Intelligent identification method and system for osteoporosis area

Abstract

The invention relates to the technical field of image data processing, in particular to an intelligent identification method and system for an osteoporosis area, and the method comprises the following steps: extracting edge sharpness and texture uniformity parameters from a two-dimensional image, adjusting the gray threshold of the image, and segmenting an osteoporosis candidate area. According to the invention, through integration of image data processing and three-dimensional modeling, the ability of analyzing the osteoporosis area is improved. When a two-dimensional image is processed, edge sharpness and texture uniformity parameters are extracted, and a gray threshold is adjusted, so that an osteoporosis candidate region is identified, a foreground and a background in the image are isolated, and a segmentation result is further refined by using intensity difference and connectivity between pixel points; therefore, high-precision data are provided in bone tissue structure and bone mineral density analysis. And the two-dimensional image is converted into the three-dimensional voxels, and the three-dimensional model is constructed, so that the analysis depth is increased, and the evaluation of the structural stability and the stress point is more comprehensive.

Description

Technical Field

[0001] The present invention relates to the technical field of image data processing, and in particular to an osteoporosis area intelligent recognition method and system thereof. Background Art

[0002] Osteoporosis area recognition methods involve the development and use of image data processing techniques to identify and analyze areas of osteoporosis. This method is often applied in medical imaging, especially when processing X-rays, CT scans or MRI images, in order to assess osteoporosis lesions at an early stage.

[0003] However, existing technologies are often limited to two-dimensional image analysis, which limits their ability to distinguish the difference between the lesion area and normal bone tissue. In addition, existing technologies are usually not sophisticated enough in grayscale adjustment and edge sharpness analysis of images, which may lead to insufficient accuracy in disease diagnosis. This technical deficiency may lead to missed diagnosis or misdiagnosis in practical applications, especially in the early stages of the disease, which may lead to missed treatment opportunities, thereby affecting the patient's final treatment effect. Summary of the invention

[0004] The purpose of the present invention is to solve the shortcomings of the prior art and to propose an intelligent identification method and system for osteoporosis areas.

[0005] In order to achieve the above object, the present invention adopts the following technical solution, a method for intelligently identifying osteoporosis areas, comprising the following steps:

[0006] Extract edge sharpness and texture uniformity parameters from the two-dimensional image, adjust the grayscale threshold of the image, segment the osteoporosis candidate area, obtain segmentation result data, use the intensity difference between pixels in the segmentation result data to analyze the connectivity of the pixels, identify and isolate the foreground and background by setting the connection threshold, and generate foreground isolation data;

[0007] Based on the foreground isolation data, density and accuracy detection are performed, bone tissue structure feature analysis and bone density analysis are performed on the foreground isolation data to generate a bone density result;

[0008] Convert each pixel of the two-dimensional image into a three-dimensional voxel, record the conversion parameters and voxel size, construct a framework of a three-dimensional model, apply volume and coordinate data, calculate the structural stability and stress point distribution of the three-dimensional model, generate stress analysis data, perform bone microstructure integrity detection based on the stress analysis data, and generate microstructure integrity data;

[0009] The stress value and distribution coordinates are extracted from the stress analysis data, the stress area in the three-dimensional model is drawn, and the resolution and color depth are optimized to generate a stress visualization map.

[0010] Preferably, the steps of obtaining the segmentation result data are:

[0011] Extracting edge sharpness parameters and texture uniformity parameters from the two-dimensional image to obtain edge sharpness data and texture uniformity data;

[0012] Based on the edge sharpness data and the texture uniformity data, adjusting the grayscale threshold of the image to obtain adjusted image data;

[0013] Based on the adjusted image data, image segmentation is performed to identify and segment the osteoporosis candidate area to obtain segmentation result data.

[0014] Preferably, the steps of acquiring the foreground isolation data are:

[0015] Using the intensity difference between the pixels in the segmentation result data, the intensity difference between each pair of adjacent pixels is calculated to obtain intensity difference data;

[0016] Based on the intensity difference data, the formula is used:

[0017]

[0018] Calculate the connectivity of each pair of pixels (i, j); where C(i, j) represents the connectivity between pixels i and j, I ik and I jk Represent the intensity of pixel i and j in the kth color channel respectively, where K is the number of color channels of the image;

[0019] Based on the connectivity, a connection threshold T is set and a judgment is made for each pair of pixels. If C(i,j)>

[0020] T, the pixel (i, j) is considered to be disconnected, and the foreground and background are identified and isolated to obtain the foreground isolation data.

