A method and system for evaluating lumbar disc herniation and degeneration based on CT images

Abstract

The present invention discloses a lumbar disc herniation degeneration assessment method and system based on CT images, the method comprising: according to lumbar CT image data, projecting the image data of the three-dimensional lumbar disc region to a multi-view plane to generate a local feature map containing the local structure of the lumbar disc and a global feature map containing the macroscopic structure of the lumbar disc; extracting the structural features of the lumbar disc annulus, nucleus pulposus and protrusion from the local feature maps and global feature maps of different view angles to reconstruct the anatomical structure of the lumbar disc in three-dimensional space to obtain a geometric deformation model of the lumbar disc; according to the geometric deformation model, constructing a multi-level assessment model of lumbar disc degeneration, wherein the multi-level assessment model is trained by a deep learning network to perform feature classification and quantitative scoring to assess the protrusion degeneration level. By using the embodiments of the present invention, the degeneration state of the lumbar disc can be comprehensively and accurately analyzed through the combination of three-dimensional reconstruction and deep learning technology.

Description

Technical Field

[0001] The present invention belongs to the technical field of image processing, and in particular to a method and system for evaluating lumbar disc herniation and degeneration based on CT images. Background Art

[0002] The lumbar intervertebral disc is an important part of the spine, consisting of the annulus fibrosus and the nucleus pulposus, which plays a role in cushioning and supporting. With the influence of factors such as aging, genetic factors, trauma, long-term improper posture or excessive strain, the lumbar intervertebral disc gradually degenerates, which may eventually lead to lumbar disc herniation. Lumbar disc herniation not only causes pain and limited mobility in patients, but in severe cases it may also compress the nerve roots, causing a series of neurological symptoms such as radiating pain, numbness and muscle weakness in the lower limbs.

[0003] At present, the evaluation of lumbar disc herniation and degeneration mostly relies on the observation of clinical symptoms and physical examination, or the use of imaging examination methods, such as MRI (magnetic resonance imaging) and CT (computed tomography). MRI has a high soft tissue contrast and is usually used to evaluate the status of the lumbar disc, but in some cases, CT images are clearer in displaying bone and calcified structures and can provide a more intuitive boundary between bone and soft tissue. However, existing image analysis methods mostly focus on traditional two-dimensional slice observations and lack a comprehensive understanding of the lumbar disc structure in three-dimensional space. This deficiency makes it impossible to fully explore the potential features and changes hidden in the image data in the degeneration assessment, affecting the accurate judgment of the degree of lesions and reducing the effectiveness of early diagnosis and personalized treatment. Summary of the invention

[0004] The purpose of the present invention is to provide a lumbar disc herniation degeneration assessment method and system based on CT images to address the deficiencies in the prior art and to comprehensively and accurately analyze the degeneration state of the lumbar disc through the combination of three-dimensional reconstruction and deep learning technology.

[0005] One embodiment of the present application provides a method for evaluating lumbar disc herniation and degeneration based on CT images, the method comprising:

[0006] According to the lumbar CT image data, the image data of the three-dimensional lumbar intervertebral disc area is projected onto a multi-view plane to generate a local feature map including the local structure of the lumbar intervertebral disc and a global feature map including the macroscopic structure of the lumbar intervertebral disc;

[0007] The structural features of the lumbar disc annulus fibrosus, nucleus pulposus and protrusion are extracted from the local feature maps and global feature maps of different perspectives to reconstruct the anatomical structure of the lumbar disc in three-dimensional space and obtain a geometric deformation model of the lumbar disc.

[0008] Based on the geometric deformation model, a multi-level evaluation model of lumbar disc degeneration is constructed. The multi-level evaluation model is trained through a deep learning network to perform feature classification and quantitative scoring to evaluate the level of prominent degeneration.

[0009] Optionally, the three-dimensional image data of the lumbar intervertebral disc region is projected onto a multi-view plane based on the lumbar CT image data to generate a local feature map including the local structure of the lumbar intervertebral disc and a global feature map including the macroscopic structure of the lumbar intervertebral disc, including:

[0010] Perform three-dimensional reconstruction based on lumbar CT image data to build a three-dimensional lumbar intervertebral disc model;

[0011] The three-dimensional lumbar intervertebral disc model is projected onto a preset multi-view plane through a projection algorithm, and each view contains information in three dimensions;

[0012] In each perspective, a local feature map containing specific key structures of the lumbar intervertebral disc is extracted, and in the same perspective, a corresponding global feature map is generated to capture the macroscopic structure of the entire lumbar intervertebral disc, ensuring that all relevant anatomical structures are included.

[0013] Optionally, the extracting structural features of the annulus fibrosus, nucleus pulposus and protrusion of the lumbar intervertebral disc from the local feature maps and the global feature maps at different viewing angles to reconstruct the anatomical structure of the lumbar intervertebral disc in three-dimensional space and obtain a geometric deformation model of the lumbar intervertebral disc includes:

[0014] The image segmentation algorithm is applied to process the local feature map to segment the annulus fibrosus, nucleus pulposus and protrusion structure of the lumbar intervertebral disc from the background, and morphological operations are used to clarify the boundaries of different segmented structures in the global feature map.

[0015] Extracting key structural features from the different segmented structures after being clarified, wherein the key structural features include: the thickness and shape of the annulus fibrosus, the volume and shape of the nucleus pulposus, and the size and position of the protrusion;

[0016] The extracted key structural features are mapped into three-dimensional space to construct a three-dimensional anatomical structure model of the lumbar intervertebral disc. The extracted feature points are smoothed and connected using an interpolation method to form a complete geometric shape, and the appearance of the lumbar intervertebral disc in different degenerative states is reconstructed. The geometric shape of the three-dimensional anatomical structure model is corrected using a numerical optimization algorithm to ensure the consistency of the model with the original CT image data, and the reconstructed three-dimensional anatomical structure model is obtained.

