Spinal structure image template generation method, spinal image analysis method and apparatus

Abstract

The application relates to a spine structure image template generation method, a spine image analysis method and device, and the method comprises the following steps: acquiring a set of vertebral region image data; registering each vertebral region image in the set of vertebral region image data in sequence to obtain target image data corresponding to each vertebra; and based on the position of the vertebra in the vertebral segment, the target image data is summarized to obtain a spine structure image template. Compared with the prior art, the technical scheme based on the application can obtain a spine structure image template with higher efficiency, and the spine structure image template obtained based on the technical scheme of the application can more accurately reflect the vertebral structure distribution.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, in particular to a spine structure image template generation method and a spine image analysis method and device. BACKGROUND

[0002] At present, the analysis of bone structures in spine CT images is mainly performed by clinical physicians based on past experience through manual analysis. However, due to the limited bone structure parts involved in manual analysis, the information dimension provided by the analysis results obtained by this method is limited.

[0003] In recent years, some studies have attempted to automatically locate the vertebral structure based on a deep learning network and label the cancellous bone and cortical bone in the complete vertebra through gray value differences, but this method does not distinguish between the vertebral body and its accessories, and the accuracy of the labeling results needs to be improved. Therefore, based on the existing technology, it is difficult to accurately and efficiently obtain the bone structure analysis results of the spine CT image. SUMMARY

[0004] Therefore, it is necessary to provide a spine structure image template generation method, a spine image analysis method and device to solve the above technical problems.

[0005] In a first aspect, the present application provides a spine structure image template generation method, which comprises:

[0006] obtaining a set of vertebral region image data;

[0007] sequentially registering each vertebral region image in the set of vertebral region image data to obtain target image data corresponding to each vertebral segment;

[0008] based on the vertebral segment positions of each vertebral segment, the target image data is summarized to obtain a spine structure image template.

[0009] In one embodiment, the sequentially registering each vertebral region image in the set of vertebral region image data to obtain target image data corresponding to each vertebral segment comprises:

[0010] selecting a first sample image corresponding to each vertebral segment and a plurality of to-be-registered images corresponding to each first sample image from the set of vertebral region image data; registering each first sample image to each to-be-registered image to obtain a plurality of first registered images corresponding to each first sample image; obtaining registration results of each vertebral segment according to each first sample image and each first registered image; and obtaining the target image data corresponding to each vertebral segment based on each registration result.

[0011] In one of the embodiments, the registering each of the first sample images to each of the images to be registered to obtain a plurality of first registration images corresponding to each of the first sample images comprises:

[0012] sequentially registering the first sample images to each of the images to be registered to obtain intermediate registration images corresponding to each of the images to be registered and first registration deformation variables corresponding to each of the intermediate registration images; inversely registering each of the intermediate registration images according to an average residual error value corresponding to each of the first registration deformation variables to obtain first registration images corresponding to each of the intermediate registration images.

[0013] In one of the embodiments, the obtaining the registration results of the vertebrae according to each of the first sample images and each of the first registration images comprises:

[0014] sequentially registering the first sample images to each of the first registration images to obtain second registration images corresponding to each of the first registration images and second registration deformation variables corresponding to each of the second registration images; inversely registering each of the second registration images according to an average residual error value corresponding to each of the second registration deformation variables to obtain third registration images corresponding to each of the second registration images; and obtaining the registration results of the vertebrae based on an average gray value of each of the third registration images.

[0015] In one of the embodiments, the obtaining the target image data corresponding to each of the vertebrae based on the registration results comprises:

[0016] In a case where the residual error values corresponding to each of the second registration deformation variables all satisfy a preset convergence threshold, obtaining the target image data corresponding to each of the vertebrae according to the registration results.

[0017] In one of the embodiments, the set of vertebral region image data comprises at least one of a set of vertebral region standard images and a set of vertebral region labeled images; and the obtaining the set of vertebral region image data comprises:

[0018] obtaining a plurality of morphological feature points in each of the vertebral region images based on the vertebral region images corresponding to each of the vertebrae; obtaining average morphological feature points corresponding to each of the morphological feature points according to average coordinate positions of each of the morphological feature points; and adjusting each of the vertebral region images so that a distance between each of the morphological feature points and the corresponding average morphological feature point is less than a preset distance threshold to obtain the set of vertebral region standard images or the set of vertebral region labeled images.

[0019] In a second aspect, the present application provides a spinal column image analysis method, which comprises:

[0020] a plurality of vertebral region instance images are acquired;

[0021] Each of the vertebral region instance images is registered with a corresponding spine structure image template to obtain a first registration result corresponding to each of the vertebral region instance images; the spine structure image template is acquired by using the spine structure image template generation method described above;

[0022] An image analysis result containing an image genomics parameter is obtained according to the pixel gray scale distribution in the first registration result.

[0023] In one embodiment, the image analysis result containing the image genomics parameter is obtained according to the pixel gray scale distribution in the first registration result, including:

[0024] Based on a preset analysis accuracy, the vertebral region boundary in each of the first registration results is adjusted to obtain a first registration result after the vertebral region boundary is adjusted; and an image analysis result containing an image genomics parameter is obtained according to the pixel gray scale distribution in the first registration result after the vertebral region boundary is adjusted.

[0025] In a third aspect, the present application also provides a spine structure image template generation device, which includes:

[0026] An image data set acquisition module is configured to acquire a vertebral region image data set.

[0027] A target image data acquisition module is configured to sequentially register each vertebral region image in the vertebral region image data set to obtain target image data corresponding to each vertebral region.

[0028] A spine structure image template output module is configured to aggregate each of the target image data based on the vertebral level position of each of the vertebral regions to obtain a spine structure image template.

[0029] In a fourth aspect, the present application also provides a spine image analysis device, which includes:

[0030] A vertebral region instance image acquisition module is configured to acquire a plurality of vertebral region instance images.

[0031] A first registration result acquisition module is configured to register each of the vertebral region instance images with a corresponding spine structure image template to obtain a first registration result corresponding to each of the vertebral region instance images; the spine structure image template is acquired by using the spine structure image template generation method described above.

[0032] An image analysis result output module is configured to obtain an image analysis result containing an image genomics parameter according to the pixel gray scale distribution in the first registration result.

[0033] In a fifth aspect, the present application provides a computer device. The computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method.

[0034] In a sixth aspect, the present application provides a computer readable storage medium. The computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the steps of the method.

[0035] In a seventh aspect, the present application provides a computer program product. The computer program product comprises a computer program, and the computer program is executed by a processor to implement the steps of the method.