[0021] Preferably, the steps of obtaining the bone density result are:

[0022] Extracting the bone tissue area from the foreground isolation data, calculating the total number of pixels in the area, and obtaining bone tissue area data;

[0023] Based on the bone tissue area data, the average density of the bone tissue area is calculated using the following formula:

[0024]

[0025] Among them, ρ is the average density, N is the total number of pixels in the bone tissue area, and D i Represents the density value of the i-th pixel;

[0026] Based on the average density, the structural characteristics of the bone tissue are analyzed, the uniformity and integrity of the bone tissue are verified, and a bone density result is generated.

[0027] Preferably, the steps of acquiring the stress analysis data are:

[0028] Convert each pixel of the two-dimensional image into a three-dimensional voxel, record the conversion parameter and voxel size of each voxel, and obtain voxel parameter data;

[0029] Based on the voxel parameter data, the framework of the three-dimensional model is constructed and the total volume of the model is calculated. The calculation formula is:

[0030]

[0031] Where S is the total volume of the model, M is the total number of voxels, and V i represents the volume of the ith voxel, C i represents the volume contribution of the i-th voxel;

[0032] The total volume of the model is used to analyze the structural stability and stress point distribution of the three-dimensional model, calculate the stress contribution of each voxel, and generate stress analysis data.

[0033] Preferably, the steps of acquiring the microstructure integrity data are:

[0034] Based on the stress analysis data, analyzing the behavior of each voxel under stress, calculating the stress influence factor of each voxel, and obtaining stress influence data;

[0035] Based on the stress impact data, the integrity index of the bone microstructure is calculated using the following formula:

[0036]

[0037] Where W is the completeness index, Z represents the total number of voxels, and P j represents the stress influence factor of the jth voxel, T j represents the theoretical stress threshold of the jth voxel;

[0038] Based on the integrity index, the integrity of the bone microstructure is judged. If the index is lower than a threshold, it is considered that the microstructure is damaged, and microstructure integrity data is generated.

[0039] Preferably, the steps of obtaining the stress visualization map are:

[0040] Extracting the stress value and three-dimensional coordinates of each voxel from the stress analysis data to obtain stress value and coordinate data;

[0041] Using the stress values ​​and coordinate data, the stress region of each voxel is drawn on a graphical interface through three-dimensional graphics rendering to obtain an unoptimized stress visualization map;

[0042] The resolution and color depth of the unoptimized stress visualization image are adjusted to generate an optimized stress visualization image.

[0043] The present invention provides an intelligent recognition system, comprising:

[0044] The image preprocessing module extracts edge sharpness and texture uniformity parameters from the two-dimensional image, adjusts the grayscale threshold of the image, segments the osteoporosis candidate area through grayscale adjustment, analyzes the connectivity of pixels using the intensity difference between pixels, sets the connectivity threshold to identify and isolate the foreground and background, and generates foreground isolation data;

[0045] A bone density analysis module performs density detection and bone tissue structure feature analysis based on foreground isolation data, analyzes bone density and calculates bone tissue structure features, and generates bone density results and bone tissue structure feature data;

[0046] The three-dimensional modeling and stress analysis module converts each pixel of the two-dimensional image into a three-dimensional voxel, records the conversion parameters and voxel size, builds the framework of the three-dimensional model, and based on the bone density results and bone tissue structure characteristic data, uses volume and coordinate data to calculate the structural stability and stress point distribution of the three-dimensional model, generates stress analysis data, extracts stress values ​​and distribution coordinates from the stress analysis data, draws the stress area in the three-dimensional model, and generates a stress visualization diagram.