[0017] Perform geometric deformation analysis on the reconstructed 3D anatomical structure model, identify the shape change and degree of degeneration, and extract geometric deformation features;

[0018] The geometric deformation features are integrated into the geometric deformation model to represent the structural changes of the lumbar intervertebral disc under different degenerative states, and the geometric deformation model can reflect the relative position of each structure and its degree of change.

[0019] Optionally, constructing a multi-level evaluation model for lumbar disc degeneration based on the geometric deformation model includes:

[0020] On the basis of the geometric deformation model, the tensor field of the annulus fibrosus and nucleus pulposus is constructed by high-order tensor analysis to provide deformation information of each region under different stress levels. Based on the tensor field, the local deformation characteristics of the lumbar intervertebral disc are analyzed through the regional deformation tensor, and the herniated displacement of the nucleus pulposus, the stress change of the annulus fibrosus, and the tension distribution between the nucleus pulposus and the intervertebral space are quantified to generate a multidimensional tensor feature map.

[0021] According to the multi-dimensional tensor feature map and the stress and deformation information it contains, the boundary area between the nucleus pulposus and the annulus fibrosus is segmented by adaptive texture features to obtain the pathological texture feature distribution of the lumbar intervertebral disc tissue;

[0022] A multi-level evaluation model for lumbar disc degeneration is constructed by using pathological texture features. The multi-level evaluation model is trained through a deep learning network to perform feature classification and quantitative scoring to evaluate the level of prominent degeneration.

[0023] Another embodiment of the present application provides a lumbar disc herniation and degeneration assessment system based on CT images, the system comprising:

[0024] A projection module is used to project the image data of the three-dimensional lumbar intervertebral disc region onto a multi-view plane based on the lumbar CT image data, so as to generate a local feature map including the local structure of the lumbar intervertebral disc and a global feature map including the macroscopic structure of the lumbar intervertebral disc;

[0025] An extraction module is used to extract the structural features of the annulus fibrosus, nucleus pulposus and protrusion of the lumbar intervertebral disc from the local feature maps and global feature maps of different viewing angles, so as to reconstruct the anatomical structure of the lumbar intervertebral disc in three-dimensional space and obtain a geometric deformation model of the lumbar intervertebral disc;

[0026] The evaluation module is used to construct a multi-level evaluation model for lumbar disc degeneration based on the geometric deformation model, wherein the multi-level evaluation model is trained through a deep learning network to perform feature classification and quantitative scoring to evaluate the level of prominent degeneration.

[0027] Yet another embodiment of the present application provides a storage medium, wherein the storage medium stores a computer program, wherein the computer program is configured to execute any of the above methods when running.

[0028] Yet another embodiment of the present application provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute any of the methods described above.

[0029] Compared with the prior art, the present invention provides a lumbar disc herniation degeneration assessment method based on CT images. According to the lumbar CT image data, the image data of the three-dimensional lumbar disc area is projected onto a multi-view plane to generate a local feature map containing the local structure of the lumbar disc and a global feature map containing the macroscopic structure of the lumbar disc; from the local feature maps and the global feature maps of different viewpoints, the structural features of the lumbar disc annulus fibrosus, nucleus pulposus and protrusion are extracted to reconstruct the anatomical structure of the lumbar disc in three-dimensional space and obtain a geometric deformation model of the lumbar disc; according to the geometric deformation model, a multi-level assessment model of lumbar disc degeneration is constructed, wherein the multi-level assessment model is trained by a deep learning network to perform feature classification and quantitative scoring to assess the protrusion degeneration level, so that the degeneration state of the lumbar disc can be comprehensively and accurately analyzed through the combination of three-dimensional reconstruction and deep learning technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 A hardware structure block diagram of a computer terminal for a method for evaluating lumbar disc herniation and degeneration based on CT images provided in an embodiment of the present invention;

[0031] Figure 2 A flow chart of a method for evaluating lumbar disc herniation and degeneration based on CT images provided in an embodiment of the present invention;

[0032] Figure 3 A schematic structural diagram of a lumbar disc herniation and degeneration assessment system based on CT images provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0033] The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, but should not be construed as limiting the present invention.

[0034] The embodiment of the present invention firstly provides a method for evaluating lumbar disc herniation and degeneration based on CT images. The method can be applied to electronic devices, such as computer terminals, specifically ordinary computers, etc.

[0035] The following describes it in detail by taking running on a computer terminal as an example. Figure 1 The hardware structure block diagram of a computer terminal for a method for evaluating lumbar disc herniation and degeneration based on CT images provided by an embodiment of the present invention. Figure 1 As shown, the computer device includes a processor, a memory, and a network interface connected via a system bus, wherein the memory may include a non-volatile storage medium and an internal memory.

[0036] The non-volatile storage medium can store an operating system and a computer program. The computer program includes program instructions, and when the program instructions are executed, the processor can execute any lumbar disc herniation degeneration assessment method based on CT images.

[0037] The processor is used to provide computing and control capabilities and support the operation of the entire computer equipment.

[0038] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor, the processor can execute any CT image-based lumbar disc herniation degeneration assessment method.

