A spinal conservative treatment aided decision method and system based on image analysis

Abstract

The present application relates to the technical field of spinal X-ray image segmentation, and particularly relates to a kind of spinal conservative treatment auxiliary decision-making method and system based on image analysis.Based on spinal X-ray image, the pixel to be analyzed is obtained by threshold segmentation, and each row of suspected spinal pixel is obtained by position and gray distribution screening.According to the difference in the number of pixels between adjacent rows, the distance between rows and the distribution characteristics, the similarity between rows is calculated, and accordingly the suspected spinal joint area is divided.The confidence of each region is evaluated by fusing the shape characteristics of the region and the number of regions, and the final spinal joint region is selected by screening.By analyzing the position offset angle, distance and offset region number of adjacent joint regions, the degree of scoliosis is comprehensively evaluated to provide decision support for conservative treatment.The present application can accurately identify the scoliosis region, so that relevant personnel can develop the most appropriate treatment auxiliary decision-making scheme for patients.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of spinal X-ray image segmentation, and in particular to a spinal conservative treatment auxiliary decision-making method and system based on image analysis. BACKGROUND

[0002] With the deep integration of medical image analysis technology and artificial intelligence, traditional spinal conservative treatment relies on doctors to evaluate X-ray, MRI and other images by naked eye, which has strong subjectivity, low measurement efficiency and difficulty in precise quantification. This leads to a lack of objective and unified standards for disease assessment, program development and efficacy tracking, affecting the accuracy and consistency of treatment. The auxiliary decision-making method based on image analysis emerges as the times require, which automatically extracts spinal geometric parameters, biomechanical features and pathological changes, providing data-driven objective basis for conservative treatment (such as rehabilitation training and brace correction), significantly improving the accuracy of diagnostic evaluation, the individualization level of treatment strategy and the continuity of efficacy management, and thus promoting the spinal disease conservative treatment to a new stage of standardization and intelligentization.

[0003] When X-ray film is taken for the patient's spine, due to different standing postures, individual differences and other problems of different patients, the X-ray film image taken has interference information, which affects the recognition of the patient's spinal feature part, and thus it is difficult for relevant personnel to develop the most appropriate treatment auxiliary decision-making scheme according to the patient's specific situation, ultimately affecting the patient's treatment effect. SUMMARY

[0004] To address the technical problem that interference information in X-ray images of a patient's spine can occur due to variations in standing posture and individual differences, affecting the identification of spinal features and hindering the development of the most suitable treatment decision-making plan, ultimately impacting treatment outcomes, this invention provides an image analysis-based method and system for assisting decision-making in conservative spinal treatment. The specific technical solution is as follows: An image analysis-based method for assisting decision-making in conservative spinal treatment includes: acquiring a patient's spinal X-ray image; performing threshold segmentation on the spinal X-ray image to obtain all pixels to be analyzed; filtering all pixels to be analyzed based on their positional and grayscale distribution to obtain suspected spinal pixels in each row; and determining the relative positions of adjacent suspected spinal pixels. The probability of similarity between adjacent rows of suspected spinal pixels is determined by the differences in the number of pixels, the distance between adjacent rows of suspected spinal pixels, and the pixel distribution characteristics of each row of suspected spinal pixels. Based on this probability, all suspected spinal pixels are divided into regions to obtain all suspected spinal joint regions. The confidence level of each suspected spinal joint region is obtained based on its shape characteristics and the number of suspected spinal joint regions. All suspected spinal joint regions are then filtered based on this confidence level to obtain all spinal joint regions. The degree of scoliosis is determined based on the positional offset angle and distance between adjacent spinal joint regions, as well as the number of spinal joint regions with positional offset. The degree of scoliosis is then used to assist in making decisions regarding conservative spinal treatment for the patient.

[0005] Furthermore, the method for obtaining each row of suspected spinal pixels includes: taking the pixel with a gray value of 1 closest to the center of each row of the spinal X-ray image as the starting point, obtaining all the consecutive pixels with a gray value of 1 on the left and right sides of the pixel, and using them together as each row of suspected spinal pixels.