[0047] Compared with the prior art, the advantages and positive effects of the present invention are:

[0048] The present invention improves the ability to analyze osteoporotic areas by integrating image data processing and three-dimensional modeling. When processing two-dimensional images, the identification of osteoporotic candidate areas is achieved by extracting edge sharpness and texture uniformity parameters and adjusting the grayscale threshold, isolating the foreground and background in the image, and further refining the segmentation results by using the intensity difference and connectivity between pixels, thereby providing high-precision data in bone tissue structure and bone density analysis. Converting two-dimensional images into stereo voxels and constructing a three-dimensional model increases the depth of analysis, making the evaluation of structural stability and stress points more comprehensive, making it possible to detect potential osteoporosis problems at an early stage. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a schematic diagram of the steps of the present invention. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0051] See also Figure 1 The present invention provides a technical solution, a method for intelligently identifying osteoporosis areas, comprising the following steps:

[0052] Extract edge sharpness and texture uniformity parameters from the two-dimensional image, adjust the grayscale threshold of the image, segment the osteoporosis candidate area, obtain the segmentation result data, use the intensity difference between pixels in the segmentation result data to analyze the connectivity of the pixels, identify and isolate the foreground and background by setting the connection threshold, and generate foreground isolation data;

[0053] Based on the foreground isolation data, density and accuracy detection are performed, bone tissue structure feature analysis and bone density analysis are performed on the foreground isolation data to generate bone density results;

[0054] Convert each pixel of the two-dimensional image into a three-dimensional voxel, record the conversion parameters and voxel size, build a framework for the three-dimensional model, apply volume and coordinate data, calculate the structural stability and stress point distribution of the three-dimensional model, generate stress analysis data, perform bone microstructure integrity detection based on the stress analysis data, and generate microstructure integrity data;

[0055] Extract stress values ​​and distribution coordinates from stress analysis data, plot stress areas in 3D models, and optimize resolution and color depth to generate stress visualization maps.

[0056] The steps to obtain the segmentation result data are:

[0057] Extracting edge sharpness parameters and texture uniformity parameters from the two-dimensional image to obtain edge sharpness data and texture uniformity data;

[0058] Based on the edge sharpness data and the texture uniformity data, the grayscale threshold of the image is adjusted to obtain adjusted image data;

[0059] Based on the adjusted image data, image segmentation is performed to identify and segment the osteoporosis candidate area to obtain segmentation result data.

[0060] Specifically, when extracting edge sharpness and texture uniformity parameters, we first analyze the changes in pixel values ​​around each pixel in the image, and determine the edge sharpness by calculating the gradient size. The gradient calculation involves taking the difference between adjacent pixel values ​​and finding the square root of the sum of their squares. This data reflects the speed and intensity of color changes in the image. For the analysis of texture uniformity, the texture regularity and consistency are evaluated by calculating the standard deviation and mean of the pixel values ​​in a specific area based on the texture repeatability around the pixel. Through these two specific numerical analyses, the key image characteristic parameters are obtained, and then the edge sharpness data and texture uniformity data are generated.

[0061] Based on the edge sharpness data and texture uniformity data obtained, these specific data are used to dynamically adjust the grayscale threshold of the image. This adjustment increases the grayscale threshold based on the higher gradient value in the edge data in order to maintain a clear boundary in the high-contrast area. At the same time, the threshold in the area with lower texture uniformity data is lowered to reduce the impact of noise and improve the overall visual effect of the image. This threshold adjustment based on specific data directly affects the accuracy of image segmentation and the efficiency of image feature extraction, and ultimately the adjusted image data is obtained.

[0062] The adjusted image data is used for image segmentation. The specific method is to set a new grayscale threshold, analyze the entire image under this threshold, and regard the part exceeding this threshold as a possible osteoporosis area. This operation is performed by comparing the relationship between the grayscale value and the threshold pixel by pixel, ensuring that only clearly distinguished areas are marked as candidate areas. Such an operation ensures high accuracy and reliability of segmentation, so that the obtained segmentation result data is both accurate and direct and practical in operation.

[0063] The steps for obtaining foreground isolation data are:

[0064] Using the intensity difference between pixels in the segmentation result data, the intensity difference between each pair of adjacent pixels is calculated to obtain intensity difference data;

[0065] Based on the intensity difference data, the formula is used:

[0066]

[0067] Calculate the connectivity of each pair of pixels (i, j); where C(i, j) represents the connectivity between pixels i and j, I ik and I jk Represent the intensity of pixel i and j in the kth color channel respectively, where K is the number of color channels of the image;

[0068] Based on connectivity, a connection threshold T is set and each pair of pixels is judged. If C(i,j)>T, the pixel (i,j) is considered to be disconnected, and the foreground and background are identified and isolated to obtain the foreground isolation data.

