Vertebral image-based osteoporotic fracture manual reduction assistance decision system

Abstract

The present application relates to the technical field of patient vertebral body image recognition, and particularly relates to a kind of osteoporotic fracture manual reduction auxiliary decision system based on vertebral body image.Threshold segmentation is used to obtain the region to be analyzed.Based on the edge shape and pixel distribution of the region, the boundary flatness is calculated, and the attribution degree of the vertebral body is evaluated by combining the flatness difference and number of adjacent regions, so as to screen the vertebral block region.The bone fragment probability is calculated by analyzing the area and distance from the vertebral block of the non-vertebral block region, and the bone fragment region is screened and classified to the corresponding vertebral body.The osteoporosis possibility is evaluated by comprehensively considering the boundary flatness, bone fragment number and gray distribution in the vertebral body region, and finally the osteoporosis degree of the patient is quantified according to the volume and region number of the vertebral block and fragment in the osteoporosis region, to provide decision support for the manual reduction of vertebral fracture.The present application can accurately identify osteoporotic fracture on the CT image of vertebral body, so that the analysis and judgment of relevant personnel for the patient are more accurate.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of patient vertebral body image recognition, and in particular to a manual reduction auxiliary decision system for osteoporotic fractures based on vertebral body images. BACKGROUND

[0002] Globally, the population structure is developing towards aging. Osteoporosis and its most severe complication, osteoporotic fractures, have become an increasingly serious global public health problem, imposing a heavy burden on the social economy and the medical system. Osteoporotic vertebral compression fractures (OVCF) are the most common type, which has a high incidence and high morbidity, and seriously affects the quality of life of the elderly population. Traditional manual reduction methods such as traditional Chinese bone setting have their value, but they are often highly dependent on the personal experience and "hand feeling" of the physician, lack objective and quantitative standards, and thus the curative effect is difficult to guarantee and standardize. At the same time, with the popularization and development of high-resolution CT, MRI and other imaging technologies, a large amount of precise image data has laid a foundation for the digital diagnosis of fractures. This provides a historic opportunity to use artificial intelligence technology to break through the bottleneck of traditional medicine, and through deep learning algorithms to realize three-dimensional reconstruction of vertebral bodies, accurate measurement of fracture parameters and reduction simulation, which is promoting the new stage of manual reduction from experience-oriented to precise and quantitative decision-making.

[0003] In the process of analyzing the osteoporotic fracture of the patient's vertebrae, the movement artifacts caused by the patient's breathing and movement caused by pain during the scanning process seriously interfere with the identification of the osteoporotic fracture on the patient's vertebral body CT image, thereby affecting the subsequent analysis and judgment of the relevant personnel on the patient. SUMMARY

[0004] In order to solve the technical problem that the movement caused by the breathing and pain of the patient during the scanning process will produce motion artifacts, which seriously interfere with the identification of osteoporotic fractures on the CT image of the vertebral body of the patient, thereby affecting the analysis and judgment of the relevant personnel on the patient, the purpose of the present application is to provide an osteoporotic fracture manual reduction auxiliary decision system based on vertebral image, and the technical scheme adopted is as follows: An osteoporotic fracture manual reduction auxiliary decision system based on vertebral image, the system comprises: an image acquisition module for acquiring the CT image of the vertebral body of the patient; an image analysis module for threshold segmentation of the CT image of the vertebral body to obtain all the analysis regions in the CT image of the vertebral body; according to the morphological characteristics of the edge of each analysis region and the distribution characteristics of the edge pixel points, the boundary flatness of each analysis region is obtained; according to the boundary flatness difference between each analysis region and all adjacent analysis regions and the number of adjacent analysis regions, the vertebral attribution degree of each analysis region is obtained; according to the vertebral attribution degree, all vertebral block regions are screened out; an analysis region other than the vertebral block region is selected as a reference region; according to the area characteristics of the reference region and the distance and area difference between the adjacent vertebral block regions, the fragment possibility degree of the reference region as a bone fragment is obtained; according to the fragment possibility degree, all bone fragment regions are screened out; according to the distance between the vertebral block region and the bone fragment region, all bone fragment regions are classified and matched with all vertebral block regions to obtain all vertebral part regions; according to the boundary flatness, the number of bone fragment regions and the gray distribution in each vertebral part region, the osteoporosis possibility degree of each vertebral part region is obtained; an auxiliary decision module is used for screening all osteoporosis regions according to the osteoporosis possibility degree; according to the volume of the vertebral block region and the volume of the bone fragment region in each osteoporosis region and the number of osteoporosis regions, the osteoporosis degree of the patient is obtained; and the manual reduction of the vertebral fracture is assisted in decision-making according to the osteoporosis degree of the patient.

[0005] Further, the method for obtaining the boundary flatness comprises: establishing a Cartesian coordinate system with the lower right corner of the CT image of the vertebral body as the origin; selecting a target region from the analysis regions; performing edge detection on the target region to obtain all edge pixel points of the analysis region; performing straight line fitting on all edge pixel points to obtain the boundary straight lines corresponding to all boundaries of the target region; taking the two straight lines passing through the upper left corner of the target region as the target straight line and the reference straight line; and obtaining the boundary flatness according to the boundary flatness calculation formula, which is as follows: In the formula, represents the boundary flatness of the i-th analysis region; represents the boundary flatness of the i-th analysis region; ​The angle between the target line and the reference line in the region to be analyzed; This indicates the number of pixels within the target area contained in the target line; This indicates the number of pixels within the target area contained in the reference line; Represents an exponential function with the natural constant as its base; This represents the absolute value function.

