Method for screening osteoporosis in patients with spinal deformity using spiral CT images

Abstract

The present invention relates to the technical field of osteoporosis screening, and specifically relates to a method for screening osteoporosis in patients with spinal deformities by using spiral CT images, including: scanning a number of patients with spinal deformities by a spiral CT machine to obtain the coronal CT images, sagittal CT images of the spine of each patient, and cross-sectional CT images of each vertebral node in the spine; obtaining the bone density excellence degree of each vertebral node in the spine of each patient; obtaining the abnormal degree of each intervertebral distance in the intervertebral distance sequence of each patient; obtaining the degree of influence of each intervertebral distance in the intervertebral distance sequence of each patient by osteoporosis; obtaining the osteoporosis risk degree of each vertebral node in the spine of each patient according to the degree of influence, the first degree of influence, and the degree of disruption of the bone density law; and evaluating the osteoporosis status of each patient. The present invention improves the screening accuracy of osteoporosis for patients with spinal deformities.

Description

Technical Field

[0001] The present invention relates to the technical field of osteoporosis screening, and particularly to a method for screening osteoporosis in patients with spinal deformities by using spiral CT images. Background Art

[0002] Spinal deformity is a clinical manifestation that may be caused by various reasons, and osteoporosis is one of them. When osteoporosis occurs, bone density and bone quality decrease, bones become fragile and brittle, and the vertebral bodies of the spine are prone to compression fractures even under slight external force or without obvious external force, thus leading to spinal deformity. Therefore, in medical applications, it is often necessary to combine spiral CT images for auxiliary analysis to achieve accurate screening of osteoporosis in patients with spinal deformities.

[0003] When screening osteoporosis in patients with spinal deformities by existing methods, threshold analysis is often performed on the bone density and vertebral space of the vertebral cross-section in the spinal CT images of the patients. When both the bone density and the vertebral space meet the threshold requirements, it is considered that the patients with spinal deformities have potential osteoporosis risks and need risk warning. However, due to the different biomechanics borne by different vertebral nodes of the spine, there are differences in the bone density of different vertebral nodes of the normal spine itself. For example, the bone density of the lumbar vertebra is greater than that of the cervical vertebra, resulting in low screening accuracy of the consistent threshold analysis of bone density in traditional screening methods. At the same time, in addition to osteoporosis, spinal deformity itself may cause changes in the vertebral space, further affecting the screening accuracy of the traditional method based on the analysis of the vertebral space. Summary of the Invention

[0004] To solve the above problems, the present invention provides a method for screening osteoporosis in patients with spinal deformities by using spiral CT images.

[0005] The method for screening osteoporosis in patients with spinal deformities by using spiral CT images of the present invention adopts the following technical solutions:

[0006] An embodiment of the present invention provides a method for screening osteoporosis in patients with spinal deformities by using spiral CT images, and the method includes the following steps:

[0007] Scanning a number of patients with spinal deformities by a spiral CT machine to obtain the coronal CT images, sagittal CT images of each patient's spine, and transverse CT images of each vertebral node in the spine;

[0008] Segment the cross-sectional CT images to obtain the cancellous bone regions in each cross-sectional CT image; according to the change of the gray value of the pixels in the cancellous bone region, obtain the bone density of each vertebral node in the spine of each patient; according to the gray value distribution of the local regions in the cancellous bone region, obtain the bone disorder degree of each vertebral node in the spine of each patient; according to the bone density and the bone disorder degree, obtain the bone density excellence degree of each vertebral node in the spine of each patient;

[0009] According to the difference in the bone density excellence degree of adjacent vertebral nodes, obtain the degree of bone density law damage of each vertebral node in the spine of each patient; obtain several vertebral node regions in each coronal CT image; obtain the gap distance between adjacent vertebral node regions in the coronal CT image of each patient's spine; obtain the trend item curve of the gap distance between adjacent vertebral node regions in the coronal CT image of each patient's spine; arrange the gap distances between all adjacent vertebral node regions in the coronal CT image of each patient's spine to obtain the gap distance sequence of each patient; according to the offset distance of the corresponding position of the gap distance in the trend item curve, obtain the abnormality degree of each gap distance in the gap distance sequence of each patient;

[0010] According to the area change of the adjacent vertebral node regions corresponding to the gap distance, obtain the edge deformation degree of the two adjacent vertebral node regions corresponding to each gap distance in the gap distance sequence of each patient; obtain the curvature angle value corresponding to each gap distance in the gap distance sequence of each patient; according to the change of the curvature angle value, the edge deformation degree and the abnormality degree, obtain the degree of influence of osteoporosis on each gap distance in the gap distance sequence of each patient;

[0011] Obtain several first vertebral node regions in each sagittal CT image; obtain the first gap distance between adjacent vertebral node regions in the sagittal CT image of each patient's spine; arrange the first gap distances between all adjacent first vertebral node regions in the sagittal CT image of each patient's spine to obtain the first gap distance sequence of each patient; obtain the first degree of influence of osteoporosis on each first gap distance in the first gap distance sequence of each patient;

[0012] According to the degree of influence, the first degree of influence and the degree of bone density law damage, obtain the osteoporosis risk degree of each vertebral node in the spine of each patient; according to the osteoporosis risk degree, evaluate the osteoporosis condition of each patient.