[0069] Specifically, the intensity difference between pixels in the segmentation result data is calculated. The original value of the pixel intensity is first used for difference comparison. The calculation method uses a pixel-by-pixel comparison of the intensity of its surrounding pixels to obtain the intensity difference between each pixel and its adjacent pixels. This step ensures the accuracy of the segmentation and the retention of image details. These intensity difference data are then classified and analyzed according to the set statistical method to determine which intensity differences represent significant boundary lines. This process involves not only simple difference calculations, but also further statistical analysis of the difference results to ensure that areas with obvious differences can be correctly identified, thereby providing accurate basic data for the next step of image processing.

[0070] The benefit of the formula is that by calculating the color intensity difference between each pair of pixels, the local color changes in the image can be accurately measured, thereby effectively separating the foreground and background in the image and improving the accuracy of image segmentation. This method is particularly suitable for images with complex color transitions or unclear boundaries, and can improve the efficiency and effect of segmentation.

[0071] Calculation process: The image has three color channels. For pixels i and j, their intensities in the blue channel are 120 and 130, the green channel is 100 and 110, and the red channel is 150 and 140. Inserting these values ​​into the formula, we get:

[0072]

[0073] The result shows that the intensity difference between pixels i and j is 17.32, which can be used to determine whether the two points belong to the same foreground or background. If the connection threshold is set to less than 17.32, i and j will be considered disconnected and thus classified into different regions.

[0074] Based on the obtained foreground isolation data, the foreground area is further refined, which includes further connectivity testing and structural confirmation of each pixel in the foreground area. Through these data analysis, a more accurate definition of the foreground area can be obtained. Further data processing includes edge smoothing and noise removal of the isolated area to optimize the visual effect and ensure that the clarity and accuracy of the foreground data are effectively improved. These processing processes not only improve the image quality, but also provide a more reliable data basis for subsequent image recognition and analysis.

[0075] The steps to obtain bone density results are:

[0076] Extract the bone tissue area from the foreground isolation data, calculate the total number of pixels in the area, and obtain the bone tissue area data;

[0077] Based on the bone tissue area data, the average density of the bone tissue area is calculated using the following formula:

[0078]

[0079] Among them, ρ is the average density, N is the total number of pixels in the bone tissue area, and D i Represents the density value of the i-th pixel;

[0080] Based on the average density, the structural characteristics of bone tissue are analyzed, the uniformity and integrity of bone tissue are verified, and bone density results are generated.

[0081] Specifically, the bone tissue area is extracted from the foreground isolation data, and the total number of pixels in the area is determined by the basic pixel counting method. Then, the color value of each pixel is converted to grayscale to improve the accuracy of the analysis. By comparing the grayscale value of each pixel with the set threshold, the pixels that meet the characteristics of bone tissue are screened out, and then accurate bone tissue area data is obtained, which lays the foundation for subsequent density calculation and finally obtains accurately divided bone tissue area data.

[0082] formula The benefit of this method is that it can provide an objective quantitative indicator by averaging the density values ​​of all pixels in the bone tissue area, helping to evaluate and compare the degree of osteoporosis in different clinical scenarios.

[0083] Calculation process: Define N as the total number of pixels representing bone tissue in the image. For example, in a small image, there are N = 1000 pixels, and the density value of each pixel is as follows:

[0084] D1=0.32,D2=0.35,…,D 1000 = 0.30, calculate the average of all density values:

[0085]

[0086] The results show that the calculated ρ value can be used to determine the healthy level of bone density. If ρ is greater than a certain threshold, it may mean that the bone density is within a healthy range; if it is lower than a certain threshold, further medical intervention may be required.

[0087] After obtaining the average density data, the changing trend of bone tissue structure is analyzed by comparing it with historical data, and the standard deviation and coefficient of variation are used to evaluate the uniformity and integrity of bone tissue density. These statistical indicators help doctors understand the distribution of bone density and identify possible abnormal areas.

[0088] The steps to obtain stress analysis data are:

[0089] Convert each pixel of the two-dimensional image into a three-dimensional voxel, record the conversion parameter and voxel size of each voxel, and obtain voxel parameter data;

[0090] Based on the voxel parameter data, the framework of the three-dimensional model is constructed and the total volume of the model is calculated. The calculation formula is:

[0091]

[0092] Where S is the total volume of the model, M is the total number of voxels, and V i represents the volume of the ith voxel, C i represents the volume contribution of the i-th voxel;

[0093] The total volume of the model is used to analyze the structural stability and stress point distribution of the 3D model, calculate the stress contribution of each voxel, and generate stress analysis data.