[0006] Furthermore, the method for obtaining the degree of vertebral body attribution includes: obtaining the degree of vertebral body attribution according to the vertebral body attribution calculation formula, which is shown below: In the formula, Indicates the first The degree of vertebral body attribution in the region to be analyzed; Indicates the first The flatness of the boundary of the region to be analyzed; Indicates the relationship with the first The number of other regions to be analyzed adjacent to the region to be analyzed; Indicates the relationship with the first The adjacent region to be analyzed is the first The flatness of the boundaries of other regions to be analyzed; Indicates the relationship with the first The adjacent region to be analyzed is the first The flatness of the boundaries of other regions to be analyzed; Represents the normalization function; This represents the absolute value function.

[0007] Furthermore, the method for obtaining the vertebral block region includes: taking the region to be analyzed that has a vertebral body attribution degree greater than a preset first threshold as the vertebral block region.

[0008] Furthermore, the method for obtaining the fragment probability degree of the reference region as a bone fragment includes: obtaining the fragment probability degree according to the fragment probability degree calculation formula, the fragment probability degree calculation formula is as follows: In the formula, Indicates the reference area number; Indicates the likelihood of a bone fragment in the reference area; This indicates the distance between the reference area and the nearest vertebral mass region; Indicates the area of ​​the reference region; This indicates that the area of ​​all vertebral mass regions in a vertebral CT image is homogeneous. Indicates the first adjacent to the reference region The area of ​​each vertebral region; Describes the minimum value function; Represents an exponential function with the natural constant as its base; denotes a normalization function; denotes an absolute value function.

[0009] Further, the method for obtaining the bone fragment region comprises: regarding each to-be-analyzed region other than the vertebral body region with the fragmentation possibility greater than the preset second threshold as a bone fragment region.

[0010] Further, the method for obtaining the vertebral body part region comprises: optionally taking one bone fragment region as a reference bone fragment region; taking the vertebral body region closest to the reference bone fragment region as a matched vertebral body region of the reference bone fragment region, and classifying the reference bone fragment region and the matched vertebral body region into one region; traversing all bone fragment regions to obtain each vertebral body region and all matched bone fragment regions, and grouping each vertebral body region and all matched bone fragment regions into one vertebral body part region.

[0011] Further, the method for obtaining the osteoporosis possibility degree comprises: taking the lower right corner as a starting point to calculate a first boundary flatness degree in the vertebral body region in each vertebral body part region; and obtaining the osteoporosis possibility degree according to an osteoporosis possibility degree calculation formula, which is as follows: In the formula, denotes an osteoporosis possibility degree of the i-th vertebral body part region; denotes an osteoporosis possibility degree of the i-th vertebral body part region; denotes a number of bone fragment regions contained in the i-th vertebral body part region; denotes a number of bone fragment regions contained in the i-th vertebral body part region; denotes a mean gray value of the i-th vertebral body part region; denotes a mean gray value of the i-th vertebral body part region; denotes a mean gray value of all vertebral body part regions in the vertebral body CT image; denotes a boundary flatness degree of the vertebral body region in the i-th vertebral body part region; denotes a first boundary flatness degree of the vertebral body region in the i-th vertebral body part region; denotes a first boundary flatness degree of the vertebral body region in the i-th vertebral body part region; denotes a first boundary flatness degree of the vertebral body region in the i-th vertebral body part region; denotes a normalization function; denotes an absolute value function.

[0012] Further, the patient osteoporosis degree comprises: taking the vertebral body part region with the osteoporosis possibility degree greater than a preset third threshold as an osteoporosis region; constructing a three-dimensional vertebral body model of the patient by using the vertebral body CT image to obtain a three-dimensional region of the osteoporosis region; and obtaining the patient osteoporosis degree according to the distribution characteristics and volume of the three-dimensional region of the osteoporosis region in the three-dimensional vertebral body model of the patient, and the calculation formula is as follows: In the formula, denotes a patient osteoporosis degree; a number of three-dimensional regions representing osteoporotic regions in a three-dimensional vertebral body model of a patient; a volume of a three-dimensional region representing a first osteoporotic region in a three-dimensional vertebral body model of a patient; a volume of a three-dimensional region representing a first osteoporotic region in a three-dimensional vertebral body model of a patient; a normalized function.

[0013] An osteoporotic fracture manual reduction decision-making method based on a vertebral body image, the method comprising: acquiring a vertebral body CT image of a patient; performing threshold segmentation on the vertebral body CT image to obtain all regions to be analyzed in the vertebral body CT image; obtaining a boundary flatness of each region to be analyzed according to a morphological feature of an edge of each region to be analyzed and a distribution feature of edge pixels; obtaining a vertebral body attribution degree of each region to be analyzed according to a difference in boundary flatness between each region to be analyzed and all adjacent regions to be analyzed and a number of adjacent regions to be analyzed; screening all vertebral block regions according to the vertebral body attribution degree; optionally selecting a region to be analyzed other than the vertebral block region as a reference region; obtaining a fragment possibility degree of the reference region as a bone fragment according to an area feature of the reference region and a distance and area difference between the reference region and adjacent vertebral block regions; screening all bone fragment regions according to the fragment possibility degree; classifying all bone fragment regions according to a distance between the vertebral block regions and the bone fragment regions, matching all bone fragment regions with all vertebral block regions, and obtaining all vertebral body partial regions; obtaining a possible osteoporosis degree of each vertebral body partial region according to a boundary flatness, a number of bone fragment regions, and a gray scale distribution in each vertebral body partial region; screening all osteoporotic regions according to the possible osteoporosis degree; obtaining a patient osteoporosis degree according to a volume of a vertebral block region and a volume of a bone fragment region in each osteoporotic region and a number of osteoporotic regions; and assisting in a decision-making of a manual reduction of a vertebral body fracture according to the patient osteoporosis degree.

