then, by removing the asymmetry index value of the lesion area, the accuracy of identifying the lesion area is ensured.

[0089] Optionally, based on the asymmetry index image, a regional segmentation mask image of the user is generated, including: performing image partitioning processing on the asymmetry index image through a preset image partitioning strategy to obtain asymmetry index images of each partition; identifying the volume value of each partition asymmetry index image, and marking the volume value of each partition asymmetry index image in each partition asymmetry index image to obtain regional segmentation mask images of the user.

[0090] In this embodiment, the terminal uses a preset image partitioning strategy to perform image partitioning on the asymmetric index image to obtain an asymmetric index image for each partition. Figure 5 As shown, each partition asymmetry index image after image partition processing, wherein each partition asymmetry index image may include part of the lesion range, or all of the lesion range.

[0091] Then, the terminal identifies the volume value of each partition asymmetric index image, and marks the volume value of each partition asymmetric index image in each partition asymmetric index image, thereby obtaining the user's region segmentation mask image.

[0092] Based on the above scheme, by marking the volume of each partition, it is convenient to count the corresponding relationship between the voxel value information of the asymmetry index image of each partition and the range of each partition, thereby improving the accuracy of statistics.

[0093] Optionally, based on the regional segmentation mask image, asymmetry index voxel values ​​of the user's scanned image are counted through a voxel statistics strategy, including: for each regional segmentation mask image, based on the asymmetry index value of each voxel of the regional segmentation mask image and the volume value of the regional segmentation mask image, sub-asymmetry index voxel values ​​of each voxel statistical type in the regional segmentation mask image are counted through a voxel statistics strategy; and the sub-asymmetry index voxel values ​​of each voxel statistical type in all regional segmentation mask images are used as the asymmetry index voxel values ​​of the user's scanned image.

[0094] In this embodiment, for each regional segmentation mask image, the terminal counts the sub-asymmetry index voxel value of each voxel statistical type in the regional segmentation mask image based on the asymmetry index value of each voxel in the regional segmentation mask image and the volume value of the regional segmentation mask image through a voxel statistical strategy. The voxel statistical type includes the total number type of voxels in the regional segmentation mask image, the maximum asymmetry index voxel value type in the segmentation mask image, the minimum asymmetry index voxel value type in the segmentation mask image, and the average asymmetry index voxel value type in the segmentation mask image.

[0095] Finally, the terminal uses the sub-asymmetry index voxel value of each voxel statistical type in all region segmentation mask images as the asymmetry index voxel value of the user's scanned image.

[0096] Based on the above scheme, the asymmetry index voxel value of the user's scanned image is obtained through partition statistics, which improves the comprehensiveness and granularity of the analysis of the asymmetry index voxel value, avoids the problems of high deviation and slow timeliness of manual analysis, and thus comprehensively improves the accuracy of identifying the asymmetry index voxel value of the scanned image.

[0097] The present application also provides a statistical example of the asymmetry index voxel value of a scanned image, such as Figure 6 As shown, the specific processing process includes the following steps:

[0098] Step S601, obtaining a scanned image of the user.

[0099] Step S602: Identify the image type of the scanned image through an image recognition network, and obtain the user's lesion type in response to the staff's lesion upload operation.

[0100] Step S603 , based on the image type, query each sample image corresponding to the image type in the history database, and based on the lesion type, select each sample image corresponding to the scanned image from each sample image.

[0101] Step S604: Based on each sample image, an initial image edge recognition algorithm is trained to obtain an image edge recognition algorithm, and the scanned image is input into the image edge recognition algorithm to obtain lesion edge information of the scanned image.

[0102] Step S605 , based on the lesion edge information, identifying the lesion range of the lesion area in the scanned image, and performing equal segmentation processing on the scanned image to obtain symmetrical scanned images.

[0103] Step S606, based on the lesion range of the lesion area in the scanned image, identifying the target symmetrical scanned image to which the lesion area belongs, and generating a mirror image of the target symmetrical scanned image through mirror mapping processing based on the target symmetrical scanned image.