[0039] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art will understand that Figure 1 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0040] It should be understood that the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0041] See also Figure 2 The embodiment of the present invention provides a method for evaluating lumbar disc herniation and degeneration based on CT images, which may include the following steps:

[0042] S201, projecting the image data of the three-dimensional lumbar intervertebral disc region onto a multi-view plane based on the lumbar CT image data, so as to generate a local feature map including the local structure of the lumbar intervertebral disc and a global feature map including the macroscopic structure of the lumbar intervertebral disc;

[0043] Based on the lumbar CT image data, this method projects the image data of the three-dimensional lumbar disc area onto multiple viewing planes, thereby generating a local feature map containing the local structure of the lumbar disc and a global feature map reflecting the macroscopic structure. Through this projection technology, we can obtain rich structural information from different perspectives, covering both subtle local features, such as the specific shape and boundaries of the annulus fibrosus and nucleus pulposus, and showing the overall anatomical structure, which lays a solid foundation for subsequent feature extraction and three-dimensional reconstruction.

[0044] The significance of this step is to make full use of the advantages of high resolution and multi-angle observation of CT images to form a three-dimensional view, which can more comprehensively present the anatomical characteristics of the lumbar disc. This not only helps to improve the accuracy of subsequent algorithms in feature extraction and model reconstruction, but also provides rich information support for evaluating the degenerative state of the lumbar disc, thereby promoting the scientific and personalized clinical diagnosis and treatment.

[0045] Specifically, three-dimensional reconstruction can be performed based on lumbar CT image data to construct a three-dimensional lumbar intervertebral disc model;

[0046] In this step, the lumbar CT image data is first collected and processed, and the two-dimensional slice data is integrated into a coherent three-dimensional model through image reconstruction algorithms (such as filter-based reconstruction technology). This model can truly reflect the spatial distribution and structural characteristics of the lumbar intervertebral disc, forming a complete three-dimensional image data set to facilitate subsequent analysis and processing. Three-dimensional reconstruction is the basis for obtaining lumbar intervertebral disc morphological information, which can effectively alleviate the limitations of traditional two-dimensional image analysis, provide the required spatial information for more in-depth structural analysis and degeneration assessment, and ensure the accuracy and reliability of the model.

[0047] When performing 3D reconstruction, you first need to collect lumbar spine CT image data, which are usually stored in DICOM format. Use professional image processing software (such as 3D Slicer or OsiriX) to load the CT image data set and perform preprocessing, including noise reduction, image enhancement, and contrast adjustment. These steps help improve the quality of the image and ensure the accuracy of the reconstructed model.

[0048] Next, use the 3D reconstruction function in these software to select a suitable reconstruction algorithm, such as an interpolation-based filter, to process multiple slice data. This process involves analyzing the pixel values ​​of each slice and filling the gaps between slices through interpolation to form continuous volume data. Users can adjust the thickness of the slices and the spatial resolution of the reconstruction to optimize the details and visualization of the final model.

[0049] After reconstruction, the 3D model is exported to a visualization format (such as OBJ or STL) for subsequent operations. When the 3D model is completed, further detail adjustment and optimization are performed through 3D visualization software (such as Blender or MeshLab) to clearly display the anatomical structure of the lumbar intervertebral disc and lay the foundation for subsequent analysis.

[0050] The three-dimensional lumbar intervertebral disc model is projected onto a preset multi-view plane through a projection algorithm, and each view contains information in three dimensions;

[0051] This step involves projecting the constructed three-dimensional lumbar disc model from multiple preset perspectives (e.g., front view, side view, and oblique view) through projection algorithms in computer graphics (such as perspective projection or orthogonal projection, or maximum intensity projection, average intensity projection, etc.). This projection process ensures that each perspective can capture the three-dimensional information of the model and fully present the complex structure of the lumbar disc. By projecting the three-dimensional model onto multiple planes, image data from different perspectives can be generated to enhance the comprehensiveness of information acquisition. This process provides multi-angle information support for the subsequent extraction of local and global features, which helps to more accurately analyze and reconstruct the anatomical structure of the lumbar disc.

[0052] After completing the 3D reconstruction, the next step is to project the 3D lumbar disc model onto multiple preset 2D viewing planes. This process uses projection algorithms in computer graphics to create images based on the set viewing angle and projection type (such as perspective projection or orthogonal projection). In this step, you first need to define the position, direction, and projection matrix of the viewing angle to determine how to convert the 3D coordinate system into a 2D plane.

[0053] In specific implementation, OpenGL or similar 3D graphics engines can be used to set the viewport and perspective parameters of the model to ensure that each perspective correctly reflects all the details of the lumbar intervertebral disc. By properly configuring camera parameters (such as field of view angle, near plane and far plane), different observation perspectives can be obtained to provide more comprehensive information.

[0054] After the projection is completed, the generated 2D images will be rendered according to the specific settings of each perspective and output as image files. These images will serve as the basis for subsequent feature extraction. When analyzed, they can present different structural details and help extract local and global features later.

[0055] In each perspective, a local feature map containing specific key structures of the lumbar intervertebral disc is extracted, and in the same perspective, a corresponding global feature map is generated to capture the macroscopic structure of the entire lumbar intervertebral disc, ensuring that all relevant anatomical structures are included.

[0056] In this step, image processing techniques (such as edge detection and region segmentation) are used to identify and extract key anatomical features of the lumbar disc, such as the boundary information of the annulus fibrosus and nucleus pulposus, from the projection images generated from each perspective. At the same time, a corresponding global feature map is generated to capture the macroscopic structure of the entire lumbar disc, ensuring that all relevant anatomical structures are taken into account. Through this step, the local and global features of the lumbar disc can be effectively separated and identified, providing a clear structural basis for subsequent geometric reconstruction. In addition, the integration of local and global features helps to improve the accuracy of the overall model and ensure that the degenerative state of the lumbar disc can be fully reflected in subsequent analysis.