[0094] Specifically, in the process of converting each pixel of a two-dimensional image into a three-dimensional voxel, it is first necessary to extract the position and color information of each pixel from the original two-dimensional image data, and then perform three-dimensional processing on each pixel, that is, convert each two-dimensional pixel into a three-dimensional voxel with a certain depth. This conversion process involves the spatial positioning of the pixel points and the determination of the voxel size, where the voxel size is calculated based on the image resolution and preset depth parameters. The depth and width of the voxel are usually set to be equivalent to the two-dimensional pixel, but the height (or depth) is an adjustable parameter, which allows the two-dimensional image to be converted into a three-dimensional model for further analysis. Finally, this step generates complete voxel parameter data.

[0095] Formula calculation process: A model consists of 10 voxels, and the volume of each voxel is V i is 1 cubic unit, and the contribution of each voxel is C i If both are 1, the formula becomes cubic unit;

[0096] This result shows that the total volume of the model is 10 cubic units, which can be directly used in subsequent structural analysis and load testing. This result indicates that the 3D model has sufficient volume to support its designed structural load. If the volume is too small, it may be necessary to redesign the voxel parameters or increase the number of voxels to ensure structural stability.

[0097] In the process of analyzing the structural stability and stress point distribution of the three-dimensional model, volume and coordinate data are used to calculate the position of each voxel in the model and its contribution to the structural stability of the model. This includes combining the volume data of each voxel with its spatial coordinates in the model, and simulating the stress response of each voxel when subjected to external force through physical modeling, so as to identify possible stress concentration points. The analysis process requires the setting of the material properties of the voxels, including but not limited to density, elastic modulus and compressive strength. The parameters are obtained based on the standard database of materials science to ensure the accuracy and reliability of the model analysis. Finally, this step generates stress analysis data, which provides important information about which parts of the model are most likely to fail in actual applications.

[0098] The steps to obtain microstructural integrity data are:

[0099] Based on the stress analysis data, the behavior of each voxel under stress is analyzed, the stress influence factor of each voxel is calculated, and the stress influence data is obtained;

[0100] Based on the stress impact data, the integrity index of bone microstructure is calculated using the following formula:

[0101]

[0102] Where W is the completeness index, Z represents the total number of voxels, and P j represents the stress influence factor of the jth voxel, T j represents the theoretical stress threshold of the jth voxel;

[0103] The integrity of the bone microstructure is judged based on the integrity index. If the index is lower than the threshold, the microstructure is considered to be damaged and microstructure integrity data is generated.

[0104] Specifically, based on the stress analysis data, the behavior of each voxel under stress is evaluated. The evaluation is carried out by monitoring the response of each voxel under different stress conditions and collecting the stress data of each voxel, so as to understand the force distribution that each voxel may suffer in actual use, thereby providing basic data for subsequent analysis and obtaining stress impact data;

[0105] The benefit of the formula is that by comparing the ratio of actual stress to theoretical threshold, the integrity of bone microstructure can be quantified, which provides a scientific basis for judging whether the material will fail in actual use. Parameter Z is the total number of voxels evaluated, which is directly obtained by counting from medical scan data. Parameter P j is the stress influencing factor of the jth voxel, obtained through the stress response data of each voxel in the physical simulation test; parameter T j is the theoretical stress threshold of the jth voxel, obtained according to the standard stress threshold data table provided in materials science research;

[0106] Calculation process: The total number of voxels Z is 100, and the stress influence factor of each voxel P j The average is 0.5, the theoretical stress threshold T j The average is 0.8, and the calculation process is as follows:

[0107]

[0108] W=0.625

[0109] The results showed that the integrity index was 0.625, which means that the integrity of the bone microstructure under the average stress it was subjected to reached 62.5% of the theoretical threshold. If this value is lower than a specific threshold, such as 0.5, it may indicate that the microstructure is damaged;

[0110] The integrity of the bone microstructure is judged based on the integrity index data. The judgment process includes using an algorithm to perform threshold analysis on the integrity index data. By setting the threshold, microstructure areas that may be damaged due to stress exceeding the tolerance range can be identified. If the index is lower than a certain threshold, such as 0.5, it is considered that the microstructure may be damaged, and microstructure integrity data is generated.