[0006] Furthermore, the method for obtaining the probability of the same property includes: obtaining the probability of the same property according to the formula for calculating the probability of the same property, which is shown below: In the formula, Indicates the first Line suspected spine pixels and the first The degree to which pixels suspected of having the same properties are compared with those of the spine. Indicates the first The number of pixels in the suspected spine pixel array; Indicates the first The number of pixels in the suspected spine pixel array; Indicates the first Line suspected spine pixels and the first The minimum number of pixels within a row of suspected spinal pixels; Indicates the first Line suspected spine pixels and the first The distance between pixels at the center of the suspected spine pixel; Represents the absolute value function; This represents the normalization function.

[0007] Furthermore, the method for obtaining the suspected spinal joint region includes: dividing two adjacent rows of suspected spinal pixels with a probability of having the same property greater than a preset first threshold into the same region, and dividing the next row of suspected spinal pixels in two adjacent rows of suspected spinal pixels with a probability of having the same property less than the preset first threshold into other regions; traversing the probability of having the same property in each pair of adjacent suspected spinal pixels, dividing each row of suspected spinal pixels into regions, and obtaining all suspected spinal joint regions.

[0008] Furthermore, the method for obtaining the confidence level includes: obtaining the confidence level according to a confidence level calculation formula, the confidence level calculation formula being as follows: In the formula, Indicates the first Confidence level of a suspected spinal joint region; Indicates the first The width of a suspected spinal joint area; This represents the average width of all suspected spinal joint areas; Indicates the relationship with the first The number of other suspected spinal joint regions with the same area as the suspected spinal joint region; Indicates the number of suspected spinal joint areas; Represents the absolute value function; This represents the normalization function.

[0009] Furthermore, the method for obtaining the spinal joint region includes: taking each suspected spinal joint region with a confidence level greater than a preset second threshold as each spinal joint region.

[0010] Further, the method for obtaining the degree of scoliosis includes: establishing a Cartesian coordinate system with the lower left corner of the spine X-ray image as the origin, obtaining the coordinate positions of the upper left corner, the lower left corner, the upper right corner and the lower right corner in each spine joint region; calculating the included angle between the line connecting the upper left corner coordinate points of two adjacent spine joint regions and the vertical direction as a first angle; calculating the included angle between the line connecting the lower left corner coordinate points of two adjacent spine joint regions and the vertical direction as a second angle; calculating the included angle between the line connecting the upper right corner coordinate points of two adjacent spine joint regions and the vertical direction as a third angle; calculating the included angle between the line connecting the lower right corner coordinate points of two adjacent spine joint regions and the vertical direction as a fourth angle; taking the mean value between the first angle, the second angle, the third angle and the fourth angle as the position offset angle feature between two adjacent spine joint regions; calculating the mean value of the distance between each vertex of two adjacent spine joint regions as the position offset distance feature between two adjacent spine joint regions; taking the product between the position offset angle feature and the position offset distance feature as the offset performance degree between two adjacent spine joint regions; when the offset performance degree between two adjacent spine joint regions is greater than a preset third threshold value, it is considered that the two adjacent spine joint regions have occurred offset, the offset performance degree between each adjacent two spine joint regions is traversed to obtain the number of spine joints that have occurred offset; the mean value of the position offset angle feature of each adjacent two spine joint regions is calculated as a first mean value, and the maximum value of the position offset angle feature of each adjacent two spine joint regions is calculated as a first numerical value; the product between the number of spine joints that have occurred offset, the first mean value and the first numerical value is normalized to obtain the degree of scoliosis of the patient.

[0011] An image analysis-based spinal conservative treatment auxiliary decision system, the system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, the processor implementing the steps of the image analysis-based spinal conservative treatment auxiliary decision method as described above when executing the computer program.