[0013] Further, the specific steps included in obtaining the bone density of each vertebral node in the spine of each patient according to the change of the gray value of the pixels in the cancellous bone region are as follows:

[0014] ;

[0015] In the formula, represents the number of pixels in the cancellous bone region of the cross-sectional CT image of the th vertebral node in the spine of the th patient; represents the gray value of the th pixel in the cancellous bone region of the cross-sectional CT image of the th vertebral node in the spine of the th patient; represents the average gray value of all pixels in the cancellous bone region of the cross-sectional CT image of the th vertebral node in the spine of the th patient; represents taking the absolute value; represents the bone density of the th vertebral node in the spine of the th patient; is the first hyperparameter to prevent the denominator from being zero.

[0016] Furthermore, obtaining the bone disorder degree of each vertebral node in the spine of each patient according to the gray value distribution of the local area in the cancellous bone region includes the following specific steps:

[0017] Perform K-means clustering on all pixels in the cancellous bone region of the cross-sectional CT image of the th vertebral node in the spine of the th patient, using the absolute difference in gray values between pixels as the distance metric, to obtain several clusters in the cancellous bone region of the cross-sectional CT image of the th vertebral node in the spine of the th patient;

[0018] ;

[0019] In the formula, represents the number of clusters in the cancellous bone region of the cross-sectional CT image of the th vertebral node in the spine of the th patient; represents the area of the th cluster in the cancellous bone region of the cross-sectional CT image of the th vertebral node in the spine of the th patient; represents the average area of all clusters in the cancellous bone region of the cross-sectional CT image of the th vertebral node in the spine of the th patient; represents the th patient's spine Degree of bone disorder of the vertebral body nodes.

[0020] Furthermore, obtaining the bone density excellence degree of each vertebral body node in the spine of each patient according to the bone density degree and the bone disorder degree includes the following specific steps:

[0021] For the th patient, the ratio of the bone density degree of the th vertebral body node in the spine to the bone disorder degree of the th vertebral body node in the spine of the th patient is normalized, and the result obtained by the normalization is used as the bone density excellence degree of the th vertebral body node in the spine of the th patient.

[0022] Furthermore, obtaining the degree of bone density law disruption of each vertebral body node in the spine of each patient according to the difference in the bone density excellence degrees of adjacent vertebral body nodes includes the following specific steps:

[0023] For the th patient, the difference value between the bone density excellence degree of the th vertebral body node in the spine and the average value of the bone density excellence degrees of the adjacent vertebral body nodes of the th patient corresponding to the th vertebral body node is normalized, and the result obtained by the normalization is used as the degree of bone density law disruption of the th vertebral body node in the spine of the th patient.

[0024] Furthermore, obtaining the trend item curve of the gap distance in the adjacent vertebral body node regions in the coronal CT image of each patient's spine includes the following specific steps:

[0025] For all the gap distances in the adjacent vertebral body node regions in the coronal CT image of the th patient's spine, a two-dimensional rectangular coordinate system is constructed with the sequential value of the gap distance from top to bottom in the spine as the abscissa and the gap distance as the ordinate; according to the first sequential value and the corresponding gap distance, the point corresponding to the gap distance in the two-dimensional rectangular coordinate system is obtained; all the points corresponding to the gap distances in the two-dimensional rectangular coordinate system are obtained to get the scatter plot of the gap distances in the adjacent vertebral body node regions in the coronal CT image of the th patient; the scatter plot is subjected to STL trend decomposition to obtain the trend item curve of the gap distances in the adjacent vertebral body node regions in the coronal CT image of the th patient, and several standard gap distances are included in the trend item curve.

[0026] Further, obtaining the abnormality degree of each gap distance in the gap distance sequence of each patient according to the offset distance at the corresponding position in the trend item curve includes the following specific steps:

[0027] Taking the absolute difference between the th gap distance in the gap distance sequence of the th patient and the standard gap distance at the corresponding position in the trend item curve as the offset distance of the th gap distance;

[0028] Normalizing the absolute difference between the offset distance of the th gap distance in the gap distance sequence of the th patient and the mean of the offset distances of all gap distances in the gap distance sequence of the th patient, and taking the result of the normalization process as the abnormality degree of the th gap distance in the gap distance sequence of the th patient.

[0029] Further, obtaining the edge deformation degree of the two adjacent vertebral node regions corresponding to each gap distance in the gap distance sequence of each patient according to the area change of the adjacent vertebral node regions corresponding to the gap distance includes the following specific steps:

[0030] Taking the ratio of the mean convex hull area of the two adjacent vertebral node regions corresponding to the th gap distance in the gap distance sequence of the th patient to the mean area of the two adjacent vertebral node regions corresponding to the th gap distance in the gap distance sequence of the th patient as the edge deformation degree of the two adjacent vertebral node regions corresponding to the th gap distance in the gap distance sequence of the th patient.

[0031] Further, obtaining the influence degree of osteoporosis on each gap distance in the gap distance sequence of each patient according to the change of the curvature angle value, the edge deformation degree and the abnormality degree includes the following specific steps:

[0032] ;

[0033] In the formula, represents the edge deformation degree of the two adjacent vertebral node regions corresponding to the th gap distance in the gap distance sequence of the th patient; represents the th gap distance in the gap distance sequence of the The abnormality degree of a gap distance; Indicates the curvature included angle value corresponding to the -th gap distance in the gap distance sequence of the -th patient; Indicates the standard curvature included angle value corresponding to the -th gap distance, and the standard curvature included angle value is a preset value by humans; Indicates taking the absolute value; Indicates a linear normalization function; Indicates the degree of influence of osteoporosis on the -th gap distance in the gap distance sequence of the -th patient; Indicates a hyperparameter to prevent the denominator from being 0.