[0036] The above-mentioned spinal structure image template generation method first acquires a set of vertebral region image data. Then, each vertebral region image in the set of vertebral region image data is sequentially registered to obtain target image data corresponding to each vertebral segment. Finally, based on the vertebral segment positions of each vertebral segment, the target image data is summarized to obtain a spinal structure image template. Based on the vertebral segment positions of each vertebral segment, the target image data corresponding to each vertebral segment is summarized to obtain a spinal structure image template. Compared with the prior art, the efficiency of obtaining a spinal structure image template based on the technical solution of the present application is higher, and the spinal structure image template obtained based on the technical solution of the present application can more accurately reflect the vertebral structure distribution.

[0037] The above-mentioned spinal image analysis method first acquires a plurality of vertebral region instance images. Then, each vertebral region instance image is registered with the corresponding spinal structure image template to obtain a first registration result corresponding to each vertebral region instance image. Finally, according to the pixel gray scale distribution in the first registration result, an image analysis result containing radiomics parameters is obtained. Based on the registration result of the vertebral region instance image and the spinal structure image template, the present application obtains the image analysis result of the vertebral region instance image containing a plurality of radiomics parameters, which not only effectively improves the efficiency of analyzing the vertebral region instance image, but also effectively improves the accuracy of the image analysis result corresponding to the vertebral region instance image. BRIEF DESCRIPTION OF DRAWINGS

[0038] Figure 1 A flowchart of a spinal structure image template generation method provided in an embodiment;

[0039] Figure 2 A flowchart of a specific way of obtaining target image data corresponding to each vertebral segment provided in an embodiment;

[0040] Figure 3 A flowchart of a specific manner of obtaining a first registration image corresponding to each intermediate registration image provided in an embodiment;

[0041] Figure 4 A flowchart of a specific manner of obtaining a registration result of each vertebra provided in an embodiment;

[0042] Figure 5 A flowchart of a specific manner of obtaining a set of vertebra region image data provided in an embodiment;

[0043] Figure 6 A flowchart of a spinal image analysis method provided in an embodiment;

[0044] Figure 7 A specific manifestation form of a second lumbar vertebra CT standard image and a second lumbar vertebra CT labeled image provided in an embodiment;

[0045] Figure 8 A specific manifestation form of a second lumbar vertebra CT example image and a first registration result corresponding thereto provided in an embodiment;

[0046] Figure 9 A schematic diagram of a whole operation flow of performing spinal image analysis provided in an embodiment;

[0047] Figure 10 A schematic diagram of a specific manner of obtaining an image analysis result containing an imageomic parameter provided in an embodiment;

[0048] Figure 11 A structural block diagram of a spinal structure image template generation device provided in an embodiment;

[0049] Figure 12 A structural block diagram of a spinal image analysis device provided in an embodiment;

[0050] Figure 13 An internal structure diagram of a computer device provided in an embodiment. DETAILED DESCRIPTION

[0051] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application is further described in detail below with reference to the accompanying drawings and embodiments. Unless otherwise defined, all technical and scientific terms used herein have the same meanings as commonly understood by those skilled in the art to which the present application belongs. It should be understood that the specific embodiments described herein are only used to explain the present application and should not be used to limit the present application.

[0052] As used herein, the singular forms "a," "an," and "the" include plural referents unless the context clearly dictates otherwise. It will be further understood that the terms "comprises," "comprising," "includes," "including," or "has," "having," and the like, when used herein, specify the presence of stated features, integers, steps, operations, elements, components, or combinations thereof, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, or combinations thereof. Also, the term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0053] In one embodiment, as shown in Figure 1 a spinal structure image template generation method is provided, comprising the following steps:

[0054] Step S110, acquiring a set of vertebral region image data.

[0055] In this step, the set of vertebral region image data refers to a set of image data containing a plurality of vertebral region images.

[0056] Specifically, the specific form of the plurality of vertebral region images in the set of vertebral region image data can be a plurality of computed tomography (CT) images of the vertebral region.

[0057] In actual application, the set of vertebral region image data can include a plurality of CT images of the cervical spine, thoracic spine, lumbar spine and other parts (i.e. parts related to the spine) obtained from a plurality of research subjects aged between 30 and 75 years old, without surgical experience, without fracture, without spinal tumor, and without congenital deformity.

[0058] Step S120, sequentially registering each vertebral region image in the set of vertebral region image data to obtain target image data corresponding to each vertebral region.

[0059] In this step, the set of vertebral region image data refers to a set of image data containing a plurality of vertebral region images; and the target image data corresponding to each vertebral region refers to standardized image data corresponding to each vertebral region obtained by sequentially registering each vertebral region image in the set of vertebral region image data.

[0060] Specifically, the specific way of sequentially registering each vertebral region image in the set of vertebral region image data can be to sequentially perform image registration processing on each vertebral region image in the set of vertebral region image data by using an automatic nonlinear image matching algorithm and an anatomical marker algorithm, and then to obtain target image data corresponding to each vertebral region.

[0061] In practical applications, the following formula can be used to perform image registration processing on each vertebral region image in the vertebral region image dataset sequentially, employing an automatic nonlinear image matching algorithm and an anatomical labeling algorithm:

[0062] First, assuming that during the registration process of each vertebral region image in the vertebral region image dataset, the displacement of the corresponding point is the point deformation, then the total deformation formed by the sum of all point deformations should be the minimum value, that is:

[0063]

[0064] Where, Φ * Image representing a specific vertebral region; I i ψ represents the i-th study object in the vertebral region image dataset; v represents the three-dimensional spatial coordinates of each pixel in the target image data corresponding to a certain vertebra; ψ i,Φ This represents the process of mapping the target image data corresponding to a certain vertebra to the image of a certain vertebra region of the i-th study subject; ψ i,Φ (v) represents the coordinates v in the target image data corresponding to a certain vertebra, mapped to the three-dimensional spatial coordinates in the image of a certain vertebra region of the i-th study subject.

[0065] Then, assuming that during the registration process of each vertebral region image in the vertebral region image dataset, the square of the gray-level difference of corresponding points is the point residual, then the sum of the residuals of all points should be the minimum value, that is:

[0066]

[0067] Where Φ(v) represents the gray value at coordinate v in the target image corresponding to a certain vertebra; I i (ψ i,Φ (v) indicates that the i-th research subject was subjected to ψ i,Φ Grayscale values ​​after registration.

[0068] Furthermore, for an image of a specific vertebral region in the i-th study subject, ψ i,Φ The following formula should be satisfied:

[0069]

[0070] Where, ψ i,Φ This represents the process of mapping the target image data corresponding to a certain vertebra to the image of a certain vertebra region of the i-th study subject.

[0071] Step S130: Based on the vertebral segment location of each vertebra, the target image data are summarized to obtain a spinal structure image template.