[0104] Step S607, performing image stitching processing on the target symmetrical scanned image and the mirror image to obtain a mask image on the opposite side of the scanned image.

[0105] Step S608, identifying each scanning voxel value of the target symmetrical scanning image and each mirror voxel value in the mirror image, and based on each scanning voxel value and each mirror voxel value, calculating the asymmetry index value of each voxel of the target symmetrical scanning image through an asymmetry index algorithm.

[0106] Step S609 , among the voxels, each lesion voxel within the lesion range is selected, and the asymmetry index value of each lesion voxel is reset to zero to obtain a target asymmetry index value of each lesion voxel.

[0107] Step S610, based on the asymmetry index value of each non-lesion voxel in the target symmetric scan image and the target scan voxel value of each lesion voxel, the target symmetric scan image is converted into a medical format image to obtain an asymmetry index image.

[0108] Step S611, performing image partition processing on the asymmetric index image using a preset image partition strategy to obtain an asymmetric index image of each partition.

[0109] Step S612, identifying the volume value of each partition asymmetry index image, and marking the volume value of each partition asymmetry index image in each partition asymmetry index image, to obtain the user's region segmentation mask image.

[0110] Step S613, for each region segmentation mask image, based on the asymmetry index value of each voxel of the region segmentation mask image and the volume value of the region segmentation mask image, a voxel statistical strategy is used to count the sub-asymmetry index voxel value of each voxel statistical type in the region segmentation mask image.

[0111] Step S614: The sub-asymmetry index voxel value of each voxel statistical type in all region segmentation mask images is used as the asymmetry index voxel value of the user's scanned image.

[0112] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0113] Based on the same inventive concept, the embodiment of the present application also provides a statistical device for the asymmetry index voxel value of a scanned image for implementing the statistical method for the asymmetry index voxel value of a scanned image involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in the embodiment of the statistical device for the asymmetry index voxel value of one or more scanned images provided below can refer to the limitations of the statistical method for the asymmetry index voxel value of a scanned image above, and will not be repeated here.

[0114] In an exemplary embodiment, Figure 7 As shown, a statistical device for calculating the asymmetry index voxel value of a scanned image is provided, comprising: an acquisition module 710, a generation module 720 and a statistical module 730, wherein:

[0115] The acquisition module 710 is used to acquire the scanned image of the user and identify the edge information of the lesion in the scanned image through an image edge recognition algorithm;

[0116] A generating module 720 is used to generate a mask image on the opposite side of the scanned image based on the lesion edge information, and to generate an asymmetry index image based on the lesion edge information and the mask image on the opposite side of the lesion by an asymmetry index algorithm;

[0117] The statistical module 730 is used to generate a regional segmentation mask image of the user based on the asymmetry index image, and to count the asymmetry index voxel values ​​of the scanned image of the user through a voxel statistical strategy based on the regional segmentation mask image.

[0118] Optionally, the acquisition module 710 is specifically configured to:

[0119] Identifying the image type of the scanned image through an image recognition network, and acquiring the lesion type of the user in response to a lesion upload operation by a staff member;

[0120] Based on the image type, querying each sample image corresponding to the image type in a historical database, and based on the lesion type, screening each sample image corresponding to the scanned image in each sample image;

[0121] Based on each of the sample images, an initial image edge recognition algorithm is trained to obtain an image edge recognition algorithm, and the scanned image is input into the image edge recognition algorithm to obtain lesion edge information of the scanned image.

[0122] Optionally, the generating module 720 is specifically configured to:

[0123] Based on the lesion edge information, identifying the lesion range of the lesion area of ​​the scanned image in the scanned image, and equally dividing the scanned image to obtain symmetrical scanned images;

[0124] Based on the lesion range of the lesion area in the scanned image, identifying the target symmetrical scanned image to which the lesion area belongs, and based on the target symmetrical scanned image, generating a mirror image of the target symmetrical scanned image through mirror mapping processing;

[0125] The target symmetrical scanned image and the mirror image are subjected to image stitching processing to obtain a mask image on the opposite side of the scanned image.