[0014] The present application has the following advantages: in order to accurately identify the osteoporotic fracture area that may occur in the vertebral body part of a patient, the present application obtains a CT image of the vertebral body of the patient; since the vertebral body area is composed of multiple small vertebral blocks, and each vertebral block is regularly distributed, and in the CT image, the vertebral body area and other body tissue areas have a large difference in gray scale characteristics, the present application performs threshold segmentation on the CT image of the vertebral body to obtain all the analysis areas in the CT image of the vertebral body, wherein the analysis areas include multiple vertebral block areas and multiple bone fragment areas; since the vertebral block has a rectangular feature, its shape feature is relatively regular, and its edge feature is relatively clear, the present application obtains the boundary flatness of each analysis area according to the shape feature of the edge of each analysis area and the distribution feature of the edge pixel points, and screens out the vertebral block area according to the boundary flatness; since the distribution feature of the vertebral body presents a feature that one vertebral block area is connected to one vertebral block area, the present application further analyzes the number of adjacent analysis areas of the target area to obtain the vertebral body attribution degree of each analysis area, thereby screening out the vertebral block area; since a certain number of free bone fragments may occur near the vertebral body area of a patient with osteoporotic fracture, these bone fragment areas have a small area and are close to the vertebral block area, and the sum of the areas of the adjacent vertebral block areas and other vertebral block areas is similar, the present application obtains the bone fragment possibility of the reference area according to the area feature of the reference area and the distance and area difference between the adjacent vertebral block areas, thereby screening out the bone fragment area; since the vertebral block area to which different bone fragment areas belong before separation may be different, and the bone fragment area separated from a certain vertebral block area is closest to the vertebral block area, the present application obtains all the vertebral body part areas; since when a certain vertebra has osteoporotic performance, the vertebral block may have the characteristics of bone density and bone mass reduction, and the bone strength is reduced, and free bone fragments may exist, the present application analyzes the osteoporosis possibility of each vertebral body part area, thereby screening out the osteoporosis area, and judges the osteoporosis degree of the patient. The present application can accurately identify the osteoporotic fracture on the CT image of the vertebral body, thereby making the analysis and judgment of the relevant personnel on the patient more accurate. 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 only show some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained from these drawings without any creative effort.

[0016] Figure 1 A block diagram of an osteoporotic fracture manual reduction auxiliary decision system based on a vertebral body image is provided for an embodiment of the present application. DETAILED DESCRIPTION

[0017] To further clarify the technical means and effects taken by the present application to achieve the predetermined object of the application, the specific implementation, structure, features and effects of a vertebral body image-based osteoporotic fracture manual reduction auxiliary decision system according to the present application are described in detail as follows in combination with the drawings and preferred embodiments. 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 vertebral body image-based osteoporotic fracture manual reduction auxiliary decision system provided by the present application is described in detail below in combination with the drawings.

[0020] Please refer to Figure 1 which shows a vertebral body image-based osteoporotic fracture manual reduction auxiliary decision system provided by one embodiment of the present application, which includes an image acquisition module 101, an image analysis module 102 and an auxiliary decision module 103, and the specific steps are as follows: the image acquisition module 101 acquires the CT image of the vertebral body of the patient.

[0021] The present embodiment is mainly applied to the scene of accurately identifying the osteoporotic fracture area that may occur in the vertebral body of the patient, so the CT image of the vertebral body of the patient is first acquired to facilitate subsequent analysis.

[0022] In one embodiment of the present application, the patient is first made to lie on his back with his arms raised to minimize the physiological bending of the spine and movement artifacts, and the CT image of the patient is acquired by using the CT device, and the image is preprocessed to enhance the contrast of the picture, and finally the CT image of the vertebral body used in the subsequent steps is obtained.

[0023] It should be noted that the method of image acquisition and image processing is a technical means known to those skilled in the art, which is not limited here.

[0024] The image analysis module 102: threshold segmentation is performed on the vertebral body CT image to obtain all the analysis regions in the vertebral body CT image; the boundary flatness of each analysis region is obtained according to the morphological characteristics of the edge of each analysis region and the distribution characteristics of the edge pixels; the vertebral body attribution degree of each analysis region is obtained according to the boundary flatness difference between each analysis region and all adjacent analysis regions and the number of adjacent analysis regions; all vertebral block regions are screened out according to the vertebral body attribution degree; an analysis region other than the vertebral block region is selected as a reference region; the fragment possibility of the reference region is obtained according to the area characteristics of the reference region and the distance and area difference between the reference region and adjacent vertebral block regions; all bone fragment regions are screened out according to the fragment possibility; all vertebral body part regions are obtained by classifying and matching all bone fragment regions with all vertebral block regions according to the distance between the vertebral block regions and the bone fragment regions; and the osteoporosis possibility of each vertebral body part region is obtained according to the boundary flatness, the number of bone fragment regions and the gray scale distribution in each vertebral body part region.