[0072] In this step, the vertebral level position of each segmental vertebra is determined based on the distribution of each segmental vertebra in the human body; each target image data is the target image data corresponding to each segmental vertebra; and the spinal structure image template is obtained by summarizing each target image data based on the vertebral level position of each segmental vertebra.

[0073] The spinal structure image template generation method described above first acquires a set of vertebral region image data. Then, each vertebral region image in the set of vertebral region image data is registered in sequence to obtain target image data corresponding to each segmental vertebra. Finally, each target image data is summarized based on the vertebral level position of each segmental vertebra to obtain a spinal structure image template. Based on the vertebral level position of each segmental vertebra, the target image data corresponding to each segmental vertebra is summarized to obtain a spinal structure image template. Compared with the prior art, the efficiency of obtaining a spinal structure image template based on the technical solution of the present application is higher, and the spinal structure image template obtained based on the technical solution of the present application can more accurately reflect the vertebral structure distribution.

[0074] For the specific way of obtaining target image data corresponding to each segmental vertebra, in one embodiment, as shown in Figure 2 the step S120 specifically includes:

[0075] In step S210, a first sample image corresponding to each segmental vertebra and a plurality of to-be-registered images corresponding to each first sample image are selected from the set of vertebral region image data.

[0076] In this step, the set of vertebral region image data refers to a set of image data containing a plurality of vertebral region images; the first sample image corresponding to each segmental vertebra refers to a first sample image selected from the set of vertebral region image data as a reference image for image registration of each vertebral region image corresponding to each segmental vertebra; and the plurality of to-be-registered images corresponding to each first sample image refers to a plurality of to-be-registered images corresponding to each first sample image selected from the set of vertebral region image data.

[0077] In actual application, the specific way of selecting a first sample image corresponding to a certain segmental vertebra and a plurality of to-be-registered images corresponding to the first sample image from the set of vertebral region image data can be to select a CT image of a certain segmental vertebra (for example, the first cervical vertebra) of a certain research subject from the set of vertebral region image data as the first sample image corresponding to the segmental vertebra, and to select the CT images of the segmental vertebra of other research subjects in the set of vertebral region image data as the plurality of to-be-registered images corresponding to the first sample image corresponding to the segmental vertebra.

[0078] Step S220, each first sample image is registered to each image to be registered, to obtain a plurality of first registration images corresponding to each first sample image.

[0079] In this step, each first sample image, i.e. the first sample image corresponding to each vertebral bone, refers to the first sample image selected from the vertebral region image data set as the reference image for image registration of each vertebral region image corresponding to each vertebral bone; each image to be registered, i.e. a plurality of images to be registered corresponding to each first sample image, refers to a plurality of images to be registered corresponding to each first sample image selected from the vertebral region image data set; and a plurality of first registration images corresponding to each first sample image refers to a plurality of first registration images corresponding to each first sample image obtained by registering each first sample image to each image to be registered.

[0080] In actual application, the specific way of registering each first sample image to each image to be registered to obtain a plurality of first registration images corresponding to each first sample image can be to register each first sample image to each image to be registered in turn, and record the result of each registration as a plurality of first registration images corresponding to each first sample image.

[0081] Step S230, according to each first sample image and each first registration image, obtaining the registration result of each vertebral bone.

[0082] In this step, each first sample image, i.e. the first sample image corresponding to each vertebral bone, refers to the first sample image selected from the vertebral region image data set as the reference image for image registration of each vertebral region image corresponding to each vertebral bone; each first registration image, i.e. a plurality of first registration images corresponding to each first sample image, refers to a plurality of first registration images corresponding to each first sample image obtained by registering each first sample image to each image to be registered; and the registration result of each vertebral bone refers to the registration result of each vertebral bone obtained according to each first sample image and each first registration image.

[0083] Step S240, based on each registration result, obtaining the target image data corresponding to each vertebral bone.

[0084] In this step, each registration result, i.e. the registration result of each vertebral bone, refers to the registration result of each vertebral bone obtained according to each first sample image and each first registration image; and the target image data corresponding to each vertebral bone refers to the target image data corresponding to each vertebral bone obtained based on each registration result.

[0085] The above embodiment effectively improves the acquisition efficiency of the target image data for generating the image template of the spine structure by selecting the first sample image corresponding to each vertebra from the set of vertebral region image data, selecting a plurality of to-be-registered images corresponding to each first sample image, and acquiring a plurality of registration results for generating the target image data of each vertebra according to each first sample image and each first registration image.

[0086] For the specific manner of acquiring the first registration image corresponding to each intermediate registration image, in one embodiment, as shown in Figure 3 The step S220 specifically includes:

[0087] In step S310, the first sample image is sequentially registered to each to-be-registered image to obtain an intermediate registration image corresponding to each to-be-registered image and a first registration deformation variable corresponding to each intermediate registration image.

[0088] In this step, the first sample image, i.e., the first sample image corresponding to each vertebra, is a first sample image selected from the set of vertebral region image data as a reference image for image registration of each vertebral region image corresponding to each vertebra; each to-be-registered image, i.e., a to-be-registered image corresponding to the first sample image; each intermediate registration image corresponding to each to-be-registered image is an intermediate registration image corresponding to each to-be-registered image obtained by sequentially registering the first sample image to a plurality of to-be-registered images corresponding thereto; and the first registration deformation variable corresponding to each intermediate registration image is a first registration deformation variable corresponding to each intermediate registration image generated in each registration process of sequentially registering the first sample image to a plurality of to-be-registered images corresponding thereto.

[0089] In step S320, each intermediate registration image is inversely registered according to the average residual error value corresponding to each first registration deformation variable to obtain a first registration image corresponding to each intermediate registration image.

[0090] In this step, the average residual error value corresponding to each first registration deformation variable is an average residual error value calculated based on the first registration deformation variables corresponding to a plurality of intermediate registration images; and the first registration image corresponding to each intermediate registration image is a first registration image corresponding to each intermediate registration image obtained by inversely registering each intermediate registration image according to the average residual error value corresponding to each first registration deformation variable.

[0091] In actual applications, after inversely registering each intermediate registration image according to the average residual error value corresponding to each first registration deformation variable to obtain a first registration image corresponding to each intermediate registration image (i.e., after step S320), a preliminary registration result of each vertebra can also be obtained based on the average gray value of the first registration image corresponding to each intermediate registration image.

[0092] The above embodiment effectively guarantees the accuracy of the registration results of each vertebra based on the first registration image by sequentially registering the first sample image to each to-be-registered image, obtaining the intermediate registration image corresponding to each to-be-registered image, and the first registration deformation variable corresponding to each intermediate registration image, and inversely registering each intermediate registration image according to the average residual value corresponding to each first registration deformation variable to obtain the first registration image corresponding to each intermediate registration image, thereby effectively ensuring the accuracy of the target image data used to generate the spinal structure image template.