[0126] Optionally, the generating module 720 is specifically configured to:

[0127] Identifying each scanned voxel value of the target symmetrical scanned image and each mirror voxel value in the mirror image, and calculating the asymmetry index value of each voxel of the target symmetrical scanned image by an asymmetry index algorithm based on each scanned voxel value and each mirror voxel value;

[0128] In each voxel, each lesion voxel within the lesion range is selected, and the asymmetry index value of each lesion voxel is reset to zero to obtain a target asymmetry index value of each lesion voxel;

[0129] Based on the asymmetry index value of each non-lesion voxel of the target symmetric scan image and the target scan voxel value of each lesion voxel, the target symmetric scan image is converted into a medical format image to obtain an asymmetry index image.

[0130] Optionally, the statistics module 730 is specifically used for:

[0131] The asymmetric index image is subjected to image partition processing by a preset image partitioning strategy to obtain an asymmetric index image of each partition;

[0132] The volume value of each partition asymmetry index image is identified, and the volume value of each partition asymmetry index image is marked in each partition asymmetry index image to obtain each region segmentation mask image of the user.

[0133] Optionally, the statistics module 730 is specifically used for:

[0134] For each regional segmentation mask image, based on the asymmetry index value of each voxel of the regional segmentation mask image and the volume value of the regional segmentation mask image, a sub-asymmetry index voxel value of each voxel statistical type in the regional segmentation mask image is counted by a voxel statistical strategy;

[0135] The sub-asymmetry index voxel value of each voxel statistical type in all region segmentation mask images is used as the asymmetry index voxel value of the scanned image of the user.

[0136] Each module in the statistical device for calculating the voxel value of the asymmetry index of the scanned image can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each module.

[0137] In an exemplary embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as shown in FIG. Figure 8As shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and the external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be realized through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a statistical method for the asymmetric index voxel value of a scanned image is implemented. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.

[0138] Those skilled in the art will understand that Figure 8 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0139] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps of any one of the methods in the first aspect when executing the computer program.

[0140] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of any one of the methods in the first aspect are implemented.

[0141] In one embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the steps of any one of the methods in the first aspect.

[0142] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0143] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0144] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0145] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A statistical method for calculating the voxel value of the asymmetry index of a scanned image, characterized in that: The method comprises: Obtaining a scanned image of the user, and identifying the edge information of the lesion in the scanned image through an image edge recognition algorithm; Based on the lesion edge information, identifying the lesion range of the lesion area of ​​the scanned image in the scanned image, and equally dividing the scanned image to obtain symmetrical scanned images; Based on the lesion range of the lesion area in the scanned image, identifying the target symmetrical scanned image to which the lesion area belongs, and based on the target symmetrical scanned image, generating a mirror image of the target symmetrical scanned image through mirror mapping processing; Performing image stitching processing on the target symmetrical scanned image and the mirror image to obtain a mask image on the opposite side of the scanned image; Based on the lesion edge information and the mask image on the opposite side of the lesion, an asymmetry index image is generated by an asymmetry index algorithm; The asymmetric index image is subjected to image partition processing by a preset image partitioning strategy to obtain an asymmetric index image of each partition; Identify the volume value of each partition asymmetry index image, and mark the volume value of each partition asymmetry index image in each partition asymmetry index image to obtain each region segmentation mask image of the user; For each regional segmentation mask image, based on the asymmetry index value of each voxel of the regional segmentation mask image and the volume value of the regional segmentation mask image, a sub-asymmetry index voxel value of each voxel statistical type in the regional segmentation mask image is counted by a voxel statistical strategy; The sub-asymmetry index voxel value of each voxel statistical type in all region segmentation mask images is used as the asymmetry index voxel value of the scanned image of the user.