[0012] A computer readable storage medium storing a computer program, the computer program implementing the steps of the image analysis-based spinal conservative treatment auxiliary decision method as described above when executed by a processor.

[0013] A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, the processor implementing the steps of the image analysis-based spinal conservative treatment auxiliary decision method as described above when executing the computer program.

[0014] The present application has the following advantages: the present application collects the X-ray image of the spine of a patient; since the human spine is composed of a plurality of continuous joints, there is a certain gray difference between the spine joints and the surrounding area in the X-ray image, therefore the threshold segmentation is performed on the X-ray image of the spine to obtain all the pixel points to be analyzed; for the frontal X-ray film, the spine is generally located in the middle of the image, therefore the pixel points of each row of the suspected spine region can be described by the position distribution and the gray distribution of the pixel points to be analyzed; since the spine part is connected by a plurality of vertebrae, the likelihood of the same nature between the adjacent two rows of suspected spine pixel points is obtained according to the number difference between the adjacent two rows of suspected spine pixel points, the inter-row distance feature of the adjacent two rows of suspected spine pixel points and the pixel point distribution feature of each row of suspected spine pixel points; since there are other regions between the vertebrae that affect the judgment of the spine region, the region division is first performed on each row of suspected spine pixel points to obtain the suspected spine joint region, and further judgment is performed on different suspected spine joint regions; since the shape of each vertebra in the spine part is similar, the confidence degree of each suspected spine joint region is obtained according to the region shape feature of each suspected spine joint region and the number of suspected spine joint regions, and then all the spine joint regions are screened out; when the patient has a spine deviation problem, the patient has a deviation phenomenon between part of the spine joints and the upper and lower connected spine parts reflected on the X-ray film, therefore the degree of scoliosis of the patient is obtained according to the position deviation angle feature, the position deviation distance feature between each adjacent two spine joint regions and the number of spine joint regions having the position deviation, and the conservative treatment of the spine of the patient is assisted in decision-making. The present application can accurately identify the scoliosis region, so that relevant personnel can develop the most suitable treatment and auxiliary decision-making scheme for the patient. BRIEF DESCRIPTION OF DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, and the advantages thereof, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained from these drawings without creative labor.

[0016] Figure 1 A flow chart of a conservative treatment auxiliary decision-making method for the spine based on image analysis provided by an embodiment of the present application; Figure 2 A block diagram of a conservative treatment auxiliary decision-making system for the spine based on image analysis provided by an embodiment of the present application. DETAILED DESCRIPTION

[0017] For further illustrating the technical means and effects taken by the present application to achieve the predetermined inventive purpose, the following describes in detail the specific implementation, structure, features and effects of a kind of image analysis-based spinal conservative treatment auxiliary decision-making method and system according to the present application, combined with the preferred embodiments and the drawings. 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.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs.

[0019] The specific scheme of the image analysis-based spinal conservative treatment auxiliary decision-making method and system provided by the present application is described in detail below in combination with the drawings.

[0020] Please refer to Figure 1 which shows an image analysis-based spinal conservative treatment auxiliary decision-making method provided by an embodiment of the present application. The method comprises the following steps: step S1: acquiring a spinal X-ray image of a patient.

[0021] The present embodiment is mainly applied to the scenario of identifying the scoliosis region of a patient, so in the present embodiment, the spinal part of the patient is photographed by using X-ray technology, and therefore the spinal X-ray image of the patient is first acquired.

[0022] In an embodiment of the present application, a standard standing position full-spinal anteroposterior X-ray film of the patient is acquired as the spinal X-ray image required for subsequent operations. It should be noted that the image acquisition method is a technical means known to those skilled in the art, which will not be described here.