[0093] For the specific manner of obtaining the registration results of each vertebra, in one embodiment, as shown in Figure 4 The above step S230 specifically includes:

[0094] Step S410, sequentially register the first sample image to each first registration image to obtain the second registration image corresponding to each first registration image and the second registration deformation variable corresponding to each second registration image.

[0095] In this step, the first sample image, i.e., the first sample image corresponding to each vertebra, refers to the first sample image selected from the set of vertebral region image data as the reference image for image registration of each vertebral region image corresponding to each vertebra; each first registration image, i.e., the first registration image corresponding to each first sample image, refers to the first registration image corresponding to each first sample image obtained by registering each first sample image to each to-be-registered image; the second registration image corresponding to each first registration image refers to the second registration image corresponding to each first registration image obtained by sequentially registering the first sample image to each first registration image; and the second registration deformation variable corresponding to each second registration image refers to the second registration deformation variable corresponding to each second registration image generated in each registration process of sequentially registering the first sample image to the corresponding first registration image.

[0096] Step S420, inversely register each second registration image according to the average residual value corresponding to each second registration deformation variable to obtain the third registration image corresponding to each second registration image.

[0097] In this step, the average residual value corresponding to each second registration deformation variable refers to the average residual value calculated based on the second registration deformation variable corresponding to each second registration image; and the third registration image corresponding to each second registration image refers to the third registration image corresponding to each second registration image obtained by inversely registering each second registration image according to the average residual value corresponding to each second registration deformation variable.

[0098] Step S430, based on the average gray value of each third registration image, the registration result of each vertebral segment is obtained.

[0099] In this step, each third registration image, i.e. the third registration image corresponding to each second registration image, refers to the third registration image corresponding to each second registration image obtained by inversely registering each second registration image according to the average residual value corresponding to each second registration deformation variable; the registration result of each vertebral segment refers to the registration result of each vertebral segment obtained based on the average gray value of each third registration image.

[0100] In practical application, the specific implementation process of steps S210 to S430 can include the following steps:

[0101] 1. In the vertebral region image data set containing n standard CT images, a first cervical vertebra I 1,C1 of a research subject is randomly selected as the starting first cervical vertebra standard CT image (i.e. the first sample image), named T C1-0 ;

[0102] 2. T C1-0 is registered to the first cervical vertebra I i,C1 of the i-th research subject one by one, the intermediate registration image is calculated and named X i,C1-0 , and the deformation variable generated by registration is calculated;

[0103] 3. The average residual of the deformation variable generated in n-1 registrations is calculated and named D C1-0 ;

[0104] 4. In the i-th research subject, D C1-0 and X i,C1-0 are inversely registered to obtain the corrected first cervical vertebra CT image of the i-th research subject, named Y i,C1-0 (i.e. the first registration image described above);

[0105] 5. The average of the sum of the gray values of the corrected first cervical vertebra CT images Y i,C1-0 of all research subjects is calculated, which is the standard CT image of the first cervical vertebra after 1 iteration, named T C1-1 (i.e. the preliminary registration result described above);

[0106] 6. T C1-0 is registered to the corrected first cervical vertebra CT image Y i,C1-0 of the i-th research subject one by one, the second registration image is calculated and named X i,C1-1 , and the deformation variable generated by registration is calculated;

[0107] 7. The average residual of the deformation variable generated in n-1 registrations is calculated and named DC1-1 ;

[0108] 8. In the ith study subject, based on the D C1-1 and X i,C1-1 , the corrected C1 CT image of the ith study subject is obtained by inverse registration, and is named as Y i,C1-1 (i.e. the third registration image described above);

[0109] 9. The average of the gray values of the corrected C1 CT images Y i,C1-1 of all the study subjects is calculated, and is the standard CT image of the first cervical vertebra after two iterations, and is named as T C1-2 (i.e. the registration result described above).

[0110] The above embodiment obtains the second registration image corresponding to each first registration image by sequentially registering the first sample image to each first registration image, obtains the second registration deformation variable corresponding to each second registration image, performs inverse registration on each second registration image according to the average residual value corresponding to each second registration deformation variable, obtains the third registration image corresponding to each second registration image, and obtains the registration result of each vertebra based on the average gray value of each third registration image, which ensures the accuracy of the registration result of each vertebra obtained based on the third registration image, and further effectively ensures the accuracy of the target image data used to generate the spinal structure image template.

[0111] For a specific manner of obtaining the target image data corresponding to each vertebra according to the registration result, in one embodiment, the step S240 specifically comprises:

[0112] When it is confirmed that the residual values corresponding to each second registration deformation variable all satisfy the preset convergence threshold, the target image data corresponding to each vertebra is obtained according to the registration result.

[0113] The second registration deformation variable corresponding to each second registration image refers to the second registration deformation variable corresponding to each second registration image generated in each registration process of sequentially registering the first sample image to the corresponding first registration image; the preset convergence threshold refers to a preset convergence threshold for determining whether the residual value corresponding to each second registration deformation variable satisfies the convergence condition; the registration result refers to the registration result of each vertebra obtained based on the average gray value of each third registration image; and the target image data corresponding to each vertebra refers to the target image data corresponding to each vertebra obtained according to the registration result when it is confirmed that the residual values corresponding to each second registration deformation variable all satisfy the preset convergence threshold.

[0114] In actual application, in a case where it is confirmed that the residual error value corresponding to a second registration deformation variable does not satisfy the preset convergence threshold, the above steps S410 to S430 can be repeatedly executed until it is confirmed that the residual error values corresponding to all the second registration deformation variables satisfy the preset convergence threshold, so as to obtain the target image data corresponding to each vertebral segment.

[0115] The above embodiment, by obtaining the target image data corresponding to each vertebral segment according to the registration results in a case where it is confirmed that the residual error values corresponding to all the second registration deformation variables satisfy the preset convergence threshold, not only effectively improves the efficiency of obtaining the target image data corresponding to each vertebral segment, guarantees the accuracy of the target image data corresponding to each vertebral segment, but also effectively improves the accuracy of the spinal structure image template obtained by aggregating the target image data corresponding to each vertebral segment.

[0116] For the specific manner of obtaining the vertebral region image data set, in one embodiment, the above vertebral region image data set includes at least one of a vertebral region standard image set and a vertebral region labeled image set; as shown in Figure 5 The above step S110 specifically includes:

[0117] Step S510, based on the vertebral region image corresponding to each vertebral segment, obtaining a plurality of morphological feature points in each vertebral region image.

[0118] In this step, each vertebral region image refers to each vertebral region image contained in the vertebral region image data set; and the plurality of morphological feature points refer to a plurality of feature points in each vertebral region image, which represent the transverse process tip, the spinous process tip, and the like, and have distinctive anatomical morphology.