[0057] In this step, image processing technology is used to extract local and global feature maps for each projection image generated from different perspectives. First, key structures of the lumbar disc, such as the annulus fibrosus and nucleus pulposus, are identified through image segmentation techniques (such as threshold segmentation, region growing, or deep learning-based segmentation algorithms). In this process, local feature maps are constructed based on specific pixel intensity and shape features to ensure that clear structural boundaries are extracted.

[0058] Next, in order to obtain the global feature map of the macro structure, the same image processing process is used to analyze the global information under the same perspective. At this point, more complex image processing algorithms may need to be used, such as edge detection algorithms (such as Canny) and template matching techniques to accurately capture the overall morphology of the lumbar intervertebral disc and its interrelationships. This provides comprehensive anatomical information for subsequent model analysis.

[0059] Finally, the local feature map is synthesized with the global feature map to ensure that no important anatomical structure is missed during the analysis. Through image fusion technology, these two types of information are integrated to form a comprehensive feature map, which lays a solid foundation for subsequent three-dimensional reconstruction and geometric deformation analysis. Through such steps, a comprehensive evaluation of the lumbar intervertebral disc can be achieved, providing necessary support for the analysis of the degree of degeneration.

[0060] S202, extracting structural features of the annulus fibrosus, nucleus pulposus and protrusion of the lumbar intervertebral disc from the local feature maps and the global feature maps at different viewing angles, so as to reconstruct the anatomical structure of the lumbar intervertebral disc in three-dimensional space and obtain a geometric deformation model of the lumbar intervertebral disc;

[0061] In this method, by extracting the structural features of the lumbar disc from the local feature maps and global feature maps at different perspectives, we can accurately analyze the key anatomical structures of the lumbar disc, including the annulus fibrosus, nucleus pulposus, and protrusions. This process integrates the information from different perspectives in the image data and provides the necessary structural basis for subsequent three-dimensional reconstruction. The extracted features can reflect the morphology, position, and size of the structure, laying a data foundation for understanding the degenerative state of the lumbar disc.

[0062] By extracting the structural features of the lumbar intervertebral disc, detailed anatomical information of the structure in three-dimensional space can be obtained, thus providing the necessary data support for building a geometric deformation model. This model not only helps to understand the morphological changes of the lumbar intervertebral disc, but also reflects its biomechanical properties under different degenerative states, which has important practical significance for clinical treatment and the formulation of personalized medical plans.

[0063] Specifically, an image segmentation algorithm can be applied to process the local feature map to segment the annulus fibrosus, nucleus pulposus and protrusion structure of the lumbar intervertebral disc from the background, and in the global feature map, morphological operations can be used to clarify the boundaries of different segmented structures;

[0064] In this step, the application of image segmentation algorithms can effectively identify and extract the key structures of the lumbar disc in the image. By using techniques such as threshold segmentation and region growing, the annulus fibrosus, nucleus pulposus and protrusion can be accurately separated from the complex background. Subsequently, morphological operations such as expansion and erosion are used to clarify the boundaries, making the separation between different structures more obvious, creating conditions for subsequent feature extraction. Through efficient image segmentation and boundary clarity, the details of the anatomical structure can be accurately extracted, providing a reliable basis for subsequent reconstruction. This is crucial for studying the degeneration mechanism and biomechanical properties of the lumbar disc, and is helpful for the diagnosis and formulation of treatment plans.

[0065] In this step, the local feature map needs to be preprocessed first to improve the image quality and lay the foundation for subsequent image segmentation. Gaussian filtering or median filtering can be used to remove noise and enhance image contrast to highlight different structures. Then, an appropriate image segmentation algorithm is selected. Common methods include threshold-based Otsu segmentation or semantic segmentation using deep learning models such as U-Net, both of which can effectively separate the structures of the annulus fibrosus, nucleus pulposus, and herniation. When performing segmentation, it is necessary to set a reasonable threshold or train the model so that it can accurately identify the area of ​​interest.

[0066] After segmentation, the resulting structures often have irregular edges and possible artifacts, so morphological operations are needed to further clarify the boundaries of different segmented structures. Specifically, morphological techniques such as opening and closing operations can be used to eliminate small noise points and smooth the edges of structures. Opening operations can be used to remove small isolated areas, while closing operations can fill small holes, making the segmented areas smoother and more coherent. Ultimately, after these processing steps, the boundaries of important structures in the local feature map will be clearer, laying a solid foundation for subsequent feature extraction.

[0067] Extracting key structural features from the different segmented structures after being clarified, wherein the key structural features include: the thickness and shape of the annulus fibrosus, the volume and shape of the nucleus pulposus, and the size and position of the protrusion;

[0068] In this step, the extraction of key structural features is the core of analyzing the anatomical morphology of the lumbar intervertebral disc. By further analyzing the segmented structures, important information such as the thickness and shape of the annulus fibrosus, the volume and morphology of the nucleus pulposus, and the size and position of the protrusion can be obtained. This process involves geometric analysis of the structure, and quantification using geometric features such as boundary points and centroids to ensure accurate feature extraction. The extracted key structural features can provide the necessary parameters for the geometric deformation analysis of the lumbar intervertebral disc. This information can not only help identify biomechanical changes under different degenerative states, but also provide an important quantitative basis for subsequent model construction, which helps to better understand the degeneration mechanism of the lumbar intervertebral disc.

[0069] At this stage, the first step is to extract features from the clarified segmentation results. The specific operation includes using contour detection technology to obtain the boundaries of each structure. The Canny edge detection algorithm can be used to quickly find the boundaries of the annulus fibrosus, nucleus pulposus, and protrusions, and combine them with the original segmentation results to obtain specific geometric features through regional attribute analysis. For example, the thickness of the annulus fibrosus can be obtained by calculating the shortest distance between any two points on its boundary, and its shape features, such as area, perimeter, and shape index, can be calculated. These indicators can quantify the geometric characteristics of the annulus fibrosus and reflect its stability and health.