[0023] Step S2: threshold segmentation is performed on the spine X-ray image to obtain all pixel points to be analyzed; according to the position distribution and the gray scale distribution of all pixel points to be analyzed, the pixel points to be analyzed are screened to obtain suspected spine pixel points in each row; according to the number difference between adjacent two rows of suspected spine pixel points, the distance feature between adjacent two rows of suspected spine pixel points, and the pixel point distribution feature of each row of suspected spine pixel points, the same property possibility between adjacent two rows of suspected spine pixel points is obtained; according to the same property possibility between each adjacent two rows of suspected spine pixel points, the suspected spine joint regions are divided to obtain all suspected spine joint regions; according to the region shape feature of each suspected spine joint region and the number of suspected spine joint regions, the confidence degree of each suspected spine joint region is obtained; according to the confidence degree, all suspected spine joint regions are screened to obtain all spine joint regions; and according to the position offset angle feature, the position offset distance feature between each adjacent two spine joint regions, and the number of spine joint regions with position offset, the degree of scoliosis of the patient is obtained.

[0024] Since the human spine is composed of a plurality of continuous joints, there is a certain gray scale difference between the spine joint and the surrounding area in the X-ray image, therefore, in the embodiment of the present application, threshold segmentation is performed on the spine X-ray image to obtain all pixel points to be analyzed. It should be noted that threshold segmentation is a technical means known to those skilled in the art, which will not be described here.

[0025] For the frontal X-ray film, the spine is generally located in the middle of the image, so the pixel points in each row of the suspected spine region can be described by the position distribution and the gray scale distribution of the pixel points to be analyzed.

[0026] Preferably, in an embodiment of the present application, the method for obtaining each row of suspected spine pixel points comprises: after threshold segmentation, the vertebrae part is taken as the foreground, and the gray scale value of the pixel points in the region is 1. However, in the spine X-ray image, there are also pixel points with a gray scale value of 1 in other regions which need to be further screened out subsequently. Therefore, the pixel point with a gray scale value of 1 closest to the center of the image in each row of the spine X-ray image is taken as the starting point, all pixel points with a gray scale value of 1 on the left and right sides of the starting point are obtained, and the obtained pixel points are collectively taken as each row of suspected spine pixel points.

[0027] Since the spine part is connected by a large number of vertebrae, in the spine X-ray image, a large number of regions with a gray scale value of 1 are connected from top to bottom, therefore, according to the number difference between adjacent two rows of suspected spine pixel points, the distance feature between adjacent two rows of suspected spine pixel points, and the pixel point distribution feature of each row of suspected spine pixel points, the same property possibility between adjacent two rows of suspected spine pixel points is obtained.

[0028] Preferably, in one embodiment of the present invention, the method for obtaining the probability of having the same property includes: obtaining the probability of having the same property according to a formula for calculating the probability of having the same property, the formula for calculating the probability of having the same property being as follows: In the formula, Indicates the first Line suspected spine pixels and the first The degree to which pixels suspected of having the same properties are compared with those of the spine. Indicates the first The number of pixels in the suspected spine pixel array; Indicates the first The number of pixels in the suspected spine pixel array; Indicates the first Line suspected spine pixels and the first The minimum number of pixels within a row of suspected spinal pixels; Indicates the first Line suspected spine pixels and the first The distance between pixels at the center of the suspected spine pixel; Represents the absolute value function; This represents the normalization function.

[0029] In the formula for calculating the probability of the same property, since the spine region is roughly rectangular, the number of pixels in each row within the spine region varies little, and the row spacing between pixels is also very small; in the formula... The smaller the value, the better. Line suspected spine pixels and the first The smaller the difference in the number of suspected spinal pixels, the better. Line suspected spine pixels and the first The more likely a row of pixels is to be in the same spinal region, the more likely it is to be in the same spinal region; since the number of pixels per row in the skull region is greater than the number of pixels per row in the spinal region, the first row... Line suspected spine pixels and the first Within a row of suspected spine pixels, the smaller the minimum number of pixels, and the greater the row spacing between two rows of suspected spine pixels, the better. The smaller the value, the more likely both rows of suspected spine pixels are to be located in the spine region, i.e., the first... Line suspected spine pixels and the first The greater the degree of similarity in properties between suspected spinal pixels, the more likely they are to share the same characteristics.