[0025] Since the vertebral body region is composed of a plurality of small vertebral blocks, each vertebral block is regularly distributed, and the gray scale characteristics of the vertebral body region and other body tissue regions are quite different in the CT image, in the embodiment of the present application, threshold segmentation is performed on the vertebral body CT image to obtain all the analysis regions in the vertebral body CT image, wherein the analysis regions include a plurality of vertebral block regions and a plurality of bone fragment regions, and the following steps are used to distinguish them.

[0026] Since the vertebral block has a rectangular feature and its shape characteristics are regular and the edge characteristics are clear, in the embodiment of the present application, the boundary flatness of each analysis region is obtained according to the morphological characteristics of the edge of each analysis region and the distribution characteristics of the edge pixels, and the vertebral block region is screened out according to the boundary flatness.

[0027] Preferably, in one embodiment of the present application, the method for obtaining the boundary flatness comprises: establishing a Cartesian coordinate system with the lower right corner of the vertebral body CT image as the origin; selecting one analysis region as a target region; performing edge detection on the target region to obtain all the edge pixels of the analysis region; performing straight line fitting on all the edge pixels to obtain the boundary straight lines corresponding to all the boundaries of the target region; and taking the two straight lines passing through the upper left corner of the target region as a target straight line and a reference straight line, and it should be noted that the boundary straight line with the smallest angle with the x-axis is taken as the target straight line and the boundary straight line with the smallest angle with the y-axis is taken as the target straight line.

[0028] The boundary flatness is obtained according to the boundary flatness calculation formula, and the boundary flatness calculation formula is as follows: In the formula, the boundary flatness of the target region is calculated according to the following formula: representing the boundary flatness of the first representing the angle between the target straight line and the reference straight line of the first representing the number of pixels in the target region contained in the target straight line; representing the number of pixels in the target region contained in the reference straight line; representing an exponential function with a natural constant as the base number; representing an absolute value function.

[0029] In the boundary flatness calculation formula, the angle between the target straight line and the reference straight line is closer to a right angle, that is, the smaller the angle is, the closer the morphological characteristics of the two boundary curves are to the morphological characteristics of the vertebral body, and the greater the boundary flatness of the region to be analyzed is; the fewer the number of pixels on the two straight lines, the more likely the edge is curved, and the smaller the boundary flatness is.

[0030] Since the distribution characteristics of the vertebral body present the feature of one vertebral block region connecting one vertebral block region, the number of adjacent regions to be analyzed of the target region is analyzed, that is, the vertebral body attribution degree of each region to be analyzed is obtained according to the difference in boundary flatness between each region to be analyzed and all adjacent regions to be analyzed and the number of adjacent regions to be analyzed.

[0031] Preferably, in an embodiment of the present application, the method for obtaining the vertebral body attribution degree comprises: obtaining the vertebral body attribution degree according to the vertebral body attribution degree calculation formula, and the vertebral body attribution degree calculation formula is as follows: In the formula, represents the vertebral body attribution degree of the first representing the boundary flatness of the first representing the number of other regions to be analyzed adjacent to the first representing the boundary flatness of the first other region to be analyzed adjacent to the first representing the boundary flatness of the first other region to be analyzed adjacent to the first representing the boundary flatness of the first other region to be analyzed adjacent to the first representing a normalization function; representing an absolute value function.

[0032] In the vertebral body attribution degree calculation formula, the first ​​​​​​​​​​The greater the boundary flatness of the to-be-analyzed region, the number of adjacent to-be-analyzed regions is 2, and the smaller the difference in boundary flatness of the adjacent to-be-analyzed regions, the more likely the first to-be-analyzed region is a vertebral body region, and the greater the vertebral body attribution degree of the first to-be-analyzed region. The greater the vertebral body attribution degree of the first to-be-analyzed region.

[0033] Preferably, in an embodiment of the present application, the method for obtaining the vertebral body region comprises: taking the to-be-analyzed region with the vertebral body attribution degree greater than a preset first threshold as the vertebral body region.

[0034] In an 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 and is not limited herein.

[0035] Since a certain number of free bone fragments can appear near the vertebral body region of a patient with osteoporotic fracture, the bone fragment region presents a small area and is close to the vertebral body region, and the sum of the area of the adjacent vertebral body region and the area of the bone fragment region is similar to the area of other vertebral body regions, a to-be-analyzed region other than the vertebral body region is selected as a reference region; and the possibility of the reference region being a bone fragment is obtained according to the area characteristics of the reference region and the distance and area difference between the reference region and the adjacent vertebral body region.

[0036] Preferably, in an embodiment of the present application, the method for obtaining the possibility of the reference region being a bone fragment comprises: obtaining the possibility of the reference region being a bone fragment according to a fragment possibility calculation formula, and the fragment possibility calculation formula is as follows: In the formula, represents the serial number of the reference region; represents the possibility of the reference region being a bone fragment; represents the distance between the reference region and the closest vertebral body region; represents the area of the reference region; represents the area homogeneity of all vertebral body regions in the CT image of the vertebral body; represents the area of the first vertebral body region adjacent to the reference region; represents a minimum function; represents an exponential function with a natural constant as the base; represents a normalization function; represents an absolute value function.