[0034] Further, obtaining the osteoporosis risk degree of each vertebral node in the spine of each patient according to the influence degree, the first influence degree, and the degree of disruption of the bone density law includes the following specific steps:

[0035] ;

[0036] Indicates the degree of disruption of the bone density law of the -th vertebral node in the spine of the -th patient; Indicates the average value of the influence degrees of osteoporosis on the two adjacent gap distances corresponding to the -th vertebral node in the spine of the -th patient; Indicates the average value of the first influence degrees of osteoporosis on the two adjacent first gap distances corresponding to the -th vertebral node in the spine of the -th patient; Indicates a linear normalization function; Indicates the osteoporosis risk degree of the -th vertebral node in the spine of the

[0037] The beneficial effects of the technical solution of the present invention are as follows: The present invention can analyze by combining the bone density laws of different vertebral nodes in the spine, the differences in the influence of spinal deformity and osteoporosis on the vertebral space, improving the screening accuracy of osteoporosis for patients with spinal deformity. When determining the bone density excellence of each vertebral node in the patient's spine, by analyzing the bone density degree and bone disorder degree of the vertebral node, the bone density excellence is determined, improving the scientificity, accuracy and objectivity of the bone density excellence. When determining the abnormality degree of each gap distance in the patient's gap distance sequence, by analyzing the offset distance of the corresponding position of the gap distance in the trend item curve, the abnormality degree is determined, accurately describing the influence of the offset of the gap distance on the gap abnormality. When determining the influence degree of each gap distance in the patient's gap distance sequence on osteoporosis, by analyzing the change of the curvature angle value, the edge deformation degree and the abnormality degree, the influence degree is determined, improving the accuracy, scientificity and objectivity of the influence degree. When determining the osteoporosis risk degree of each vertebral node in the patient's spine, by comprehensively and comprehensively analyzing the sagittal CT image and the coronal CT image, the osteoporosis risk degree is determined, improving the scientificity, accuracy and objectivity of the osteoporosis risk degree. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0039] Figure 1 It is a flowchart of the steps of a method for screening osteoporosis in patients with spinal deformity using spiral CT images provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0040] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the intended invention purpose, the following, in combination with the drawings and preferred embodiments, details the specific implementation manners, structures, features and their effects of the method for screening osteoporosis in patients with spinal deformity using spiral CT images proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0041] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs.

[0042] The following specifically describes the specific solution of the method for osteoporosis screening of patients with spinal deformities using spiral CT images provided by the present invention in conjunction with the accompanying drawings.

[0043] Please refer to Figure 1 , which shows a flowchart of the steps of a method for osteoporosis screening of patients with spinal deformities using spiral CT images provided by an embodiment of the present invention. The method includes the following steps:

[0044] Step S001: Scan a number of patients with spinal deformities using a spiral CT machine to obtain the coronal CT images, sagittal CT images of each patient's spine, and cross-sectional CT images of each vertebral node in the spine.

[0045] It should be noted that the main purpose of this embodiment is to evaluate and screen the osteoporosis risk of patients with spinal deformities. Before starting the analysis, relevant data is first obtained.

[0046] It should be noted that osteoporosis is one of the influencing factors for spinal deformities in patients with spinal deformities. The osteoporosis screening of patients with spinal deformities is often achieved by analyzing the spiral CT images of the patients with spinal deformities. For a normal spine, there is no osteoporosis phenomenon inside the bone, and there is no abnormal curvature deformation on its surface.

[0047] Specifically, scan a number of patients with spinal deformities using a spiral CT machine to obtain the coronal CT images, sagittal CT images of each patient's spine, and cross-sectional CT images of each vertebral node in the spine.

[0048] It should be noted that scanning the patients with spinal deformities using a spiral CT machine to obtain the coronal CT images, sagittal CT images of the patients' spines, and cross-sectional CT images at each vertebral node in the spine is an existing method, which will not be elaborated in this embodiment.

[0049] So far, the coronal CT images, sagittal CT images of each patient's spine, and cross-sectional CT images of each vertebral node in the spine have been obtained.

[0050] Step S002: Segment the cross-sectional CT images to obtain the cancellous bone regions in each cross-sectional CT image; according to the change in the gray value of the pixels in the cancellous bone region, obtain the bone density of each vertebral node in each patient's spine; according to the gray value distribution of the local region in the cancellous bone region, obtain the bone disorder degree of each vertebral node in each patient's spine; according to the bone density and bone disorder degree, obtain the bone density excellence degree of each vertebral node in each patient's spine.

[0051] It should be noted that the spine is composed of connected vertebral nodes. When a patient with spinal deformity has osteoporosis, the bone mineral in the vertebrae will decrease, which will in turn cause the bone substance in the vertebrae to become sparse and disordered. Therefore, the bone density excellence of each vertebral node is evaluated according to the density characteristics of the bone mass in the cross-sectional CT images of each vertebral node.

[0052] Specifically, the cross-sectional CT images are segmented to obtain the cancellous bone regions in each cross-sectional CT image, as follows:

[0053] Take any cross-sectional CT image of a vertebral node as the target cross-sectional CT image; perform semantic segmentation on the target cross-sectional CT image to obtain the cancellous bone region in the target cross-sectional CT image.

[0054] It should be noted that the semantic segmentation network used in this embodiment is the DeepLabV3 network, which is a well-known technology. The specific structure and training method of this network will not be elaborated in this embodiment. The main object affected by osteoporosis is the bone mass in the cancellous bone region.