[0030] Since other regions may exist between the vertebrae and affect the judgment of the spinal region, we first divide each row of suspected spinal pixel points into regions to obtain suspected spinal joint regions, and then make further judgments on different suspected spinal joint regions.

[0031] Preferably, in one embodiment of the present application, the method for obtaining the suspected spinal joint region comprises: dividing the two adjacent suspected spinal pixel rows with the same property possibility greater than a preset first threshold into the same region, and dividing the last suspected spinal pixel row of the two adjacent suspected spinal pixel rows with the same property possibility less than the preset first threshold into other regions; traversing the same property possibility of each two adjacent suspected spinal pixel rows, and performing region division on each suspected spinal pixel row to obtain all suspected spinal joint regions.

[0032] In one embodiment of the present application, the preset first threshold is set to 0.8. It should be noted that the preset first threshold can be set by itself, which is not limited herein.

[0033] Since the shape of each vertebra is similar in the spinal part, in the embodiment of the present application, the confidence degree of each suspected spinal joint region is obtained according to the region shape feature of each suspected spinal joint region and the number of suspected spinal joint regions.

[0034] Preferably, in one embodiment of the present application, the method for obtaining the confidence degree comprises: obtaining the confidence degree according to a confidence degree calculation formula, and the confidence degree calculation formula is as follows: In the formula, C represents the confidence degree of the i-th suspected spinal joint region; W represents the width of the i-th suspected spinal joint region; W represents the average width of all suspected spinal joint regions; N represents the number of suspected spinal joint regions; N represents the number of other suspected spinal joint regions with the same area as the i-th suspected spinal joint region; and represents the absolute value function.

[0035] In the confidence degree calculation formula, the smaller the difference between the width of the i-th suspected spinal joint region and the average width of all suspected spinal joint regions, the smaller the width difference between the i-th suspected spinal joint region and the overall width, and the more likely the i-th suspected spinal joint region is the spinal region, that is, the greater the confidence degree of the i-th suspected spinal joint region; and the more the number of other suspected spinal joint regions with the same area as the i-th suspected spinal joint region accounts for the total number of suspected spinal joint regions, the more likely the i-th suspected spinal joint region is the spinal region, that is, the greater the confidence degree of the i-th suspected spinal joint region. ​​​​​​​​​​​​​​​​​The greater the confidence degree of each suspected spinal joint region is, the greater the confidence degree of the spinal joint region is.

[0036] Preferably, in an embodiment of the present application, the method for obtaining the spinal joint region comprises: taking each suspected spinal joint region with the confidence degree greater than the preset second threshold as each spinal joint region.

[0037] In an embodiment of the present application, the preset second threshold is set to 0.85. It should be noted that the preset second threshold can be set by itself and is not limited herein.

[0038] When the patient has a spinal deviation problem, it is reflected on the X-ray image that there is a deviation phenomenon between the part of the spinal joint and the upper and lower connected part of the spine. Therefore, in the embodiment of the present application, the degree of scoliosis of the patient is obtained according to the position deviation angle feature between each two adjacent spinal joint regions, the position deviation distance feature, and the number of spinal joint regions with position deviation.

[0039] Preferably, in an embodiment of the present application, the method for obtaining the degree of scoliosis comprises: establishing a Cartesian coordinate system with the lower left corner of the spinal X-ray image as the origin, obtaining the coordinate positions of the upper left corner, the lower left corner, the upper right corner and the lower right corner in each spinal joint region; calculating the included angle between the line connecting the upper left corner coordinate points of the two adjacent spinal joint regions and the vertical direction as a first angle; calculating the included angle between the line connecting the lower left corner coordinate points of the two adjacent spinal joint regions and the vertical direction as a second angle; calculating the included angle between the line connecting the upper right corner coordinate points of the two adjacent spinal joint regions and the vertical direction as a third angle; and calculating the included angle between the line connecting the lower right corner coordinate points of the two adjacent spinal joint regions and the vertical direction as a fourth angle.