[0070] For the feature extraction of the nucleus pulposus, the focus is on its volume and morphological characteristics. Through three-dimensional volume calculation, the segmented area of ​​the nucleus pulposus can be regarded as a three-dimensional shape and its volume can be calculated. When further analyzing the morphology of the nucleus pulposus, important parameters such as aspect ratio and surface smoothness can be extracted. These characteristic indicators can reflect the degeneration of the nucleus pulposus. The analysis of the protrusion focuses on its size and position. By calculating the height, width and coordinates of the protrusion in three-dimensional space, the severity of the protrusion can be effectively evaluated. The extraction of these features not only provides data support for subsequent geometric modeling, but is also an important basis for clinical diagnosis.

[0071] Finally, all the extracted key structural features need to be organized into a structured data form to facilitate subsequent analysis and modeling. This can be done by establishing a database to record each structural feature in a table, with the numerical value and descriptive statistical information of each feature attached, for subsequent analysis. When the model building phase begins, these extracted features will be directly input into the corresponding algorithm to provide sufficient basic data for subsequent 3D reconstruction and geometric deformation analysis. At the same time, the integration of these data will also help improve the accuracy and effectiveness of subsequent models.

[0072] The extracted key structural features are mapped into three-dimensional space to construct a three-dimensional anatomical structure model of the lumbar intervertebral disc. The extracted feature points are smoothed and connected using an interpolation method to form a complete geometric shape, and the appearance of the lumbar intervertebral disc in different degenerative states is reconstructed. The geometric shape of the three-dimensional anatomical structure model is corrected using a numerical optimization algorithm to ensure the consistency of the model with the original CT image data, and the reconstructed three-dimensional anatomical structure model is obtained.

[0073] In this step, the extracted key structural features are mapped into three-dimensional space to construct a three-dimensional anatomical model of the lumbar intervertebral disc. The gaps between the feature points are smoothly connected by interpolation methods to form a complete model while ensuring geometric continuity. In addition, numerical optimization algorithms (such as the least squares method) are used to adjust the model so that it is highly consistent with the original CT image data to ensure the biological authenticity and physical availability of the final model. By constructing an accurate three-dimensional anatomical structure model, basic data can be provided for subsequent geometric deformation analysis and biomechanical simulation. This model can not only well reflect the actual appearance of the lumbar intervertebral disc, but also provide an important reference for clinical evaluation and the design of personalized treatment plans, which helps to achieve more effective pathological analysis and intervention measures.

[0074] The first step of this step is to perform three-dimensional mapping of the extracted key structural features. By defining the position coordinates of each key feature, the anatomical structure model of the lumbar intervertebral disc is reconstructed in the three-dimensional coordinate system. During the mapping process, the accuracy of the relative position of each feature point needs to be considered to ensure that it can reflect the true anatomical features. In this process, computer graphics technology can be used to form preliminary geometric shapes by defining the connection relationship between points. In the process of initially constructing a three-dimensional model, methods such as Delaunay triangulation can be used to ensure the smoothness of the connection and the accuracy of the model.

[0075] Subsequently, in order to make the three-dimensional model smoother and more natural, interpolation methods are used to further process the connections between feature points. Interpolation methods can include bilinear interpolation or spline interpolation, which can effectively generate intermediate points to fill the gaps between feature points and form continuous geometric shapes. In this process, the interpolation algorithm not only improves the smoothness of the model, but also better reflects the morphological changes of the lumbar intervertebral disc under different degenerative states. The visualization effect of the model is also significantly improved, making subsequent analysis and observation more intuitive.

[0076] To ensure the accuracy of the model, the last step is to correct the geometric shape through a numerical optimization algorithm. By comparing with the original CT image data, the difference between the model and the image data is calculated, and the model parameters are adjusted using the least squares method or other optimization algorithms. This process aims to eliminate possible shape deviations in the model to ensure that it accurately reflects the actual anatomical structure. After numerical optimization, the resulting three-dimensional anatomical structure model can not only provide effective support for clinical research, but also lay an important foundation for subsequent geometric deformation analysis, thereby improving the reliability and effectiveness of the research.

[0077] Perform geometric deformation analysis on the reconstructed 3D anatomical structure model, identify the shape change and degree of degeneration, and extract geometric deformation features;

[0078] In this step, the reconstructed 3D anatomical model will be used for geometric deformation analysis. By comparing the models under different degenerative states, we can identify the shape changes and degree of degeneration of the lumbar disc. This analysis involves not only the quantitative calculation of the geometric features of the model, but also comparative analysis to clarify the structural differences between different degenerative stages. Geometric deformation analysis can reveal the laws and mechanisms of lumbar disc degeneration and provide a scientific basis for clinical intervention. By identifying shape changes, doctors can make more accurate diagnoses and treatment plans, thereby improving patients' medical outcomes and quality of life.

[0079] The first task of geometric deformation analysis is to perform detailed shape recognition on the reconstructed 3D anatomical structure model. This process involves various measurements and analyses of the geometric properties of the model, focusing on its shape change, volume reduction, and thickness difference. Finite element analysis (FEA) can be used to simulate the behavior of the model under various stress conditions. By applying biomechanical stress to the model and observing its response under different loads and stresses, key features of the model during deformation can be extracted.