[0055] Furthermore, according to the change of the gray value of the pixels in the cancellous bone region, the bone density degree of each vertebral node in the spine of each patient is obtained, as follows:

[0056] ;

[0057] In the formula, represents the number of pixels in the cancellous bone region of the cross-sectional CT image of the th vertebral node in the spine of the th patient; represents the gray value of the th pixel in the cancellous bone region of the cross-sectional CT image of the th vertebral node in the spine of the th patient; represents the average gray value of all pixels in the cancellous bone region of the cross-sectional CT image of the th vertebral node in the spine of the th patient; represents taking the absolute value; represents the th vertebral node in the spine of the th patient; To prevent the denominator from being 0, the first hyperparameter is for description in this embodiment.

[0058] Specifically, according to the gray value distribution of the local region in the cancellous bone region, the bone disorder degree of each vertebral node in the spine of each patient is obtained, including the following steps:

[0059] First, cluster the pixels in the cancellous bone region to obtain several clusters in the cancellous bone region of the cross-sectional CT image of each vertebral node in the spine of each patient.

[0060] Secondly, according to the change in the cluster area, obtain the degree of bone disorder of each vertebral node in the spine of each patient.

[0061] Furthermore, cluster the pixels in the cancellous bone region to obtain several clusters in the cancellous bone region of the cross-sectional CT image of each vertebral node in the spine of each patient, specifically as follows:

[0062] For the th patient, perform K-means clustering on all the pixels in the cancellous bone region of the cross-sectional CT image of the th vertebral node in the spine. The distance metric uses the absolute difference in gray values between pixels to obtain several clusters in the cancellous bone region of the cross-sectional CT image of the th patient's th vertebral node. It should be noted that in this embodiment, the K value of K-means clustering is determined by the elbow method.

[0063] Furthermore, according to the change in the cluster area, obtain the degree of bone disorder of each vertebral node in the spine of each patient, specifically as follows:

[0064] ;

[0065] In the formula, represents the number of clusters in the cancellous bone region of the cross-sectional CT image of the th patient's th vertebral node; represents the area of the th cluster in the cancellous bone region of the cross-sectional CT image of the th vertebral node in the spine of the th patient; represents the average area of all clusters in the cancellous bone region of the cross-sectional CT image of the th patient's th vertebral node; represents the degree of bone disorder of the th patient's th vertebral node.

[0066] It should be noted that in this embodiment, the cluster area is represented by the number of pixels in the cluster;

[0067] Furthermore, according to the bone density and the degree of bone disorder, obtain the bone density excellence of each vertebral node in the spine of each patient, specifically as follows:

[0068] ;

[0069] Wherein, represents the bone density of the th vertebra node in the spine of the th patient; represents the bone disorder degree of the th vertebra node in the spine of the th patient; represents a linear normalization function for normalization processing; represents the bone density excellence degree of the th vertebra node in the spine of the th patient.

[0070] Thus, the bone density excellence degree of each vertebra node in the spine of each patient is obtained.

[0071] Step S003: According to the difference in the bone density excellence degree of adjacent vertebra nodes, obtain the degree of disruption of the bone density pattern of each vertebra node in the spine of each patient; obtain several vertebra node regions in each coronal CT image; obtain the gap distance between adjacent vertebra node regions in the coronal CT image of each patient's spine; obtain the trend item curve of the gap distance between adjacent vertebra node regions in the coronal CT image of each patient's spine; arrange the gap distances between all adjacent vertebra node regions in the coronal CT image of each patient's spine to obtain the gap distance sequence of each patient; according to the offset distance of the corresponding position of the gap distance in the trend item curve, obtain the abnormality degree of each gap distance in the gap distance sequence of each patient.

[0072] It should be noted that considering the differences in the biomechanics borne by different vertebra nodes in the spine, such as the lumbar vertebrae bearing greater biomechanics than the cervical vertebrae, and the bone density of the vertebrae being positively correlated with the biomechanics that the vertebrae can bear, the normal spinal vertebrae should show an increasing bone density from top to bottom. If osteoporosis exists at a certain vertebra node, it will disrupt the regular change of the increasing bone density of the vertebrae from top to bottom in the spine. Therefore, the degree of disruption of the bone density pattern of each vertebra node is evaluated.

[0073] It should be noted that at the same time, since the bone volume of the vertebra nodes closer to the coccyx in the spine is larger, the demand for the interstitial connective tissues such as tissue fluid and ligaments between the two vertebra nodes is higher, and the vertebra gap is larger. Therefore, the normal vertebra gap also shows a pattern of increasing gap closer to the coccyx. Osteoporosis and spinal deformity can both lead to non-compliance with the gap change pattern for the vertebra gap. Therefore, the osteoporosis confidence degree of the gap bone is determined based on the degree of vertebral deformity and the gap edge characteristics shown in the coronal and sagittal CT images, and thus the osteoporosis risk degree of each vertebra node is calculated.

[0074] Specifically, according to the difference in the bone density excellence degree of adjacent vertebral nodes, the degree of disruption of the bone density pattern of each vertebral node in the spine of each patient is obtained as follows:

[0075] ;

[0076] In the formula, represents the bone density excellence degree of the th vertebral node in the spine of the th patient; represents the average value of the bone density excellence degrees of the adjacent vertebral nodes to the th vertebral node in the spine of the th patient; represents taking the absolute value; represents a linear normalization function for normalization processing; represents the degree of disruption of the bone density pattern of the th vertebral node in the spine of the th patient.