[0040] The mean value between the first angle, the second angle, the third angle and the fourth angle is taken as the position deviation angle feature between the two adjacent spinal joint regions. The greater the mean value between the first angle, the second angle, the third angle and the fourth angle is, the greater the angle deviation between the two adjacent spinal joint regions is, and the greater the position deviation angle feature is. The mean value of the line distance between each vertex of the two adjacent spinal joint regions is taken as the position deviation distance feature between the two adjacent spinal joint regions. The greater the mean value of the line distance is, the greater the distance deviation between the two adjacent spinal joint regions is, and the greater the position deviation distance feature is. The product between the position deviation angle feature and the position deviation distance feature is taken as the deviation performance degree between the two adjacent spinal joint regions. The greater the position deviation angle feature is and the greater the position deviation distance feature is, the greater the deviation performance degree between the two adjacent spinal joint regions is.

[0041] When the offset manifestation degree between two adjacent spinal joint regions is greater than a preset third threshold, it is considered that the two adjacent spinal joint regions are offset, the offset manifestation degree between each two adjacent spinal joint regions is traversed, and the number of spinal joints that are offset is obtained; in an embodiment of the present application, the preset third threshold is set to 0, and it needs to be noted that the preset third threshold can be set by itself and is not limited herein.

[0042] The average of the position offset angle features of each two adjacent spinal joint regions is calculated as a first average, and the maximum of the position offset angle features of each two adjacent spinal joint regions is calculated as a first value; the product between the number of spinal joints that are offset, the first average and the first value is normalized to obtain the degree of scoliosis of the patient. An embodiment of the present application provides a formula for calculating the degree of scoliosis, and the specific formula is as follows: In the formula, represents the degree of scoliosis of the patient; represents the number of spinal joints that are offset; represents the average of the position offset angle features of each two adjacent spinal joint regions; represents the maximum of the position offset angle features of each two adjacent spinal joint regions.

[0043] In the formula for calculating the degree of scoliosis, when there are more spinal regions that are offset in the patient, and the offset angle of the spinal region is also large, it indicates that the current spinal offset problem of the patient is large, that is, the degree of scoliosis of the patient is larger.

[0044] Step S3: assisting in decision-making for the conservative treatment of the patient's spine according to the degree of scoliosis.

[0045] In the above process, the analysis of the degree of scoliosis of the patient's spine is completed, and further based on the analysis of the degree of scoliosis of the patient's spine to help give the decision scheme of the conservative treatment of the patient's spine. However, the decision scheme for treatment of patients of different ages and different physical qualities should be different. Therefore, the personal data of the patient needs to be combined to help develop a conservative treatment decision scheme, and the specific steps are as follows: in an embodiment of the present application, a multi-modal fusion decision network is selected to construct an intelligent decision model for the conservative treatment of scoliosis based on multi-modal feature fusion and reinforcement learning, and it needs to be noted that the model construction method is a technical means known to those skilled in the art and is not described herein. The personal data of the patient (including age, various physical quality indicators) and the data of the degree of scoliosis of the patient obtained by the above analysis are input into the trained model to generate a conservative treatment decision-making method for the spine.

[0046] In summary, a spine X-ray image of a patient is collected; threshold segmentation is performed on the spine X-ray image to obtain all pixel points to be analyzed; all pixel points to be analyzed are screened according to the position distribution and gray scale distribution of the pixel points to obtain suspected spine pixel points in each row; the likelihood of the same nature between adjacent two rows of suspected spine pixel points is obtained according to the number difference between the adjacent two rows of suspected spine pixel points, the distance feature between the adjacent two rows of suspected spine pixel points, and the pixel point distribution feature of each row of suspected spine pixel points; all suspected spine joint regions are obtained by region division of all suspected spine pixel points according to the likelihood of the same nature between each adjacent two rows of suspected spine pixel points; the confidence degree of each suspected spine joint region is obtained according to the region shape feature of each suspected spine joint region and the number of suspected spine joint regions; all spine joint regions are screened according to the confidence degree to obtain all spine joint regions; the degree of scoliosis of the patient is obtained according to the position offset angle feature, the position offset distance feature between each adjacent two spine joint regions, and the number of spine joint regions with position offset; and the conservative treatment of the spine of the patient is assisted in decision-making according to the degree of scoliosis.