2. The method according to claim 1, characterized in that The step of identifying the edge information of the lesion in the scanned image by using an image edge recognition algorithm includes: Identifying the image type of the scanned image through an image recognition network, and acquiring the lesion type of the user in response to a lesion upload operation by a staff member; Based on the image type, querying each sample image corresponding to the image type in a historical database, and based on the lesion type, screening each sample image corresponding to the scanned image in each sample image; Based on each of the sample images, an initial image edge recognition algorithm is trained to obtain an image edge recognition algorithm, and the scanned image is input into the image edge recognition algorithm to obtain lesion edge information of the scanned image.

3. The method according to claim 1, characterized in that: The step of generating an asymmetry index image based on the lesion edge information and the mask image on the opposite side of the lesion by using an asymmetry index algorithm comprises: Identifying each scanned voxel value of the target symmetrical scanned image and each mirror voxel value in the mirror image, and calculating the asymmetry index value of each voxel of the target symmetrical scanned image by an asymmetry index algorithm based on each scanned voxel value and each mirror voxel value; In each voxel, each lesion voxel within the lesion range is selected, and the asymmetry index value of each lesion voxel is reset to zero to obtain a target asymmetry index value of each lesion voxel; Based on the asymmetry index value of each non-lesion voxel of the target symmetric scan image and the target scan voxel value of each lesion voxel, the target symmetric scan image is converted into a medical format image to obtain an asymmetry index image.

4. A statistical device for calculating the voxel value of the asymmetry index of a scanned image, characterized in that: The device comprises: An acquisition module is used to acquire a scanned image of a user and identify the edge information of a lesion in the scanned image through an image edge recognition algorithm; A generation module is used to identify the lesion range of the lesion area of ​​the scanned image in the scanned image based on the lesion edge information, and equally divide the scanned image to obtain symmetrical scanned images; identify the target symmetrical scanned image to which the lesion area belongs based on the lesion range of the lesion area in the scanned image, and generate a mirror image of the target symmetrical scanned image through mirror mapping based on the target symmetrical scanned image; perform image stitching processing on the target symmetrical scanned image and the mirror image to obtain a lesion contralateral mask image of the scanned image; generate an asymmetry index image through an asymmetry index algorithm based on the lesion edge information and the lesion contralateral mask image; A statistical module is used to perform image partitioning processing on the asymmetry index image through a preset image partitioning strategy to obtain each partition asymmetry index image; identify the volume value of each partition asymmetry index image, and mark the volume value of each partition asymmetry index image in each partition asymmetry index image to obtain each regional segmentation mask image of the user; for each regional segmentation mask image, based on the asymmetry index value of each voxel of the regional segmentation mask image and the volume value of the regional segmentation mask image, through a voxel statistical strategy, count the sub-asymmetry index voxel value of each voxel statistical type in the regional segmentation mask image; and use the sub-asymmetry index voxel value of each voxel statistical type in all regional segmentation mask images as the asymmetry index voxel value of the scanned image of the user.

5. The device according to claim 4, characterized in that The acquisition module is specifically used for: Identifying the image type of the scanned image through an image recognition network, and acquiring the lesion type of the user in response to a lesion upload operation by a staff member; Based on the image type, querying each sample image corresponding to the image type in a historical database, and based on the lesion type, screening each sample image corresponding to the scanned image in each sample image; Based on each of the sample images, an initial image edge recognition algorithm is trained to obtain an image edge recognition algorithm, and the scanned image is input into the image edge recognition algorithm to obtain lesion edge information of the scanned image.

6. The device according to claim 4, characterized in that The generating module is specifically used for: Identifying each scanned voxel value of the target symmetrical scanned image and each mirror voxel value in the mirror image, and calculating the asymmetry index value of each voxel of the target symmetrical scanned image by an asymmetry index algorithm based on each scanned voxel value and each mirror voxel value; In each voxel, each lesion voxel within the lesion range is selected, and the asymmetry index value of each lesion voxel is reset to zero to obtain a target asymmetry index value of each lesion voxel; Based on the asymmetry index value of each non-lesion voxel of the target symmetric scan image and the target scan voxel value of each lesion voxel, the target symmetric scan image is converted into a medical format image to obtain an asymmetry index image.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 3 are implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 3 are implemented.

9. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 3 are implemented.

Images