[0077] It should be noted that if the difference between the bone density excellence degree of a vertebral node in the spine and the average value of the bone density excellence degrees of the adjacent vertebral nodes is greater, it indicates that the bone density of this vertebral node less meets the changing rule of increasing density from top to bottom of the vertebral nodes, and the degree of disruption of the bone density pattern is greater.

[0078] Specifically, several vertebral node regions are obtained from each coronal CT image as follows:

[0079] Any coronal CT image is denoted as the target coronal CT image; semantic segmentation is performed on the target coronal CT image to obtain all vertebral node regions in the target coronal CT image.

[0080] Furthermore, several vertebral node regions are obtained from each sagittal CT image. It should be noted that the specific obtaining method is the same as the method for obtaining several vertebral node regions from each coronal CT image, and will not be elaborated here.

[0081] It should be noted that the gap value of a normal spine should show that the value is larger closer to the coccyx. Osteoporosis will change the edge shape of the vertebral nodes, causing abnormal changes in the vertebral node gaps. Therefore, the degree of violation of the increasing gap of the vertebral node gaps is analyzed.

[0082] Specifically, the gap distances between adjacent vertebral node regions in the coronal CT images of the spines of each patient are obtained. It should be noted that obtaining the gap distance between two vertebral node regions is an existing method, and will not be elaborated in this embodiment.

[0083] Further, obtain the trend item curve of the gap distances in the adjacent vertebral node regions in the coronal CT images of each patient's spine, specifically as follows:

[0084] For the gap distances in all adjacent vertebral node regions in the coronal CT image of the th patient's spine, construct a two-dimensional rectangular coordinate system with the sequential values of the gap distances from top to bottom in the spine as the abscissa and the gap distances as the ordinate; according to the first sequential value and the corresponding gap distance, obtain the point corresponding to the gap distance in the two-dimensional rectangular coordinate system; obtain all the points corresponding to the gap distances in the two-dimensional rectangular coordinate system, and obtain the scatter plot of the gap distances in the adjacent vertebral node regions in the coronal CT image of the th patient's spine; perform STL trend decomposition on the scatter plot to obtain the trend item curve of the gap distances in the adjacent vertebral node regions in the coronal CT image of the th patient's spine, and several standard gap distances are included in the trend item curve.

[0085] It should be noted that performing STL trend decomposition on the scatter plot to obtain the trend item curve of the gap distances in the adjacent vertebral node regions in the coronal CT image of the th patient's spine is an existing method of STL trend decomposition, which will not be elaborated in this embodiment.

[0086] Further, arrange the gap distances in all adjacent vertebral node regions in the coronal CT image of each patient's spine to obtain the gap distance sequence of each patient, specifically as follows:

[0087] Arrange the gap distances in all adjacent vertebral node regions in the coronal CT image of the th patient's spine in the order from top to bottom in the spine, and the obtained sequence is denoted as the gap distance sequence of the th patient.

[0088] Further, according to the offset distance of the gap distance at the corresponding position in the trend item curve, obtain the abnormality degree of each gap distance in the gap distance sequence of each patient, specifically as follows:

[0089] Take the absolute difference between the th gap distance in the gap distance sequence of the th patient and the standard gap distance at the corresponding position in the trend item curve as the offset distance of the th gap distance.

[0090] ;

[0091] In the formula, represents the th gap distance in the gap distance sequence of the The offset distance of a gap distance; Denote the Mean value of the offset distances of all gap distances in the gap distance sequence of the th patient; Denote taking the absolute value; Denote the linear normalization function for normalization processing; Denote the abnormality degree of the

[0092] th gap distance in the gap distance sequence of the th patient. It should be noted that the larger the offset distance of the th gap distance in the gap distance sequence of the th patient compared to the mean value of the offset distances of all gap distances in the gap distance sequence of the th patient, the more abnormal the

[0093] th gap distance in the gap distance sequence of the

[0094] th patient, and the greater the abnormality degree.

[0095] Thus far, the abnormality degree of each gap distance in the gap distance sequence of each patient is obtained.

[0096] Step S004: According to the area change situation of the adjacent vertebral node regions corresponding to the gap distance, obtain the edge deformation degree of the two adjacent vertebral node regions corresponding to each gap distance in the gap distance sequence of each patient; obtain the curvature included angle value corresponding to each gap distance in the gap distance sequence of each patient; according to the change situation of the curvature included angle value, the edge deformation degree and the abnormality degree, obtain the influence degree of osteoporosis on each gap distance in the gap distance sequence of each patient.

[0097] ;

[0098] In the formula, represents the mean of the convex hull areas of two adjacent vertebral node regions corresponding to the th gap distance in the gap distance sequence of the th patient, where the convex hull is obtained by a convex hull detection algorithm; represents the mean of the areas of two adjacent vertebral node regions corresponding to the th gap distance in the gap distance sequence of the th patient; represents the edge deformation degree of two adjacent vertebral node regions corresponding to the th gap distance in the gap distance sequence of the

[0099] It should be noted that the smaller the area of the corresponding vertebral node is compared to the area of the convex hull, the greater the degree of depression of the vertebral node edges corresponding to the upper and lower bounds of the gap between two adjacent vertebral node regions corresponding to the th gap distance, the greater the probability of osteoporosis, and the more obvious the edge deformation.