[0119] Specifically, the specific form of each vertebral region image can be each vertebral region standard image in the vertebral region standard image set, or each vertebral region labeled image in the vertebral region labeled image set; the specific form of the vertebral region standard image can be a vertebral region standard CT image in the form of a three-dimensional image, which can include vertebral standard CT images corresponding to 7 cervical vertebrae, 12 thoracic vertebrae, and 5 lumbar vertebrae; the specific form of the vertebral region labeled image can be a vertebral region labeled CT image in which the cancellous bone, cortical bone, and vertebral basilar vein foramen of the vertebral body, vertebral arch, and vertebral arch root regions of different vertebral segments are labeled respectively; the vertebral region standard CT image and the vertebral region labeled CT image each correspond to an image size and resolution, and there is a one-to-one matching relationship between them.

[0120] In practical applications, the specific manner of obtaining the vertebral region images corresponding to each vertebral segment can be to classify each vertebral region image according to the vertebral segment where it is located, to obtain a plurality of vertebral region images corresponding to each vertebral segment. Based on this, the specific manner of obtaining a plurality of morphological feature points in each vertebral region image based on the vertebral region images corresponding to each vertebral segment can be to classify each vertebral region image according to the vertebral segment where it is located, to obtain a plurality of vertebral region images corresponding to each vertebral segment, and then determine the coordinate positions of a plurality of morphological feature points existing in each vertebral region image corresponding to the same vertebral segment; the specific obtaining steps of the vertebral region standard CT image can include: first, cutting each original spinal standard CT image after uniform resolution along the vertebral center point to obtain an image of uniform size, then rotating the cut image to make the vertebral central axis in the image perpendicular to the horizontal plane, thereby completing the conversion from the original spinal standard CT image to the vertebral region standard CT image; the original spinal standard CT image can be an original spinal standard CT image obtained from the original spinal CT image data set, which can include a plurality of CT images of the cervical spine, thoracic spine, lumbar spine and other parts obtained from a plurality of research subjects aged 30 to 75 years old, without surgery, without fracture, without spinal tumor, without congenital deformity, and a plurality of original spinal standard CT images constructed segment by segment; the specific obtaining steps of the vertebral region labeled CT image are similar to the specific obtaining steps of the vertebral region standard CT image, except that each original spinal standard CT image in the specific obtaining steps of the vertebral region standard CT image is replaced by each original spinal labeled CT image; the original spinal labeled CT image can be obtained by labeling the cancellous bone, cortical bone and vertebral venous foramen of the vertebral body, vertebral arch and pedicle of different vertebral segments in each original spinal standard CT image. Since the original spinal labeled CT image can be used to represent the locations of cancellous bone, cortical bone and vertebral venous foramen in the vertebral body, vertebral arch and pedicle of different vertebral segments, based on this, an image set composed of a plurality of original spinal labeled CT images can be referred to as an original spinal atlas.

[0121] In step S520, the average morphological feature point corresponding to each morphological feature point is obtained according to the average coordinate position of each morphological feature point.

[0122] In this step, the morphological feature points refer to feature points in each vertebral region image, which represent the transverse process tip, the spinous process tip and the like with distinctive anatomical morphology. The average coordinate position of each morphological feature point refers to the average coordinate position of each morphological feature point in each vertebral region image corresponding to the same vertebral segment. The average morphological feature point corresponding to each morphological feature point refers to obtaining the average coordinate position of each morphological feature point in each vertebral region image according to the average coordinate position of each morphological feature point in each vertebral region image corresponding to the same vertebral segment, and determining the aforementioned average coordinate position as the coordinate position of the average morphological feature point corresponding to each morphological feature point.

[0123] Step S530 adjusts each vertebral region image to obtain a vertebral region standard image set or a vertebral region labeled image set, with the distance between each morphological feature point and the corresponding average morphological feature point being less than a preset distance threshold.

[0124] In this step, the preset distance threshold refers to a preset distance threshold for adjusting the distance between the coordinate position of each morphological feature point and the coordinate position of the average morphological feature point corresponding to each morphological feature point.

[0125] Specifically, the smaller the preset distance threshold, the closer the distance between the coordinate position of each morphological feature point and the coordinate position of the average morphological feature point corresponding to each morphological feature point, and the higher the similarity between each morphological feature point and the average morphological feature point corresponding to each morphological feature point. The specific form of the preset distance threshold can be a minimum distance value between the coordinate position of each morphological feature point and the coordinate position of the average morphological feature point corresponding to each morphological feature point.

[0126] In actual application, the specific manner of adjusting each vertebral region image can be rotation, scaling, translation, stretching and the like for each vertebral region image, so that the distance between the coordinate position of each morphological feature point in each vertebral region image and the coordinate position of the average morphological feature point corresponding to each morphological feature point can be less than the preset distance threshold, reaching the minimum distance value.

[0127] Further, before the adjusting each vertebra region image to obtain the vertebra region standard image set or the vertebra region labeled image set, each vertebra region image can also be adjusted to the same resolution by resampling each vertebra region image, each vertebra region image can be cropped to a uniform image size by uniform size cropping of each vertebra region image, and the gray value range corresponding to each vertebra region image can be made as close as possible by CT window width and window level adjustment of each vertebra region image. The execution order of the foregoing resampling, the foregoing uniform size cropping, and the foregoing CT window width and window level adjustment can be sequential execution or separate execution, and the specific execution steps of the foregoing resampling, the foregoing uniform size cropping, and the foregoing CT window width and window level adjustment are not strictly limited. The specific form of the spinal structure image template obtained through the above step S130 can include at least one of a vertebra region standard CT image template obtained based on a vertebra region standard image and a vertebra region labeled CT image template obtained based on a vertebra region labeled image. The vertebra region labeled CT image template obtained based on a vertebra region labeled image can be used to represent the specific position distribution of structures such as vertebral spongiosa and pedicle spongiosa in each vertebra region. Based on this, an image set composed of a plurality of vertebra region labeled CT image templates can be referred to as a vertebra sub-region map set.

[0128] The above embodiment obtains average morphological feature points corresponding to each morphological feature point based on the average value of the coordinate positions of each morphological feature point in each vertebra region image, and adjusts each vertebra region image to obtain a vertebra region standard image set or a vertebra region labeled image set, so that each vertebra region labeled image in the vertebra region labeled image set can more accurately reflect the distribution of the vertebra structure, thereby effectively ensuring the data accuracy of the spinal structure image template obtained based on the vertebra region standard image set or the vertebra region labeled image set.

[0129] In one embodiment, as shown in FIG. 6, a spinal image analysis method is provided, including the following steps: Figure 6

[0130] Step S610: Obtain a plurality of vertebra region instance images.