[0037] ​The smaller the distance between the reference region and the nearest vertebral block region and the smaller the area of the reference region, the more the reference region conforms to the morphological characteristics of the free bone fragment, and the greater the likelihood of the reference region being a bone fragment. The smaller the difference between the area sum of the reference region and the adjacent smallest vertebral block region and the mean value of the area of the vertebral block region, the more likely the reference region is a bone fragment region separated from the vertebral block region, and the greater the likelihood of the reference region being a bone fragment. The smaller the difference between the area sum of the reference region and the adjacent smallest vertebral block region and the mean value of the area of the vertebral block region, the more likely the reference region is a bone fragment region separated from the vertebral block region, and the greater the likelihood of the reference region being a bone fragment.

[0038] According to the likelihood of the fragment, all bone fragment regions are screened. Preferably, in an embodiment of the present application, the method for obtaining the bone fragment region comprises: regarding each region to be analyzed outside the vertebral block region with a likelihood of the fragment greater than a preset second threshold value as a bone fragment region.

[0039] In an embodiment of the present application, the preset second threshold value is set to 0.7. It should be noted that the preset second threshold value can be set as desired and is not limited herein.

[0040] Since the vertebral block region to which different bone fragment regions belong before separation can be different, and the bone fragment region separated from a certain vertebral block region is closest to the vertebral block region, in an embodiment of the present application, all bone fragment regions are classified and matched with all vertebral block regions according to the distance between the vertebral block region and the bone fragment region, to obtain all vertebral body part regions.

[0041] Preferably, in an embodiment of the present application, the method for obtaining the vertebral body part region comprises: selecting an optional bone fragment region as a reference bone fragment region; selecting the vertebral block region closest to the reference bone fragment region as the matching vertebral block region of the reference bone fragment region, and classifying the reference bone fragment region and the matching vertebral block region as a region; traversing all bone fragment regions to obtain each vertebral block region and all matching bone fragment regions, and grouping each vertebral block region and all matching bone fragment regions into a vertebral body part region.

[0042] Since when a certain vertebra has osteoporosis, the vertebral block can have the characteristics of decreased bone density and bone mass, and reduced bone strength, and can have free bone fragments. Therefore, in an embodiment of the present application, the osteoporosis likelihood of each vertebral body part region is obtained according to the boundary smoothness, the number of bone fragment regions, and the gray scale distribution in each vertebral body part region.

[0043] Preferably, in one embodiment of the present invention, the method for obtaining the degree of osteoporosis includes: calculating the flatness of the first boundary in the vertebral block region within each vertebral body region, starting from the lower right corner. It should be noted that two straight lines passing through the lower right corner of the vertebral block region are used as the target line and the reference line, respectively. The criteria for judging the target line and the reference line remain unchanged, and the calculation formula for the flatness of the first boundary is exactly the same as the calculation formula for the flatness of the boundary, which will not be elaborated here.

[0044] The probability of osteoporosis is determined using the formula shown below: In the formula, Indicates the first The degree of osteoporosis in certain areas of the vertebral body; Indicates the first The number of areas containing bone fragments in each vertebral body region; Indicates the first Mean gray value of a portion of the vertebral body; This represents the average grayscale value of all vertebral body regions in a vertebral CT image. Indicates the first The smoothness of the boundary of the vertebral mass region within a vertebral body region; Indicates the first The flatness of the first boundary of the vertebral mass region within a vertebral body portion; Represents the normalization function; This represents the absolute value function.

[0045] In the formula for calculating the probability of osteoporosis, the first... The difference between the mean gray value of a vertebral body region and the mean gray value of a vertebral body region The larger the value, the more likely it is to be the first. The greater the decrease in bone density in certain regions of the vertebral body, the higher the degree of bone density loss, and the more pronounced the decrease in bone density in the first vertebral body region. The greater the number of bone fragment regions within a vertebral body region, the more significant the influence of bone fragments. The more free bone fragments near the vertebral mass region within a vertebral body segment, the greater the degree of bone quality loss. The greater the degree of osteoporosis in a particular region of the vertebral body, the more likely it is to be; The difference between the smoothness of the boundary of the vertebral mass region within a vertebral body segment and the smoothness of the first boundary. The larger the value, the greater the likelihood of free bone fragments being generated. The greater the degree of osteoporosis in a particular region of the vertebral body, the more likely it is to be.

[0046] Decision Support Module 103: Based on the degree of osteoporosis, all osteoporotic areas are screened out; based on the volume of the vertebral mass area and the volume of the bone fragment area in each osteoporotic area, as well as the number of osteoporotic areas, the degree of osteoporosis of the patient is obtained; and the manual reduction of vertebral fractures is assisted in making decisions based on the degree of osteoporosis of the patient.

[0047] Based on the degree of osteoporosis obtained from the above steps, the areas where osteoporotic fractures may occur in patients can be identified, i.e., osteoporotic regions. Furthermore, the degree of osteoporosis in patients can be determined by the number and internal characteristics of these osteoporotic regions.

[0048] Preferably, in one embodiment of the present invention, the degree of osteoporosis of the patient includes: the vertebral body portion with a possible degree of osteoporosis greater than a preset third threshold is defined as the osteoporosis region; in one embodiment of the present invention, the preset third threshold is set to 0.6, and can be set by the user, and is not limited here.

[0049] Using vertebral CT images, a three-dimensional vertebral model of the patient is constructed to obtain the three-dimensional region of osteoporosis. This step is a well-known technique in the art and will not be described in detail here.