[0100] Specifically, the curvature angle values corresponding to each gap distance in the gap distance sequence of each patient are obtained as follows:

[0101] For the th gap distance and the th gap distance in the gap distance sequence of the th patient, which correspond to two adjacent vertebral node regions; the closed region between the two adjacent vertebral node regions corresponding to the th gap distance is used as the gap region corresponding to the th gap distance; the centroid of the gap region corresponding to the th gap distance and the two first centroids of the two adjacent vertebral node regions corresponding to the th gap distance are obtained; the centroid is connected to the two first centroids by straight lines respectively, and the minimum included angle value formed by the connection is used as the curvature angle value corresponding to the th gap distance in the gap distance sequence of the

[0102] It should be noted that the closed region is the closed region formed by connecting the four endpoints of the maximum width line segment of two adjacent vertebral node regions.

[0103] Furthermore, according to the change situation of the curvature angle value, the edge deformation degree and the abnormal degree, the influence degree of osteoporosis on each gap distance in the gap distance sequence of each patient is obtained as follows:

[0104] ;

[0105] In the formula, represents the edge deformation degree of the -th gap distance in the gap distance sequence of the -th patient corresponding to two adjacent vertebral node regions; represents the abnormality degree of the -th gap distance in the gap distance sequence of the -th patient; represents the curvature angle value corresponding to the -th gap distance in the gap distance sequence of the -th patient; represents the standard curvature angle value corresponding to the -th gap distance, and the standard curvature angle value is a preset value set by humans and is used to represent the curvature angle value of a normal person; represents taking the absolute value; represents a linear normalization function for normalization processing; represents the degree of influence of osteoporosis on the -th gap distance in the gap distance sequence of the -th patient; represents a hyperparameter to prevent the denominator from being 0. In this embodiment, is used for description.

[0106] It should be noted that the smaller , the smaller the difference in the curvature angle corresponding to the gap between a certain vertebral node and the curvature angle of the normal spine. At the same time, when the edge deformation degree of the -th gap distance in the gap distance sequence of the -th patient corresponding to two adjacent vertebral node regions is larger, the abnormality degree of the -th gap distance in the gap distance sequence of the -th patient is greater, indicating that the gap deformation is more affected by osteoporosis, and the degree of influence of osteoporosis on the -th gap distance in the gap distance sequence of the -th patient is greater.

[0107] Thus, the degree of influence of osteoporosis on each gap distance in the gap distance sequence of each patient is obtained.

[0108] Step S005: Obtain the osteoporosis risk degree of each vertebral node in the spine of each patient according to the degree of influence, the first degree of influence, and the degree of damage to the bone density law; evaluate the osteoporosis condition of each patient according to the osteoporosis risk degree.

[0109] It should be noted that the above analysis of the coronal CT images of the patient's spine obtained the degree of influence of osteoporosis on each intervertebral space distance in the intervertebral space distance sequence of each patient. To more accurately analyze the osteoporosis condition of the patient, this embodiment also analyzes in combination with the sagittal CT images.

[0110] Specifically, a number of first vertebral node regions are obtained in each sagittal CT image; the first intervertebral space distance between adjacent vertebral node regions in the sagittal CT image of each patient's spine is obtained; the first intervertebral space distances of all adjacent first vertebral node regions in the sagittal CT image of each patient's spine are arranged to obtain the first intervertebral space distance sequence of each patient; the first degree of influence of osteoporosis on each first intervertebral space distance in the first intervertebral space distance sequence of each patient is obtained.

[0111] It should be noted that the above steps analyzed the coronal CT images of the patient's spine to obtain the degree of influence of osteoporosis on each intervertebral space distance in the intervertebral space distance sequence of each patient. Here, the same analysis needs to be performed on the sagittal CT images of each patient to obtain the first degree of influence of osteoporosis on each first intervertebral space distance in the first intervertebral space distance sequence of each patient. The specific implementation method is the same, and this embodiment will not be elaborated further. It is only necessary to replace the coronal CT image with the sagittal CT image for analysis.

[0112] Furthermore, according to the degree of influence, the first degree of influence, and the degree of disruption of the bone density law, the osteoporosis risk degree of each vertebral node in the spine of each patient is obtained, as follows:

[0113] ;

[0114] represents the degree of disruption of the bone density law of the th vertebral node in the spine of the th patient; represents the average value of the degree of influence of osteoporosis on the two adjacent intervertebral space distances corresponding to the th vertebral node in the spine of the th patient; represents the average value of the first degree of influence of osteoporosis on the two adjacent first intervertebral space distances corresponding to the th vertebral node in the spine of the th patient; represents the linear normalization function for normalization processing;

[0115] It should be noted that represents from the The osteoporosis condition of vertebral nodes is obtained through comprehensive analysis of the coronal CT images and sagittal CT images of each patient. The greater it is, the higher the osteoporosis risk level. At the same time, for the th patient, the greater the degree of disruption of the bone density pattern of the th vertebral node in the spine, indicating that the th patient is more likely to have osteoporosis in the th vertebral node in the spine, and the osteoporosis risk level of the vertebral nodes in the patient's spine is relatively high.

[0116] Specifically, according to the osteoporosis risk level, the osteoporosis condition of each patient is evaluated, and the steps include:

[0117] First, according to the osteoporosis risk level, the comprehensive osteoporosis risk index of each patient is obtained.

[0118] Secondly, according to the magnitude of the comprehensive osteoporosis risk index, the osteoporosis condition of each patient is evaluated.

[0119] Furthermore, according to the osteoporosis risk level, the comprehensive osteoporosis risk index of each patient is obtained, specifically as follows:

[0120] Taking the average value of the osteoporosis risk levels of all vertebral nodes in the spine of the th patient as the comprehensive osteoporosis risk level of the th patient.