[0131] In this step, the plurality of vertebra region instance images refer to a plurality of vertebra region instance images of patients with image analysis requirements.

[0132] ​Specifically, the specific form of the vertebral region instance image can be a plurality of CT instance images of the vertebral region of the patient with image analysis requirements.

[0133] In practical applications, the specific manner of obtaining a plurality of vertebral region instance images can include the following steps: first, a plurality of original spine CT instance images of the patient with image analysis requirements are collected; then, each original spine CT instance image is processed according to the above-mentioned "conversion between the original spine standard CT image and the vertebral region standard CT image" step, and the corresponding vertebral region CT instance image of each original spine CT instance image is obtained.

[0134] In step S620, each vertebral region instance image is registered with the corresponding spine structure image template to obtain a first registration result corresponding to each vertebral region instance image. The aforementioned spine structure image template is obtained by using the above-mentioned spine structure image template generation method.

[0135] In this step, each vertebral region instance image refers to each vertebral region instance image of the patient with image analysis requirements; and the first registration result corresponding to each vertebral region instance image refers to the first registration result corresponding to each vertebral region instance image obtained by registering each vertebral region instance image with the corresponding spine structure image template.

[0136] Specifically, the specific form of the spine structure image template can include at least one of a vertebral region standard CT image template obtained based on a vertebral region standard image and a vertebral region labeled CT image template obtained based on a vertebral region labeled image.

[0137] In practical applications, the specific manner of registering each vertebral region instance image with the corresponding spine structure image template to obtain the first registration result corresponding to each vertebral region instance image can be to register each vertebral region instance image with its corresponding spine structure image template (which can include the above-mentioned vertebral region standard CT image template and the above-mentioned vertebral region labeled CT image template) according to the location of the vertebral segment of the spine, so as to align the image features and spatial coordinates of each vertebral region instance image with its corresponding spine structure image template, and then obtain the first registration result corresponding to each vertebral region instance image (which can include the first registration result between each vertebral region instance image and its corresponding vertebral region standard CT image template, and the first registration result between each vertebral region instance image and its corresponding vertebral region labeled CT image template).

[0138] In step S630, the image analysis result containing the radiomics parameters is obtained according to the pixel gray scale distribution in the first registration result.

[0139] In this step, the first registration result corresponding to each vertebral region instance image, i.e., the first registration result, refers to the first registration result corresponding to each vertebral region instance image obtained by registering each vertebral region instance image with the corresponding spine structure image template; and the image analysis result containing radiomics parameters refers to the image analysis result containing radiomics parameters obtained by calculating each first registration result according to the pixel gray distribution in each first registration result.

[0140] Specifically, the radiomics parameters can include gray mean value, gray peak value, volume, and other forms of radiomics parameters.

[0141] In actual application, the specific way of obtaining the image analysis result containing radiomics parameters according to the pixel gray distribution in the first registration result can be to calculate each first registration result according to the pixel gray value and its relative distribution in the first registration result, and then obtain the image analysis result containing gray mean value, gray peak value, volume, and other forms of radiomics parameters. The image analysis result can be used by an analyst to analyze the morphological changes of different vertebral regions of a patient, the mineral distribution, and other conditions, so as to obtain quantitative analysis results of the mineral concentration distribution in the vertebral body. The specific implementation process of steps S610 to S630 can be specifically shown in the form of Figure 9 as shown (i.e., steps S610 to S630 can be implemented by the following specific operation steps):

[0142] First, a plurality of original spine CT instance images of a patient with image analysis requirements are collected. Each original spine CT instance image is processed according to the above-mentioned “specific acquisition steps of the vertebral region standard CT image” (i.e., the “original spine CT instance image” is used to replace the “original spine standard CT image” in the specific acquisition steps), to obtain a plurality of vertebral region CT instance images corresponding to each original spine CT instance image.

[0143] Then, the plurality of vertebral region CT instance images are registered with the spine structure image template obtained based on steps S110 to S130 to obtain the first registration result corresponding to each vertebral region CT instance image.

[0144] Next, based on the pixel gray distribution in the plurality of first registration results, the parameters of each first registration result are calculated to obtain the image analysis result containing radiomics parameters.

[0145] The above spine image analysis method first acquires a plurality of vertebral region instance images. Then, each vertebral region instance image is registered with a corresponding spine structure image template to obtain a first registration result corresponding to each vertebral region instance image. Finally, an image analysis result containing an imageomic parameter is obtained according to the pixel gray scale distribution in the first registration result. Based on the registration result of the vertebral region instance image and the spine structure image template, the image analysis result containing a plurality of imageomic parameters corresponding to the vertebral region instance image is obtained, which not only effectively improves the efficiency of analyzing the vertebral region instance image, but also effectively improves the accuracy of the image analysis result corresponding to the vertebral region instance image.

[0146] For the specific way of obtaining the image analysis result containing the imageomic parameter, in one embodiment, as shown in Figure 10 the above step S630 specifically includes:

[0147] Step S1010, based on the preset analysis accuracy, adjusting the vertebral region boundary in each first registration result to obtain a first registration result after adjusting the vertebral region boundary.

[0148] In this step, the preset analysis accuracy refers to the preset analysis accuracy that meets the analysis requirements of image analysis on a plurality of vertebral region instance images; each first registration result, i.e., the first registration result corresponding to each vertebral region instance image, refers to the first registration result corresponding to each vertebral region instance image obtained by registering each vertebral region instance image with a corresponding spine structure image template; and the first registration result after adjusting the vertebral region boundary refers to the first registration result after adjusting the vertebral region boundary in each first registration result based on the preset analysis accuracy.

[0149] Specifically, the specific form of the preset analysis accuracy can be a probability threshold that can be used to adjust the vertebral region boundary range in each first registration result. By adjusting the probability threshold of the vertebral region boundary range in each first registration result, the precise control of the vertebral region boundary distribution in each first registration result can be achieved.

[0150] In actual application, it is assumed that the specific form of the second lumbar vertebra CT standard image is as shown in the left part of Figure 7 the specific form of the second lumbar vertebra CT standard image is as shown in the left part of Figure 7The probability threshold of the vertebral region boundary range in the first registration result corresponding to the second lumbar CT instance image can be set to 0.5. The first registration result can be in the form shown in the right part of FIG. 1C. Figure 8 The vertebral region instance image used to obtain the first registration result can be in the form shown in the right part of FIG. 1C. Figure 8 The vertebral region instance image used to obtain the first registration result can be in the form shown in the right part of FIG. 1C.

[0151] In step S1020, the image analysis result containing the radiomics parameters is obtained according to the pixel gray distribution in the first registration result after adjustment of the vertebral region boundary.