[0047] An embodiment of the present application provides a spine conservative treatment assisted decision-making system based on image analysis, which comprises a memory, a processor and a computer program, wherein the memory is used for storing the corresponding computer program, the processor is used for running the corresponding computer program, and the computer program can realize the method described in steps S1-S3 when running in the processor, and specifically comprises: an image acquisition module 101 used for collecting a spine X-ray image of a patient; an image analysis module 102 used for performing threshold segmentation on the spine X-ray image to obtain all pixel points to be analyzed; screening all pixel points to be analyzed according to the position distribution and gray scale distribution of the pixel points to obtain suspected spine pixel points in each row; obtaining the likelihood of the same nature between adjacent two rows of suspected spine pixel points according to the number difference between the adjacent two rows of suspected spine pixel points, the distance feature between the adjacent two rows of suspected spine pixel points, and the pixel point distribution feature of each row of suspected spine pixel points; obtaining all suspected spine joint regions by region division of all suspected spine pixel points according to the likelihood of the same nature between each adjacent two rows of suspected spine pixel points; obtaining the confidence degree of each suspected spine joint region according to the region shape feature of each suspected spine joint region and the number of suspected spine joint regions; screening all suspected spine joint regions according to the confidence degree to obtain all spine joint regions; obtaining the degree of scoliosis of the patient according to the position offset angle feature, the position offset distance feature between each adjacent two spine joint regions, and the number of spine joint regions with position offset; and an assisted decision-making module 103 used for assisting in decision-making of the conservative treatment of the spine of the patient according to the degree of scoliosis.

[0048] A third object of the embodiments of the present application is to provide a computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, the processor implementing the method described in steps S1-S3 when executing the computer program.

[0049] A fourth object of the embodiments of the present application is to provide a computer readable storage medium storing a computer program, the computer program implementing the method described in steps S1-S3 when executed by a processor.

[0050] It should be noted that the above-mentioned order of the embodiments of the present application is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are also possible or can be advantageous.

[0051] Each of the embodiments in the specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other. Each embodiment mainly describes the difference from other embodiments.