[0121] ;

[0122] In the formula, represents the comprehensive osteoporosis risk level of the th patient; represents the average value of the comprehensive osteoporosis risk levels of all patients; represents the linear normalization function for normalization processing; represents the comprehensive osteoporosis risk index of the th patient.

[0123] It should be noted that if the comprehensive osteoporosis risk level of the th patient is greater than the average value of the comprehensive osteoporosis risk levels of all patients, it indicates that the th patient is more likely to have osteoporosis, and the comprehensive osteoporosis risk index of the th patient is larger.

[0124] Furthermore, according to the magnitude of the comprehensive osteoporosis risk index, the osteoporosis condition of each patient is evaluated, specifically as follows:

[0125] If the comprehensive osteoporosis risk index of the th patient is less than , it is considered that the probability of osteoporosis of the th patient is relatively low, and fewer subsequent reexaminations can be performed; if the comprehensive osteoporosis risk index of the th patient is greater than or equal to , but less than , auxiliary analysis needs to be combined with dual-energy X-ray to ensure accurate osteoporosis assessment of the patient and achieve precise screening of osteoporosis in patients with spinal deformity; if the comprehensive osteoporosis risk index of the th patient is greater than or equal to , it is considered that the probability of osteoporosis of the th patient is relatively high, and medical staff need to focus on the osteoporosis manifestations of the patient in the later stage and formulate a treatment and nursing plan accordingly; and are the preset first threshold and second threshold respectively. In this embodiment, , is used for narration.

[0126] Through the above steps, the method for screening osteoporosis in patients with spinal deformity by using spiral CT images is completed.

[0127] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for osteoporosis screening of spinal deformity patients using spiral CT images, characterized in that, The method includes the following steps: Scan a number of patients with spinal deformities using a spiral CT scanner to obtain the coronal CT images, sagittal CT images of the spine, and cross-sectional CT images of each vertebral node in the spine for each patient; Segment the cross-sectional CT images to obtain the cancellous bone regions in each cross-sectional CT image; according to the change in the gray value of the pixels in the cancellous bone region, obtain the bone density of each vertebral node in the spine of each patient; according to the gray value distribution of the local region in the cancellous bone region, obtain the bone disorder degree of each vertebral node in the spine of each patient; according to the bone density and the bone disorder degree, obtain the bone density excellence degree of each vertebral node in the spine of each patient; According to the difference in the bone density excellence degree of adjacent vertebral nodes, obtain the degree of disruption of the bone density pattern of each vertebral node in the spine of each patient; obtain a number of vertebral node regions in each coronal CT image; obtain the gap distance between adjacent vertebral node regions in the coronal CT image of each patient's spine; obtain the trend term curve of the gap distance between adjacent vertebral node regions in the coronal CT image of each patient's spine; arrange the gap distances between all adjacent vertebral node regions in the coronal CT image of each patient's spine to obtain the gap distance sequence of each patient; according to the offset distance of the gap distance at the corresponding position in the trend term curve, obtain the abnormality degree of each gap distance in the gap distance sequence of each patient; According to the change in the area of the adjacent vertebral node regions corresponding to the gap distance, obtain the edge deformation degree of the two adjacent vertebral node regions corresponding to each gap distance in the gap distance sequence of each patient; obtain the curvature angle value corresponding to each gap distance in the gap distance sequence of each patient; according to the change in the curvature angle value, the edge deformation degree, and the abnormality degree, obtain the degree of influence of osteoporosis on each gap distance in the gap distance sequence of each patient; Obtain a number of first vertebral node regions in each sagittal CT image; obtain the first gap distance between adjacent vertebral node regions in the sagittal CT image of each patient's spine; arrange the first gap distances between all adjacent first vertebral node regions in the sagittal CT image of each patient's spine to obtain the first gap distance sequence of each patient; obtain the first degree of influence of osteoporosis on each first gap distance in the first gap distance sequence of each patient; According to the degree of influence, the first degree of influence, and the degree of disruption of the bone density pattern, obtain the osteoporosis risk degree of each vertebral node in the spine of each patient; according to the osteoporosis risk degree, evaluate the osteoporosis condition of each patient.

2. The method for osteoporosis screening of patients with spinal deformity using spiral CT images according to claim 1, characterized in that, The specific steps included in obtaining the bone density of each vertebral node in the spine of each patient according to the change in the gray value of the pixels in the cancellous bone region are as follows: ； Wherein, represents the number of pixels in the cancellous bone area of the cross-sectional CT image of the th vertebral node in the spine of the th patient; represents the gray value of the th pixel in the cancellous bone area of the cross-sectional CT image of the th vertebral node in the spine of the th patient; represents the average gray value of all pixels in the cancellous bone area of the cross-sectional CT image of the th vertebral node in the spine of the th patient; represents taking the absolute value; represents the bone density of the th vertebral node in the spine of the th patient; is the first hyperparameter to prevent the denominator from being zero.

3. The method for osteoporosis screening of spinal deformity patients using spiral CT images according to claim 1, characterized in that, The specific steps included in obtaining the bone disorder degree of each vertebral node in the spine of each patient according to the gray value distribution of the local region in the cancellous bone region are as follows: For the th patient, all pixels in the cancellous bone region of the cross-sectional CT image of the th vertebral node in the spine are subjected to K-means clustering. The distance metric uses the absolute difference in gray values between pixels, obtaining several clusters in the cancellous bone region of the cross-sectional CT image of the th patient's spine at the th vertebral node; ； In the formula, represents the number of clusters in the cancellous bone region of the cross-sectional CT image of the th vertebral node in the spine of the th patient; represents the area of the th cluster in the cancellous bone region of the cross-sectional CT image of the th vertebral node in the spine of the th patient; represents the average area of all clusters in the cancellous bone region of the cross-sectional CT image of the th vertebral node in the spine of the th patient; represents the degree of bone disorder of the th vertebral node in the spine of the th patient.