[0152] In this step, the first registration result after adjustment of the vertebral region boundary refers to each first registration result after adjustment of the vertebral region boundary in each first registration result based on the preset analysis accuracy. The image analysis result containing the radiomics parameters refers to the image analysis result containing the radiomics parameters obtained by calculating each first registration result after adjustment of the vertebral region boundary according to the pixel gray distribution in each first registration result after adjustment of the vertebral region boundary.

[0153] Specifically, the radiomics parameters can include the gray mean value, the gray peak value, the volume, and the like.

[0154] The embodiments of the present application adjust the vertebral region boundary in each first registration result based on the preset analysis accuracy to obtain the first registration result after adjustment of the vertebral region boundary, and obtain the image analysis result containing the radiomics parameters according to the pixel gray distribution in the first registration result after adjustment of the vertebral region boundary. In this way, the fine adjustment of the vertebral region boundary in the first registration result is realized, which meets the clinical diagnosis requirements and effectively improves the analysis efficiency of the vertebral region instance image.

[0155] It should be understood that although each step in the flowchart involved in each embodiment as described above is displayed in sequence according to the arrow, these steps are not necessarily executed in sequence according to the arrow. Unless otherwise specified herein, the execution of these steps is not strictly limited in sequence, and these steps can be executed in other sequences. Moreover, at least part of the steps in the flowchart involved in each embodiment as described above can include multiple steps or stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution sequence of these steps or stages is not necessarily sequential, but can be executed in rotation or alternation with at least part of other steps or stages in other steps.

[0156] Based on the same inventive concept, the present application also provides a spinal structure image template generation device for implementing the above-mentioned spinal structure image template generation method. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme described in the above method, so the specific limitations in one or more spinal structure image template generation device embodiments provided below can refer to the limitations of the spinal structure image template generation method described above, which will not be repeated here.

[0157] In one embodiment, as shown in Figure 11 a spinal structure image template generation device is provided, which comprises:

[0158] an image data set obtaining module 1110, configured to obtain a set of vertebral region image data;

[0159] a target image data obtaining module 1120, configured to sequentially register each vertebral region image in the set of vertebral region image data to obtain target image data corresponding to each vertebral segment;

[0160] a spinal structure image template output module 1130, configured to aggregate the target image data based on the vertebral segment positions of each vertebral segment to obtain a spinal structure image template.

[0161] In one embodiment, the target image data obtaining module 1120 is specifically configured to select a first sample image corresponding to each vertebral segment and a plurality of to-be-registered images corresponding to each first sample image from the set of vertebral region image data; register each first sample image to each to-be-registered image to obtain a plurality of first registered images corresponding to each first sample image; obtain registration results of each vertebral segment according to each first sample image and each first registered image; and obtain the target image data corresponding to each vertebral segment based on each registration result.

[0162] In one embodiment, the target image data obtaining module 1120 is further configured to sequentially register the first sample image to each to-be-registered image to obtain an intermediate registered image corresponding to each to-be-registered image and a first registration deformation variable corresponding to each intermediate registered image; and inversely register each intermediate registered image according to an average residual value corresponding to each first registration deformation variable to obtain a first registered image corresponding to each intermediate registered image.

[0163] In one of the embodiments, the target image data obtaining module 1120 is further configured to sequentially register the first sample image to each of the first registration images to obtain a second registration image corresponding to each of the first registration images and a second registration deformation variable corresponding to each of the second registration images; perform inverse registration on each of the second registration images according to an average residual error value corresponding to each of the second registration deformation variables to obtain a third registration image corresponding to each of the second registration images; and obtain the registration result of each of the vertebrae based on an average gray value of each of the third registration images.

[0164] In one of the embodiments, the target image data obtaining module 1120 is further configured to, in a case where it is confirmed that the residual error values corresponding to each of the second registration deformation variables all satisfy a preset convergence threshold, obtain the target image data corresponding to each of the vertebrae according to the registration result.

[0165] In one of the embodiments, the set of vertebra region image data includes at least one of a set of vertebra region standard images and a set of vertebra region labeled images; the image data set obtaining module 1110 is specifically configured to obtain a plurality of morphological feature points in each of the vertebra region images based on the vertebra region images corresponding to each of the vertebrae; obtain an average morphological feature point corresponding to each of the morphological feature points according to an average coordinate position of each of the morphological feature points; and adjust each of the vertebra region images to obtain the set of vertebra region standard images or the set of vertebra region labeled images, with a distance between each of the morphological feature points and the corresponding average morphological feature point being less than a preset distance threshold as a target.

[0166] Each of the above modules in the spinal column structure image template generation apparatus can be realized in whole or in part by software, hardware, and a combination thereof. Each of the above modules can be embedded in or independent of a processor in a computer device in a hardware form, or can be stored in a memory in a computer device in a software form, so as to be called and executed by a processor to perform the operations corresponding to each of the above modules.

[0167] Based on the same inventive concept, the embodiments of the present application also provide a spinal column image analysis apparatus for implementing the spinal column image analysis method described above. The implementation scheme for solving problems provided by the apparatus is similar to the implementation scheme described in the above method, and therefore the specific limitations in one or more spinal column image analysis apparatus embodiments provided below can be referred to the limitations of the spinal column image analysis method described above, which will not be described herein again.

[0168] In one embodiment, as shown in Figure 12 a spinal column image analysis apparatus is provided, which comprises:

[0169] The vertebra region instance image obtaining module 1210 is configured to obtain a plurality of vertebra region instance images.

[0170] The first registration result obtaining module 1220 is configured to register each of the vertebral region instance images with a corresponding spine structure image template to obtain a first registration result corresponding to each of the vertebral region instance images; the spine structure image template is obtained by using the spine structure image template generation method described above;

[0171] The image analysis result output module 1230 is configured to obtain an image analysis result containing an imageomic parameter according to a pixel gray scale distribution in the first registration result.

[0172] In one of the embodiments, the image analysis result output module 1230 is specifically configured to adjust a vertebral region boundary in each of the first registration results based on a preset analysis accuracy to obtain a first registration result after adjustment of the vertebral region boundary; and obtain an image analysis result containing an imageomic parameter according to a pixel gray scale distribution in the first registration result after adjustment of the vertebral region boundary.

[0173] The above modules in the spine image analysis device can be realized by software, hardware, or a combination thereof in whole or in part. The above modules can be embedded in or independent of a processor in a computer device in a hardware form, or can be stored in a memory in a computer device in a software form, so as to be called and executed by a processor to perform operations corresponding to the above modules.