[0111] The steps to obtain the stress visualization diagram are:

[0112] Extract the stress value and three-dimensional coordinate of each voxel from the stress analysis data to obtain stress value and coordinate data;

[0113] Using stress values ​​and coordinate data, the stress area of ​​each voxel is drawn on a graphical interface through three-dimensional graphics rendering to obtain an unoptimized stress visualization map;

[0114] The resolution and color depth of the unoptimized stress visualization map are adjusted to generate an optimized stress visualization map.

[0115] Specifically, in the data processing stage, the stress analysis data needs to be loaded and parsed first, including reading the stress value of the voxel and its three-dimensional coordinates in the model from the data storage file. During the data parsing process, the integrity and format correctness of the data will be verified to ensure that the data of each voxel can be read accurately. After that, these data are stored in a structured manner in the memory to facilitate efficient calling in the subsequent graphics rendering process. This structured method usually involves creating an array of voxel objects, where each object stores the stress value and coordinate information of a voxel. The obtained stress value and coordinate data provide prepared input materials for the next step of three-dimensional rendering;

[0116] This step involves converting the structured stress values ​​and coordinate data into a graphical representation. During the graphics rendering process, the basic parameters of the rendering environment are first set, such as the viewing angle, light source position, and intensity. These settings are to ensure that the viewing effect of the three-dimensional image reflects the physical model as realistically as possible. Next, the data of each voxel is traversed, and the color depth of the voxel on the graphical interface is adjusted according to the stress value of the voxel. For example, higher stress values ​​can be mapped to red and lower stress values ​​can be mapped to blue. Such color coding helps users to intuitively identify and analyze stress concentration areas. The visual overlap and transparency between voxels are also calculated during the rendering process to ensure that the depth and spatial sense of the three-dimensional model are correctly displayed. The generated unoptimized stress visualization provides the basis for the final adjustment;

[0117] The last step focuses on adjusting the unoptimized stress visualization. During the adjustment process, the image resolution is adjusted according to the display characteristics of the output device. For example, for high-resolution displays, the image resolution can be increased to ensure the clarity of image details. The color saturation and contrast can be adjusted according to different output requirements so that the image can maintain the best effect on different display devices. For example, the contrast is increased when viewing in a dark environment to better observe the details. After completing these adjustments, the optimized stress visualization is output.

[0118] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. An intelligent identification method for osteoporosis area, characterized in that: The following steps are involved: Extract edge sharpness and texture uniformity parameters from the two-dimensional image, adjust the grayscale threshold of the image, segment the osteoporosis candidate area, obtain segmentation result data, use the intensity difference between pixels in the segmentation result data to analyze the connectivity of the pixels, identify and isolate the foreground and background by setting the connection threshold, and generate foreground isolation data; Based on the foreground isolation data, density and accuracy detection are performed, bone tissue structure feature analysis and bone density analysis are performed on the foreground isolation data to generate a bone density result; Convert each pixel of the two-dimensional image into a three-dimensional voxel, record the conversion parameters and voxel size, construct a framework of a three-dimensional model, apply volume and coordinate data, calculate the structural stability and stress point distribution of the three-dimensional model, generate stress analysis data, perform bone microstructure integrity detection based on the stress analysis data, and generate microstructure integrity data; The stress value and distribution coordinates are extracted from the stress analysis data, the stress area in the three-dimensional model is drawn, and the resolution and color depth are optimized to generate a stress visualization map.

2. The method for intelligently identifying osteoporosis areas according to claim 1, characterized in that: The steps for obtaining the segmentation result data are as follows: Extracting edge sharpness parameters and texture uniformity parameters from the two-dimensional image to obtain edge sharpness data and texture uniformity data; Based on the edge sharpness data and the texture uniformity data, adjusting the grayscale threshold of the image to obtain adjusted image data; Based on the adjusted image data, image segmentation is performed to identify and segment the osteoporosis candidate area to obtain segmentation result data.