[0050] Based on the distribution characteristics and volume of the osteoporotic region in the three-dimensional vertebral model of the patient, the degree of osteoporosis of the patient is obtained, and the calculation formula is as follows: In the formula, Indicates the degree of osteoporosis in the patient; This indicates the number of three-dimensional regions representing osteoporotic areas in a patient's three-dimensional vertebral model. This indicates the first vertebra in the patient's three-dimensional vertebral model. The volume of a three-dimensional region in an osteoporotic area; This indicates the first vertebra in the patient's three-dimensional vertebral model. The volume of bone fragments within a three-dimensional region of an osteoporotic area; This represents the normalization function.

[0051] In the formula for calculating the degree of osteoporosis in patients, the more three-dimensional regions of osteoporotic areas there are, and the more the number of three-dimensional vertebral models in the model increases, the better. The larger the volume of the three-dimensional region of the osteoporotic area, the larger the total volume of bone fragments, indicating that the osteoporotic fracture of the patient's vertebral body is more severe.

[0052] A manual reduction auxiliary decision report is generated based on the patient's degree of osteoporosis to assist in the decision-making process for manual reduction of vertebral fractures.

[0053] In summary, a CT image of a patient's vertebral body is acquired; threshold segmentation is performed on the CT image of the vertebral body to obtain all regions to be analyzed in the CT image of the vertebral body; the boundary flatness of each region to be analyzed is obtained according to the morphological features of the edge of each region to be analyzed and the distribution features of the edge pixel points; the vertebral body attribution degree of each region to be analyzed is obtained according to the boundary flatness difference between each region to be analyzed and all adjacent regions to be analyzed and the number of adjacent regions to be analyzed; all vertebral block regions are screened out according to the vertebral body attribution degree; an optional region to be analyzed other than the vertebral block region is selected as a reference region; the fragment possibility of the reference region as a bone fragment is obtained according to the area features of the reference region and the distance and area difference between the reference region and adjacent vertebral block regions; all bone fragment regions are screened out according to the fragment possibility; all bone fragment regions are classified and matched with all vertebral block regions according to the distance between the vertebral block regions and the bone fragment regions to obtain all vertebral body partial regions; the osteoporosis possibility of each vertebral body partial region is obtained according to the boundary flatness, the number of bone fragment regions and the gray distribution in each vertebral body partial region; all osteoporosis regions are screened out according to the osteoporosis possibility; the osteoporosis degree of the patient is obtained according to the volume of the vertebral block regions and the volume of the bone fragment regions in each osteoporosis region and the number of osteoporosis regions; and the method is used to assist in decision-making for the reduction of vertebral fracture.

[0054] One embodiment of the present application provides a method for assisting decision of manual reduction of osteoporotic fracture based on vertebral body image, the method comprising: acquiring a CT image of a vertebral body of a patient; performing threshold segmentation on the CT image of the vertebral body to obtain all to-be-analyzed regions in the CT image of the vertebral body; obtaining a boundary flatness of each to-be-analyzed region according to a morphological feature of an edge of each to-be-analyzed region and a distribution feature of edge pixels; obtaining a vertebral body attribution degree of each to-be-analyzed region according to a difference in boundary flatness between each to-be-analyzed region and all adjacent to-be-analyzed regions and a number of adjacent to-be-analyzed regions; screening all vertebral block regions according to the vertebral body attribution degree; optionally selecting a to-be-analyzed region other than the vertebral block region as a reference region; obtaining a fragment possibility degree of the reference region as a bone fragment according to an area feature of the reference region and a distance and area difference between the reference region and adjacent vertebral block regions; screening all bone fragment regions according to the fragment possibility degree; classifying all bone fragment regions according to a distance between the vertebral block regions and the bone fragment regions, matching all bone fragment regions with all vertebral block regions, and obtaining all vertebral body partial regions; obtaining a bone osteoporosis possibility degree of each vertebral body partial region according to the boundary flatness, the number of bone fragment regions and a gray scale distribution in each vertebral body partial region; screening all bone osteoporosis regions according to the bone osteoporosis possibility degree; obtaining a bone osteoporosis degree of the patient according to a volume of the vertebral block regions and a volume of the bone fragment regions in each bone osteoporosis region and a number of the bone osteoporosis regions; and assisting decision of manual reduction of vertebral fracture according to the bone osteoporosis degree of the patient.

[0055] It should be noted that the above-mentioned sequence 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.

[0056] 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 differences from other embodiments.