[0080] During the dynamic analysis process, the relevant deformations of each structure in the model and their degree of degeneration can be identified. This can be achieved by calculating local strains, displacements, and geometric indicators. For example, the thickness changes of the annulus fibrosus, the protrusion of the nucleus pulposus, and the morphological changes of the protrusion can be analyzed in detail to form a series of degeneration assessment indicators. The quantification of these characteristics not only provides data support for the evaluation, but also provides an important basis for doctors in clinical decision-making. Through these analyses, we can fully understand the biomechanical characteristics of the lumbar intervertebral disc and its changes in different stages of degeneration.

[0081] Finally, the extracted geometric deformation features are integrated into the geometric deformation model to fully represent the structural changes of the lumbar disc under different degenerative states. This not only provides specific geometric deformation information, but also reflects the relative position of each structure and its degree of change. The results of this comprehensive analysis can provide an important basis for subsequent clinical treatment and research, and provide clear guidance for doctors. At the same time, the establishment of the geometric deformation model also provides a good theoretical basis for future related research and promotes in-depth research on lumbar disc degeneration.

[0082] The geometric deformation features are integrated into the geometric deformation model to represent the structural changes of the lumbar intervertebral disc under different degenerative states, and the geometric deformation model can reflect the relative position of each structure and its degree of change.

[0083] In this final step, the geometric deformation features extracted previously are integrated into the geometric deformation model to form a comprehensive model that is used to represent the structural changes of the lumbar intervertebral disc under different degenerative states. This model not only shows the relative positions of various anatomical structures (such as the annulus fibrosus and the nucleus pulposus), but also can intuitively present the degree of their changes. This process requires the use of geometric modeling and data fusion technology to match and integrate various feature data to generate a complete geometric deformation model that can reflect different degenerative states.

[0084] By integrating geometric deformation features, the geometric deformation model formed can more comprehensively reveal the degeneration mechanism of lumbar disc. This model provides an intuitive tool for clinicians to evaluate, helps to observe and analyze the degree of degeneration in imaging, and provides scientific support for the formulation of subsequent treatment plans and effect evaluation. In addition, the model can also be used for teaching and research to improve the understanding of lumbar disc degeneration.

[0085] The work at this stage focuses on integrating the extracted geometric deformation features into the geometric deformation model to build a complete model that reflects the lumbar disc in different degenerative states. First, the features extracted in the degeneration analysis phase (such as strain, displacement, thickness change, etc.) need to be input into the geometric deformation model in a structured manner, and the relationship between them and the various anatomical structures of the model needs to be defined. This process requires the model to accurately represent the relative position, geometric relationship and stress distribution of each structure in different states.

[0086] To achieve this goal, standardized feature mapping methods can be used, such as principal component analysis (PCA) to reduce the dimensionality of high-dimensional features, simplify the complexity of the model, and make it easier to understand. At the same time, with the help of graphic visualization technology, corresponding visualization interfaces can be created to enable doctors and researchers to intuitively see the changes in lumbar discs under different degenerative states. These visualization results can not only help clinicians better understand pathological changes, but also provide patients with clearer interpretations, thereby enhancing patients' confidence in treatment plans.

[0087] Finally, the integrated geometric deformation model should be predictable. To this end, the model can be verified and adjusted based on the existing degeneration data and the theory of the biomechanical model. Through deep learning or machine learning algorithms, the model can be trained using existing clinical data so that it can make accurate predictions in new clinical scenarios. This model is not only of practical significance and has a guiding role in the evaluation of lumbar disc degeneration, but also lays a good foundation for future research and clinical applications, and improves the scientific nature of lumbar disc health management.

[0088] S203, constructing a multi-level evaluation model for lumbar disc degeneration based on the geometric deformation model, wherein the multi-level evaluation model is trained through a deep learning network to perform feature classification and quantitative scoring to evaluate the level of prominent degeneration.

[0089] It is an effective method to construct a multi-level evaluation model for lumbar disc degeneration based on the geometric deformation model. By deeply analyzing the geometric deformation of the lumbar disc, the differences between different degeneration levels can be identified. This process uses a deep learning network to train the extracted features to establish a complex evaluation model. The model can automatically identify and classify the degeneration features of the lumbar disc and achieve quantitative scoring. This evaluation not only improves the accuracy of diagnosis, but also provides a scientific basis for the formulation of clinical treatment plans.

[0090] The construction of a multi-level evaluation model has important clinical application value and academic research significance. First, it provides a quantitative evaluation system that helps clinicians develop more accurate treatment strategies for patients with lumbar disc herniation. Second, the accuracy and reliability of this model can be improved through continuous deep learning training, thereby promoting the correlation between imaging analysis and pathological characteristics, and promoting a deeper understanding of the pathological mechanism of degeneration in the medical field.

[0091] Specifically, based on the geometric deformation model, high-order tensor analysis can be applied to construct the tensor field of the annulus fibrosus and the nucleus pulposus to provide deformation information about each region under different stresses, and based on the tensor field, the local deformation characteristics of the lumbar intervertebral disc are analyzed through the regional deformation tensor, and the herniated displacement of the nucleus pulposus, the force change of the annulus fibrosus and the tension distribution between the nucleus pulposus and the intervertebral space are quantified to generate a multi-dimensional tensor feature map;

[0092] This step uses high-order tensor analysis technology to construct the tensor field of the annulus fibrosus and nucleus pulposus, aiming to quantify the strain behavior of the lumbar disc under different loading conditions. By analyzing the tensor field of each region, the deformation characteristics of the local tissue under the action of external force can be understood more accurately, thereby providing reliable data for subsequent evaluation. By quantifying the local deformation characteristics, doctors can more clearly understand the extent of the lesion and its impact on the surrounding tissues. This process not only helps to improve the objectivity of diagnosis, but also provides an important basis for the selection of treatment options.