Claims

1. An image analysis based decision aid method for conservative treatment of the spine, characterized in that, The method includes: acquiring X-ray images of the patient's spine; performing threshold segmentation on the X-ray images to obtain all pixels to be analyzed; filtering all pixels to be analyzed based on their positional and grayscale distributions to obtain suspected spinal pixels in each row; obtaining the probability of similarity between suspected spinal pixels in adjacent rows based on the difference in the number of suspected spinal pixels in adjacent rows, the distance between adjacent rows of suspected spinal pixels, and the pixel distribution characteristics of each row of suspected spinal pixels; dividing all suspected spinal pixels into regions based on the probability of similarity between adjacent rows of suspected spinal pixels to obtain all suspected spinal joint regions; and obtaining the confidence level of each suspected spinal joint region based on the region shape characteristics and the number of suspected spinal joint regions. All suspected spinal joint regions are screened based on the confidence level to obtain all spinal joint regions; the degree of scoliosis of the patient is obtained based on the positional offset angle feature, positional offset distance feature, and the number of spinal joint regions with positional offset between each pair of adjacent spinal joint regions; the degree of scoliosis is used to assist in the decision-making of conservative treatment of the patient's spine; the method for obtaining each row of suspected spinal pixels includes: taking the pixel with a gray value of 1 closest to the center of each row of the spinal X-ray image as the starting point, obtaining all pixels with a gray value of 1 on both sides of the pixel, and using them together as the suspected spinal pixels of each row; the method for obtaining the probability of the same property includes: obtaining the probability of the same property according to the formula for calculating the probability of the same property, which is as follows: In the formula, Indicates the first Line suspected spine pixels and the first The degree to which pixels suspected of having the same properties are compared with those of the spine. Indicates the first The number of pixels in the suspected spine pixel array; Indicates the first The number of pixels in the suspected spine pixel array; Indicates the first Line suspected spine pixels and the first The minimum number of pixels within a row of suspected spinal pixels; Indicates the first Line suspected spine pixels and the first The distance between pixels at the center of the suspected spine pixel; Represents the absolute value function; The normalization function is represented here. The method for obtaining the suspected spinal joint region includes: dividing adjacent rows of suspected spinal pixels with a probability of similarity greater than a preset first threshold into the same region; dividing the next row of suspected spinal pixels among adjacent rows of suspected spinal pixels with a probability of similarity less than the preset first threshold into other regions; traversing the probability of similarity between each pair of adjacent suspected spinal pixels, dividing each row of suspected spinal pixels into regions, and obtaining all suspected spinal joint regions; the method for obtaining the confidence level includes: obtaining the confidence level according to the confidence level calculation formula, which is shown below: In the formula, Indicates the first Confidence level of a suspected spinal joint region; Indicates the first The width of a suspected spinal joint area; This represents the average width of all suspected spinal joint areas; Indicates the relationship with the first The number of other suspected spinal joint regions with the same area as the suspected spinal joint region; Indicates the number of suspected spinal joint areas; Represents the absolute value function; The normalized function is represented; the acquisition method of the spinal joint region comprises: taking each suspected spinal joint region with a confidence degree greater than a preset second threshold value as each spinal joint region; the acquisition method of the spinal scoliosis degree comprises: establishing a Cartesian coordinate system with the lower left corner of the spinal X-ray image as the origin, obtaining the coordinate positions of the upper left corner, the lower left corner, the upper right corner and the lower right corner in each spinal joint region; calculating the included angle between the line connecting the upper left corner coordinate points of two adjacent spinal joint regions and the vertical direction as a first angle; calculating the included angle between the line connecting the lower left corner coordinate points of two adjacent spinal joint regions and the vertical direction as a second angle; calculating the included angle between the line connecting the upper right corner coordinate points of two adjacent spinal joint regions and the vertical direction as a third angle; calculating the included angle between the line connecting the lower right corner coordinate points of two adjacent spinal joint regions and the vertical direction as a fourth angle; taking the mean value between the first angle, the second angle, the third angle and the fourth angle as the position offset angle feature between two adjacent spinal joint regions; calculating the mean value of the line distance between each vertex of two adjacent spinal joint regions as the position offset distance feature between two adjacent spinal joint regions; taking the product between the position offset angle feature and the position offset distance feature as the offset performance degree between two adjacent spinal joint regions; when the offset performance degree between two adjacent spinal joint regions is greater than a preset third threshold value, it is considered that the two adjacent spinal joint regions are offset, the offset performance degree between each two adjacent spinal joint regions is traversed, and the number of spinal joints with offset phenomenon is obtained; the mean value of the position offset angle feature of each two adjacent spinal joint regions is calculated as a first mean value, and the maximum value of the position offset angle feature of each two adjacent spinal joint regions is calculated as a first value; the product between the number of spinal joints with offset phenomenon, the first mean value and the first value is normalized to obtain the spinal scoliosis degree of the patient.

2. An image analysis based conservative treatment of the spine decision support system, the system comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that, The computer program is executed by the processor to implement the steps of the image analysis based decision support method for conservative treatment of the spine according to claim 1.

3. A computer-readable storage medium storing a computer program, the computer-readable storage medium comprising instructions that, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 2. The computer program is executed by the processor to implement the steps of the image analysis based decision support method for conservative treatment of the spine according to claim 1.

4. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The computer program is executed by the processor to implement the steps of the image analysis based decision support method for conservative treatment of the spine according to claim 1.

Images