Claims

1. A decision-making system for manual reduction of osteoporotic fractures based on vertebral imaging, characterized in that, The system includes: The image acquisition module is used to acquire CT images of the patient's vertebral body; The image analysis module is used to perform threshold segmentation on the vertebral CT image to obtain all regions to be analyzed in the vertebral CT image; obtain the boundary smoothness of each region to be analyzed based on the morphological features of the edge of each region to be analyzed and the distribution features of edge pixels; obtain the vertebral body affiliation degree of each region to be analyzed based on the difference in boundary smoothness between each region to be analyzed and all adjacent regions to be analyzed, and the number of adjacent regions to be analyzed; filter out all vertebral block regions based on the vertebral body affiliation degree; randomly select one region to be analyzed other than the vertebral block region as a reference region; obtain the fragment probability of the reference region being a bone fragment based on the area features of the reference region and the distance and area difference between it and adjacent vertebral block regions; filter out all bone fragment regions based on the fragment probability degree; classify all bone fragment regions based on the distance between the vertebral block region and the bone fragment region, match them with all vertebral block regions to obtain all vertebral body partial regions; obtain the osteoporosis probability of each vertebral body partial region based on the boundary smoothness, the number of bone fragment regions, and the grayscale distribution within each vertebral body partial region. The auxiliary decision-making module is used to filter out all osteoporotic areas based on the degree of osteoporosis; to obtain the patient's degree of osteoporosis based on the volume of the vertebral mass area and the volume of the bone fragment area within each osteoporotic area, as well as the number of osteoporotic areas; and to make auxiliary decisions on the manual reduction of vertebral fractures based on the patient's degree of osteoporosis. The method for obtaining the boundary smoothness includes: establishing a Cartesian coordinate system with the lower right corner of the vertebral CT image as the origin; selecting any region to be analyzed as the target region; performing edge detection on the target region to obtain all edge pixels of the region to be analyzed; performing line fitting on all edge pixels to obtain the boundary lines corresponding to all boundaries of the target region; using two lines passing through the upper left corner of the target region as the target line and the reference line, respectively; and obtaining the boundary smoothness according to the boundary smoothness calculation formula, which is shown below: In the formula, Indicates the first The flatness of the boundary of the region to be analyzed; Indicates the first The angle between the target line and the reference line in the region to be analyzed; This indicates the number of pixels within the target area contained in the target line; This indicates the number of pixels within the target area contained in the reference line; Represents an exponential function with the natural constant as the base; Represents the absolute value function; The method for obtaining the degree of vertebral body attribution includes: obtaining the degree of vertebral body attribution according to the vertebral body attribution calculation formula, which is shown below: In the formula, Indicates the first The degree of vertebral body attribution in the region to be analyzed; Indicates the first The flatness of the boundary of the region to be analyzed; Indicates the relationship with the first The number of other regions to be analyzed adjacent to the region to be analyzed; Indicates the relationship with the first The adjacent region to be analyzed is the first The flatness of the boundaries of other regions to be analyzed; Indicates the relationship with the first The adjacent region to be analyzed is the first The flatness of the boundaries of other regions to be analyzed; Represents the normalization function; Represents the absolute value function; The method for obtaining the fragment probability degree of a bone fragment in the reference area includes: obtaining the fragment probability degree according to the fragment probability degree calculation formula, which is as follows: In the formula, Indicates the reference area number; Indicates the likelihood of a bone fragment in the reference area; This indicates the distance between the reference area and the nearest vertebral mass region; Indicates the area of ​​the reference region; This indicates that the area of ​​all vertebral mass regions in a vertebral CT image is homogeneous. Indicates the first adjacent to the reference region The area of ​​each vertebral region; Describes the minimum value function; Represents an exponential function with the natural constant as the base; Represents the normalization function; Represents the absolute value function; The method for obtaining the vertebral body partial region includes: selecting any bone fragment region as a reference bone fragment region; taking the vertebral block region closest to the reference bone fragment region as the matching vertebral block region of the reference bone fragment region; classifying the reference bone fragment region and the matching vertebral block region into one category; traversing all bone fragment regions to obtain each vertebral block region and all matching bone fragment regions; and combining each vertebral block region and all matching bone fragment regions into a vertebral body partial region. The method for obtaining the degree of osteoporosis probability includes: calculating the flatness of the first boundary of the vertebral mass region within each vertebral body region, starting from the lower right corner; and obtaining the degree of osteoporosis probability according to the osteoporosis probability calculation formula, which is shown below: In the formula, Indicates the first The degree of osteoporosis in certain areas of the vertebral body; Indicates the first The number of areas containing bone fragments in each vertebral body region; Indicates the first Mean grayscale value of a portion of the vertebral body; This represents the average grayscale value of all vertebral body regions in a vertebral CT image. Indicates the first The smoothness of the boundary of the vertebral mass region within a vertebral body region; Indicates the first The flatness of the first boundary of the vertebral mass region within a vertebral body portion; Represents the normalization function; Represents the absolute value function; The degree of osteoporosis in the patient includes: defining the vertebral body region where the potential degree of osteoporosis is greater than a preset third threshold as the osteoporotic region; constructing a three-dimensional vertebral body model of the patient using the vertebral body CT images to obtain the three-dimensional region of the osteoporotic region; and obtaining the degree of osteoporosis based on the distribution characteristics and volume of the three-dimensional region of the osteoporotic region in the patient's three-dimensional vertebral body model, using the following calculation formula: In the formula, Indicates the degree of osteoporosis in the patient; This indicates the number of three-dimensional regions representing osteoporotic areas in a patient's three-dimensional vertebral model. This indicates the first vertebra in the patient's three-dimensional vertebral model. The volume of a three-dimensional region in an osteoporotic area; This indicates the first vertebra in the patient's three-dimensional vertebral model. The volume of bone fragments within a three-dimensional region of an osteoporotic area; This represents the normalization function.

2. The decision-making system for manual reduction of osteoporotic fractures based on vertebral imaging according to claim 1, characterized in that, The method for obtaining the bone fragment region includes: taking each region to be analyzed, excluding the vertebral block region where the probability of fragmentation is greater than a preset second threshold, as the bone fragment region.