4. The method for osteoporosis screening of patients with spinal deformity using spiral CT images according to claim 1, characterized in that, The specific steps included in obtaining the bone density excellence degree of each vertebral node in the spine of each patient according to the bone density and the bone disorder degree are as follows: For the th vertebra node in the spine of the th patient, the ratio of the bone density to the bone disorder degree of the th vertebra node in the spine of the th patient is normalized, and the result obtained from the normalization is used as the bone density excellence degree of the th vertebra node in the spine of the th patient.

5. The method for osteoporosis screening of patients with spinal deformity using spiral CT images according to claim 1, characterized in that, Obtaining the degree of disruption of the bone density pattern of each vertebral node in the spine of each patient according to the difference in the excellence of bone density of adjacent vertebral nodes, including the following specific steps: For the th vertebra node in the spine of the th patient, the difference value between the bone density excellence degree of the th vertebra node in the spine of the th patient and the average value of the bone density excellence degrees of the adjacent vertebra nodes to the th vertebra node in the spine of the th patient is normalized, and the result obtained from the normalization is used as the degree of bone density law disruption of the th vertebra node in the spine of the th patient.

6. The method for osteoporosis screening of patients with spinal deformity using spiral CT images according to claim 1, wherein Obtaining the trend item curve of the gap distance in the adjacent vertebral node region in the coronal CT image of the spine of each patient, including the following specific steps: For the gap distances of all adjacent vertebral node regions in the coronal CT image of the spine of the th patient, a two-dimensional rectangular coordinate system is constructed with the sequential values of the gap distances from top to bottom in the spine as the abscissa and the gap distances as the ordinate; according to the first sequential value and the corresponding gap distance, the point corresponding to the gap distance in the two-dimensional rectangular coordinate system is obtained; all the points corresponding to the gap distances in the two-dimensional rectangular coordinate system are acquired to obtain the scatter plot of the gap distances of the adjacent vertebral node regions in the coronal CT image of the spine of the th patient; the STL trend decomposition is performed on the scatter plot to obtain the trend item curve of the gap distances of the adjacent vertebral node regions in the coronal CT image of the spine of the th patient, and several standard gap distances are included in the trend item curve.

7. The method for osteoporosis screening of patients with spinal deformity by using spiral CT images according to claim 6, characterized in that Obtaining the abnormality degree of each gap distance in the gap distance sequence of each patient according to the offset distance of the gap distance at the corresponding position in the trend item curve, including the following specific steps: Take the absolute difference between the -th gap distance in the gap distance sequence of the -th patient and the standard gap distance at the corresponding position in the trend item curve as the offset distance of the -th gap distance; For the th patient, the absolute difference between the offset distance of the th gap distance in the gap distance sequence and the mean of the offset distances of all gap distances in the gap distance sequence of the th patient is normalized, and the result obtained from the normalization is taken as the abnormality degree of the th gap distance in the gap distance sequence of the th patient.

8. The method for osteoporosis screening of patients with spinal deformity using spiral CT images according to claim 1, characterized in that Obtaining the edge deformation degree of two adjacent vertebral node regions corresponding to each gap distance in the gap distance sequence of each patient according to the area change of the gap distance corresponding to the adjacent vertebral node region, including the following specific steps: Take the ratio of the average convex hull area of the two adjacent vertebral node regions corresponding to the -th gap distance in the gap distance sequence of the -th patient to the average area of the two adjacent vertebral node regions corresponding to the -th gap distance in the gap distance sequence of the -th patient, and use it as the edge deformation degree of the two adjacent vertebral node regions corresponding to the -th gap distance in the gap distance sequence of the -th patient.

9. The method for osteoporosis screening of patients with spinal deformity using spiral CT images according to claim 1, characterized in that, Obtaining the degree of influence of osteoporosis on each gap distance in the gap distance sequence of each patient according to the change of the curvature angle value, the edge deformation degree and the abnormality degree, including the following specific steps: ； Wherein, represents the edge deformation degree of the -th gap distance in the gap distance sequence of the -th patient corresponding to two adjacent vertebral node regions; represents the abnormality degree of the -th gap distance in the gap distance sequence of the -th patient; represents the curvature angle value corresponding to the -th gap distance in the gap distance sequence of the -th patient; represents the standard curvature angle value corresponding to the -th gap distance; represents taking the absolute value; represents the linear normalization function; represents the influence degree of osteoporosis on the -th gap distance in the gap distance sequence of the -th patient; represents a hyperparameter to prevent the denominator from being zero.

10. The method for osteoporosis screening of patients with spinal deformity using spiral CT images according to claim 1, characterized in that, Obtaining the osteoporosis risk degree of each vertebral node in the spine of each patient according to the degree of influence, the first degree of influence and the degree of disruption of the bone density pattern, including the following specific steps: ; Indicates the degree of disruption of the bone density pattern of the th vertebral node in the spine of the th patient; Indicates the average degree of influence of osteoporosis on the distance between two adjacent gaps corresponding to the th vertebral node in the spine of the th patient; Indicates the average degree of the first influence of osteoporosis on the distance between two adjacent first gaps corresponding to the th vertebral node in the spine of the th patient; Indicates a linear normalization function; Indicates the degree of osteoporosis risk of the th vertebral node in the spine of the th patient.

Images