3. The method for intelligently identifying osteoporosis areas according to claim 1, characterized in that: The steps for obtaining the foreground isolation data are: Using the intensity difference between the pixels in the segmentation result data, the intensity difference between each pair of adjacent pixels is calculated to obtain intensity difference data; Based on the intensity difference data, the formula is used: Calculate the connectivity of each pair of pixels (i, j); where C(i, j) represents the connectivity between pixels i and j, I ik and I jk Represent the intensity of pixel i and j in the kth color channel respectively, where K is the number of color channels of the image; Based on the connectivity, a connection threshold T is set, and each pair of pixels is judged. If C(i, j)>T, the pixel (i, j) is considered to be disconnected, and the foreground and background are identified and isolated to obtain foreground isolation data.

4. The method for intelligently identifying osteoporosis areas according to claim 1, characterized in that: The steps for obtaining the bone density result are: Extracting the bone tissue area from the foreground isolation data, calculating the total number of pixels in the area, and obtaining bone tissue area data; Based on the bone tissue area data, the average density of the bone tissue area is calculated using the following formula: Among them, ρ is the average density, N is the total number of pixels in the bone tissue area, and D i Represents the density value of the i-th pixel; Based on the average density, the structural characteristics of the bone tissue are analyzed, the uniformity and integrity of the bone tissue are verified, and a bone density result is generated.

5. The method for intelligently identifying osteoporosis areas according to claim 1, characterized in that: The steps for obtaining the stress analysis data are as follows: Convert each pixel of the two-dimensional image into a three-dimensional voxel, record the conversion parameter and voxel size of each voxel, and obtain voxel parameter data; Based on the voxel parameter data, the framework of the three-dimensional model is constructed and the total volume of the model is calculated. The calculation formula is: Where S is the total volume of the model, M is the total number of voxels, and V i represents the volume of the ith voxel, C i represents the volume contribution of the i-th voxel; The total volume of the model is used to analyze the structural stability and stress point distribution of the three-dimensional model, calculate the stress contribution of each voxel, and generate stress analysis data.

6. The method for intelligently identifying osteoporosis areas according to claim 1, characterized in that: The steps for obtaining the microstructure integrity data are: Based on the stress analysis data, analyzing the behavior of each voxel under stress, calculating the stress influence factor of each voxel, and obtaining stress influence data; Based on the stress impact data, the integrity index of the bone microstructure is calculated using the following formula: Where W is the completeness index, Z represents the total number of voxels, and P j represents the stress influence factor of the jth voxel, T j represents the theoretical stress threshold of the jth voxel; Based on the integrity index, the integrity of the bone microstructure is judged. If the index is lower than a threshold, it is considered that the microstructure is damaged, and microstructure integrity data is generated.

7. The method for intelligently identifying osteoporosis areas according to claim 1, characterized in that: The steps for obtaining the stress visualization diagram are as follows: Extracting the stress value and three-dimensional coordinates of each voxel from the stress analysis data to obtain stress value and coordinate data; Using the stress values ​​and coordinate data, the stress region of each voxel is drawn on a graphical interface through three-dimensional graphics rendering to obtain an unoptimized stress visualization map; The resolution and color depth of the unoptimized stress visualization image are adjusted to generate an optimized stress visualization image.

8. An intelligent identification system according to the method for intelligent identification of osteoporosis areas according to any one of claims 1 to 7, characterized in that: include: The image preprocessing module extracts edge sharpness and texture uniformity parameters from the two-dimensional image, adjusts the grayscale threshold of the image, segments the osteoporosis candidate area through grayscale adjustment, analyzes the connectivity of pixels using the intensity difference between pixels, sets the connectivity threshold to identify and isolate the foreground and background, and generates foreground isolation data; A bone density analysis module performs density detection and bone tissue structure feature analysis based on foreground isolation data, analyzes bone density and calculates bone tissue structure features, and generates bone density results and bone tissue structure feature data; The three-dimensional modeling and stress analysis module converts each pixel of the two-dimensional image into a three-dimensional voxel, records the conversion parameters and voxel size, builds the framework of the three-dimensional model, and based on the bone density results and bone tissue structure characteristic data, uses volume and coordinate data to calculate the structural stability and stress point distribution of the three-dimensional model, generates stress analysis data, extracts stress values ​​and distribution coordinates from the stress analysis data, draws the stress area in the three-dimensional model, and generates a stress visualization diagram.

Images