3. A decision-making aid method for closed reduction of osteoporotic fractures based on vertebral imaging, characterized in that, The method includes: Obtain CT images of the patient's vertebral body; Threshold segmentation is performed on the vertebral CT images to obtain all regions to be analyzed. The boundary smoothness of each region is determined based on its edge morphology and pixel distribution. The vertebral body affiliation of each region is determined based on the difference in boundary smoothness between each region and all adjacent regions, as well as the number of adjacent regions. All vertebral block regions are filtered out based on their affiliation. One region other than the vertebral block region is selected as a reference region. The likelihood of the reference region being a bone fragment is determined based on its area characteristics and the distance and area difference between it and adjacent vertebral block regions. All bone fragment regions are filtered out based on this likelihood. All bone fragment regions are classified based on the distance between the vertebral block regions and the bone fragment regions, and matched with all vertebral block regions to obtain all vertebral body partial regions. The osteoporosis probability of each vertebral body partial region is determined based on its boundary smoothness, the number of bone fragment regions, and its grayscale distribution. Based on the degree of osteoporosis, all osteoporotic areas are screened out; the patient's degree of osteoporosis is obtained based on the volume of the vertebral mass area and the volume of the bone fragment area within each osteoporotic area, as well as the number of osteoporotic areas; the patient's degree of osteoporosis is used to assist in decision-making regarding the manual reduction of vertebral fractures. The method for obtaining the boundary smoothness includes: establishing a Cartesian coordinate system with the lower right corner of the vertebral CT image as the origin; selecting any region to be analyzed as the target region; performing edge detection on the target region to obtain all edge pixels of the region to be analyzed; performing line fitting on all edge pixels to obtain the boundary lines corresponding to all boundaries of the target region; using two lines passing through the upper left corner of the target region as the target line and the reference line, respectively; and obtaining the boundary smoothness according to the boundary smoothness calculation formula, which is shown below: In the formula, Indicates the first The flatness of the boundary of the region to be analyzed; Indicates the first The angle between the target line and the reference line in the region to be analyzed; This indicates the number of pixels within the target area contained in the target line; This indicates the number of pixels within the target area contained in the reference line; Represents an exponential function with the natural constant as the base; Represents the absolute value function; The method for obtaining the degree of vertebral body attribution includes: obtaining the degree of vertebral body attribution according to the vertebral body attribution calculation formula, which is shown below: In the formula, Indicates the first The degree of vertebral body attribution in the region to be analyzed; Indicates the first The flatness of the boundary of the region to be analyzed; Indicates the relationship with the first The number of other regions to be analyzed adjacent to the region to be analyzed; Indicates the relationship with the first The adjacent region to be analyzed is the first The flatness of the boundaries of other regions to be analyzed; Indicates the relationship with the first The adjacent region to be analyzed is the first The flatness of the boundaries of other regions to be analyzed; Represents the normalization function; Represents the absolute value function; The method for obtaining the fragment probability degree of a bone fragment in the reference area includes: obtaining the fragment probability degree according to the fragment probability degree calculation formula, which is as follows: In the formula, Indicates the reference area number; Indicates the likelihood of a bone fragment in the reference area; This indicates the distance between the reference area and the nearest vertebral mass region; Indicates the area of ​​the reference region; This indicates that the area of ​​all vertebral mass regions in a vertebral CT image is homogeneous. Indicates the first adjacent to the reference region The area of ​​each vertebral region; Describes the minimum value function; Represents an exponential function with the natural constant as the base; Represents the normalization function; Represents the absolute value function; The method for obtaining the vertebral body partial region includes: selecting any bone fragment region as a reference bone fragment region; taking the vertebral block region closest to the reference bone fragment region as the matching vertebral block region of the reference bone fragment region; classifying the reference bone fragment region and the matching vertebral block region into one category; traversing all bone fragment regions to obtain each vertebral block region and all matching bone fragment regions; and combining each vertebral block region and all matching bone fragment regions into a vertebral body partial region. The method for obtaining the degree of osteoporosis probability includes: calculating the flatness of the first boundary of the vertebral mass region within each vertebral body region, starting from the lower right corner; and obtaining the degree of osteoporosis probability according to the osteoporosis probability calculation formula, which is shown below: In the formula, Indicates the first The degree of osteoporosis in certain areas of the vertebral body; Indicates the first The number of areas containing bone fragments in each vertebral body region; Indicates the first Mean grayscale value of a portion of the vertebral body; This represents the average grayscale value of all vertebral body regions in a vertebral CT image. Indicates the first The smoothness of the boundary of the vertebral mass region within a vertebral body region; Indicates the first The flatness of the first boundary of the vertebral mass region within a vertebral body portion; Represents the normalization function; Represents the absolute value function; The degree of osteoporosis in the patient includes: defining the vertebral body region where the potential degree of osteoporosis is greater than a preset third threshold as the osteoporotic region; constructing a three-dimensional vertebral body model of the patient using the vertebral body CT images to obtain the three-dimensional region of the osteoporotic region; and obtaining the degree of osteoporosis based on the distribution characteristics and volume of the three-dimensional region of the osteoporotic region in the patient's three-dimensional vertebral body model, using the following calculation formula: In the formula, Indicates the degree of osteoporosis in the patient; This indicates the number of three-dimensional regions representing osteoporotic areas in a patient's three-dimensional vertebral model. This indicates the first vertebra in the patient's three-dimensional vertebral model. The volume of a three-dimensional region in an osteoporotic area; This indicates the first vertebra in the patient's three-dimensional vertebral model. The volume of bone fragments within a three-dimensional region of an osteoporotic area; This represents the normalization function.

Images