[0093] On the basis of the geometric deformation model, it is first necessary to apply high-order tensor analysis to construct the tensor field of the annulus fibrosus and nucleus pulposus. This process begins with a detailed analysis of the geometric deformation model of the lumbar intervertebral disc. In this process, by selecting an appropriate high-order tensor library, a tensor field describing the structure of the lumbar intervertebral disc is constructed to accurately reflect the deformation information of each region under different stress levels. Specifically, the reference points and significant feature points of the tensor field must first be determined so that the response of each tissue can be clearly recorded when different biomechanical loads are applied. Then, through numerical simulation technology, the strain and displacement data under various load conditions are obtained, and these data are converted into tensor form, thereby forming a multidimensional tensor feature map for the lumbar intervertebral disc.

[0094] After obtaining the tensor field, the next step is to analyze the local deformation characteristics of the lumbar disc based on the tensor field. This process uses the regional deformation tensor method to analyze different regions of the nucleus pulposus and annulus fibrosus one by one to quantify the protrusion displacement of the nucleus pulposus, the force changes of the annulus fibrosus, and the tension distribution between the nucleus pulposus and the intervertebral disc. This analysis not only helps to reveal the stress state of the lumbar disc, but also can accurately describe the relationship between various structures, forming a multi-level evaluation of deformation characteristics at all levels.

[0095] Finally, the above analysis results are integrated to generate a comprehensive multi-dimensional tensor feature map to characterize the geometric deformation characteristics of the lumbar disc under different degenerative states. This feature map will contain the deformation information of the annulus fibrosus and nucleus pulposus and their relative position changes, and intuitively display the dynamic response of the lumbar disc under stress. This tensor feature map will provide a solid foundation for the subsequent multi-level evaluation model construction, ensuring that the subsequent analysis has sufficient details and accuracy.

[0096] According to the multi-dimensional tensor feature map and the stress and deformation information it contains, the boundary area between the nucleus pulposus and the annulus fibrosus is segmented by adaptive texture features to obtain the pathological texture feature distribution of the lumbar intervertebral disc tissue;

[0097] This step aims to locate and segment the boundary area between the nucleus pulposus and the annulus fibrosus through the rich information provided by the multidimensional tensor feature map. The adaptive texture feature segmentation method can more accurately identify the pathological characteristics of local tissues, thereby achieving in-depth analysis of lumbar disc degeneration. This process significantly improves the visualization of tissue pathological changes, provides clinicians with a clearer relationship between anatomy and pathology, and helps improve diagnostic accuracy and the formulation of personalized treatment plans.

[0098] In this step, we first need to analyze the multidimensional tensor feature map in detail to extract key stress and deformation information. The core of this process is to deeply understand the physical properties and texture characteristics of each area in the image. Therefore, we use a variety of image processing algorithms for feature extraction. First, we use local contrast enhancement and other techniques to improve the visualization of the feature map, making the subtle tissue structure more obvious. Next, according to the texture characteristics of the image, we select a suitable adaptive segmentation algorithm, which usually includes threshold-based or clustering methods, so as to accurately segment the boundary area between the nucleus pulposus and the annulus fibrosus from the background area.

[0099] In the process of boundary area segmentation, morphological processing methods are used to further clarify the segmentation results to eliminate noise and artifacts and ensure the accuracy of the segmentation contour. This process not only helps to improve the accuracy of the segmentation boundary, but also enhances the coherence of the segmented features. Through this processing, each boundary area can accurately capture the structural characteristics of the annulus fibrosus and nucleus pulposus.

[0100] Next, the pathological texture features of the lumbar disc are extracted from the clarified segmentation results. These features include the thickness, shape and relative position changes of the annulus fibrosus and nucleus pulposus, which are crucial for understanding the pathological changes of the lumbar disc under degenerative conditions. Finally, the extracted pathological texture features are summarized to form a comprehensive texture feature distribution map, which provides reliable basic data for the subsequent multi-level evaluation model, so that the pathological condition of the lumbar disc can be scientifically and systematically analyzed and evaluated.

[0101] A multi-level evaluation model for lumbar disc degeneration is constructed by using pathological texture features. The multi-level evaluation model is trained through a deep learning network to perform feature classification and quantitative scoring to evaluate the level of prominent degeneration.

[0102] In this step, pathological texture features are used to construct a multi-level evaluation model, which, combined with deep learning technology, can comprehensively evaluate the degenerative state of the lumbar disc. By training the deep learning network, the extracted features can be classified and quantified to reflect different levels of degeneration. This multi-level evaluation model provides a quantitative tool that enables clinicians to develop more effective treatment strategies based on objective data, and also lays the foundation for future related research.

[0103] When building a multi-level evaluation model, the extracted pathological texture features must first be organized and standardized to meet the input requirements of the subsequent deep learning network. Specifically, each feature must be converted into a numerical format and normalized to prevent the quantitative differences between features from affecting the training of the model. At the same time, in order to ensure the diversity and adequacy of the training data, the training set can be expanded through data augmentation techniques such as rotation, flipping, or adding noise to enhance the universality and robustness of the model.

[0104] Next, select a deep learning network architecture suitable for lumbar disc degeneration assessment. Generally, convolutional neural networks (CNNs) are more appropriate because they have good feature extraction capabilities when processing image data. In the network architecture design, the number of network layers and neurons needs to be reasonably configured according to the complexity of the features and the capacity of the model. In addition, appropriate activation functions and loss functions need to be selected to optimize the network learning process. During the model training process, the network weights are adjusted through the back-propagation algorithm to ensure that the features can be effectively classified and identified.