[0174] In one embodiment, a computer device is provided, which can be a terminal, and an internal structure diagram of the computer device can be as shown in FIG. 8. Figure 13The computer device shown in the figure includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. Among them, 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. Among them, the processor of the computer device is used to provide computing and control capability. 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 operating system and the computer program in the non-volatile storage medium to run. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals in a wired or wireless manner. The wireless manner can be realized through WIFI, mobile cellular network, NFC (near field communication) or other technologies. The computer program is executed by the processor to realize a spine structure image template generation method or a spine image analysis method. 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. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device, or an external keyboard, touchpad or mouse, etc.

[0175] Those skilled in the art can understand that, Figure 13 The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different component arrangement.

[0176] In one embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps in the above method embodiments.

[0177] In one embodiment, a computer readable storage medium is provided, storing a computer program, and the computer program is executed by a processor to implement the steps in the above method embodiments.

[0178] In one embodiment, a computer program product is provided, including a computer program, and the computer program is executed by a processor to implement the steps in the above method embodiments.

[0179] It should be noted that the user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of related data need to comply with relevant laws, regulations and standards of the country and region.

[0180] A person of ordinary skill in the art can understand that all or part of the processes in the above method embodiments can be completed by instructing related hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above embodiments of the method can be included. In the embodiments provided in the present application, any reference to a memory, a database or other medium can include at least one of a non-volatile and a volatile memory. The non-volatile memory can include a read-only memory (ROM), a magnetic tape, a floppy disk, a flash memory, an optical storage, a high-density embedded non-volatile memory, a resistive memory (ReRAM), a magnetoresistive random access memory (MRAM), a ferroelectric memory (FRAM), a phase change memory (PCM), a graphene memory, etc. The volatile memory can include a random access memory (RAM) or an external cache memory, etc. As an illustration but not limitation, the RAM can be in various forms, such as a static random access memory (SRAM) or a dynamic random access memory (DRAM), etc. The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a block chain, etc., and is not limited thereto. The processor involved in the embodiments provided in the present application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., and is not limited thereto.

[0181] The technical features of the above embodiments can be combined in any manner. To make the description concise, not all possible combinations of the technical features in the above embodiments are described, but as long as the combinations of the technical features do not exist contradictions, they should be considered as the scope of the present application.

[0182] The above-described embodiments are merely illustrative of several embodiments of the present application, and the description is relatively specific and detailed, but should not be understood as a limitation on the scope of the patent. It should be noted that for those skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are all within the scope of the present application. Therefore, the scope of protection of the present application should be subject to the appended claims.

Claims

1. A method for generating a spinal structure image template, characterized in that, The method includes: Obtain a dataset of vertebral region images; From the vertebral region image data set, select the first sample image corresponding to each vertebra and several images to be registered corresponding to each first sample image; The first sample image is sequentially registered to each of the images to be registered, to obtain intermediate registered images corresponding to each of the images to be registered, and first registration deformation corresponding to each of the intermediate registered images; Based on the average residual value corresponding to each of the first registration deformations, reverse registration is performed on each of the intermediate registration images to obtain the first registration image corresponding to each of the intermediate registration images; The first sample image is sequentially registered to each of the first registered images to obtain the second registered image corresponding to each of the first registered images, and the second registered shape corresponding to each of the second registered images; Based on the average residual value corresponding to each of the second registration deformations, reverse registration is performed on each of the second registration images to obtain the third registration image corresponding to each of the second registration images; Based on the average grayscale value of each of the third registration images, the registration result of each vertebra is obtained; based on the registration result, the target image data corresponding to each vertebra is obtained. Based on the vertebral segment location of each vertebra, the target image data are summarized to obtain a spinal structure image template.

2. The method according to claim 1, characterized in that, The process of obtaining the target image data corresponding to each vertebra based on the registration results includes: If the residual values ​​corresponding to each of the second registration deformation variables are confirmed to meet the preset convergence threshold, the target image data corresponding to each vertebra is obtained according to the registration results.

3. The method according to claim 1 or 2, characterized in that, The vertebral region image data set includes at least one of the standard vertebral region image set and the labeled vertebral region image set; The acquired vertebral region image data set includes: Based on the vertebral region images corresponding to each vertebra, multiple morphological feature points are obtained in each vertebral region image. The average morphological feature point is obtained based on the average coordinate position of each morphological feature point. With the goal of the distance between each morphological feature point and the corresponding average morphological feature point being less than a preset distance threshold, the images of each vertebral region are adjusted to obtain a set of standard images of the vertebral region or a set of labeled images of the vertebral region.

4. A method for analyzing spinal images, characterized in that, The method includes: Obtain several instance images of vertebral regions; Each of the vertebral region instance images is registered with the corresponding spinal structure image template to obtain the first registration result corresponding to each of the vertebral region instance images; the spinal structure image template is obtained by the method of any one of claims 1 to 3. Based on the pixel grayscale distribution in the first registration result, image analysis results containing radiomics parameters are obtained.

5. The method according to claim 4, characterized in that, The step of obtaining image analysis results containing radiomics parameters based on the pixel grayscale distribution in the first registration result includes: Based on the preset analysis precision, the vertebral region boundaries in each of the first registration results are adjusted to obtain the first registration results with adjusted vertebral region boundaries. Based on the pixel grayscale distribution in the first registration result after the boundary adjustment of the vertebral region, image analysis results containing radiomics parameters are obtained.

6. A spinal structure image template generation device, characterized in that, The device includes: The image data set acquisition module is used to acquire image data sets of the vertebral region; The target image data acquisition module is used to select, from the vertebral region image data set, a first sample image corresponding to each vertebra and several images to be registered corresponding to each first sample image; sequentially register the first sample images to each image to be registered to obtain intermediate registration images and first registration deformations corresponding to each intermediate registration image; perform reverse registration on each intermediate registration image according to the average residual value corresponding to each first registration deformation to obtain a first registration image corresponding to each intermediate registration image; sequentially register the first sample images to each first registration image to obtain a second registration image and a second registration deformation corresponding to each first registration image; perform reverse registration on each second registration image according to the average residual value corresponding to each second registration deformation to obtain a third registration image corresponding to each second registration image; obtain the registration result of each vertebra based on the average grayscale value of each third registration image; and obtain the target image data corresponding to each vertebra based on the registration results. The spinal structure image template output module is used to summarize the target image data based on the vertebral segment location of each vertebra to obtain a spinal structure image template.

7. A spinal image analysis device, characterized in that, The device includes: The vertebral region instance image acquisition module is used to acquire several vertebral region instance images; The first registration result acquisition module is used to register each of the vertebral region instance images with the corresponding spinal structure image template to obtain the first registration result corresponding to each of the vertebral region instance images; the spinal structure image template is acquired by the method of any one of claims 1 to 3. The image analysis result output module is used to obtain image analysis results containing radiomics parameters based on the pixel grayscale distribution in the first registration result.

Images