[0105] Finally, the trained deep learning model is used to evaluate and predict new input data. Based on the output of the model, the degeneration level of the lumbar disc can be quantified and scored to form a standardized evaluation report. This score not only reflects the health status of the lumbar disc, but also provides decision support for clinicians. In order to ensure the accuracy of the model, the model can be verified and updated regularly to adapt to new clinical data and research progress, thereby improving the reliability and applicability of the evaluation system.

[0106] It can be seen that according to the lumbar CT image data, the image data of the three-dimensional lumbar intervertebral disc area is projected onto a multi-view plane to generate a local feature map containing the local structure of the lumbar intervertebral disc and a global feature map containing the macroscopic structure of the lumbar intervertebral disc; the structural features of the lumbar intervertebral disc annulus fibrosus, nucleus pulposus and protrusion are extracted from the local feature maps and global feature maps of different viewpoints to reconstruct the anatomical structure of the lumbar intervertebral disc in three-dimensional space and obtain a geometric deformation model of the lumbar intervertebral disc; according to the geometric deformation model, a multi-level evaluation model of lumbar intervertebral disc degeneration is constructed, wherein the multi-level evaluation model is trained by a deep learning network to perform feature classification and quantitative scoring to evaluate the level of protrusion degeneration, so that the degeneration state of the lumbar intervertebral disc can be comprehensively and accurately analyzed through the combination of three-dimensional reconstruction and deep learning technology.

[0107] Another embodiment of the present invention provides a lumbar disc herniation degeneration assessment system based on CT images, see Figure 3 , the system may include:

[0108] The projection module 301 is used to project the image data of the three-dimensional lumbar intervertebral disc region onto a multi-view plane according to the lumbar CT image data, so as to generate a local feature map including the local structure of the lumbar intervertebral disc and a global feature map including the macroscopic structure of the lumbar intervertebral disc;

[0109] An extraction module 302 is used to extract the structural features of the annulus fibrosus, nucleus pulposus and protrusion of the lumbar intervertebral disc from the local feature maps and the global feature maps of different viewing angles, so as to reconstruct the anatomical structure of the lumbar intervertebral disc in three-dimensional space and obtain a geometric deformation model of the lumbar intervertebral disc;

[0110] The evaluation module 303 is used to construct a multi-level evaluation model for lumbar disc degeneration based on the geometric deformation model, wherein the multi-level evaluation model is trained through a deep learning network to perform feature classification and quantitative scoring to evaluate the level of prominent degeneration.

[0111] It can be seen that according to the lumbar CT image data, the image data of the three-dimensional lumbar intervertebral disc area is projected onto a multi-view plane to generate a local feature map containing the local structure of the lumbar intervertebral disc and a global feature map containing the macroscopic structure of the lumbar intervertebral disc; the structural features of the lumbar intervertebral disc annulus fibrosus, nucleus pulposus and protrusion are extracted from the local feature maps and global feature maps of different viewpoints to reconstruct the anatomical structure of the lumbar intervertebral disc in three-dimensional space and obtain a geometric deformation model of the lumbar intervertebral disc; according to the geometric deformation model, a multi-level evaluation model of lumbar intervertebral disc degeneration is constructed, wherein the multi-level evaluation model is trained by a deep learning network to perform feature classification and quantitative scoring to evaluate the level of protrusion degeneration, so that the degeneration state of the lumbar intervertebral disc can be comprehensively and accurately analyzed through the combination of three-dimensional reconstruction and deep learning technology.

[0112] An embodiment of the present invention further provides a storage medium, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above method embodiments when running.

[0113] Specifically, in this embodiment, the above storage medium may be configured to store a computer program for performing the following steps:

[0114] S201, projecting the image data of the three-dimensional lumbar intervertebral disc region onto a multi-view plane based on the lumbar CT image data, so as to generate a local feature map including the local structure of the lumbar intervertebral disc and a global feature map including the macroscopic structure of the lumbar intervertebral disc;

[0115] S202, extracting structural features of the annulus fibrosus, nucleus pulposus and protrusion of the lumbar intervertebral disc from the local feature maps and the global feature maps at different viewing angles, so as to reconstruct the anatomical structure of the lumbar intervertebral disc in three-dimensional space and obtain a geometric deformation model of the lumbar intervertebral disc;

[0116] S203, constructing a multi-level evaluation model for lumbar disc degeneration based on the geometric deformation model, wherein the multi-level evaluation model is trained through a deep learning network to perform feature classification and quantitative scoring to evaluate the level of prominent degeneration.

[0117] It can be seen that according to the lumbar CT image data, the image data of the three-dimensional lumbar intervertebral disc area is projected onto a multi-view plane to generate a local feature map containing the local structure of the lumbar intervertebral disc and a global feature map containing the macroscopic structure of the lumbar intervertebral disc; the structural features of the lumbar intervertebral disc annulus fibrosus, nucleus pulposus and protrusion are extracted from the local feature maps and global feature maps of different viewpoints to reconstruct the anatomical structure of the lumbar intervertebral disc in three-dimensional space and obtain a geometric deformation model of the lumbar intervertebral disc; according to the geometric deformation model, a multi-level evaluation model of lumbar intervertebral disc degeneration is constructed, wherein the multi-level evaluation model is trained by a deep learning network to perform feature classification and quantitative scoring to evaluate the level of protrusion degeneration, so that the degeneration state of the lumbar intervertebral disc can be comprehensively and accurately analyzed through the combination of three-dimensional reconstruction and deep learning technology.

[0118] An embodiment of the present invention further provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0119] Specifically, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0120] Specifically, in this embodiment, the processor may be configured to perform the following steps through a computer program:

[0121] S201, projecting the image data of the three-dimensional lumbar intervertebral disc region onto a multi-view plane based on the lumbar CT image data, so as to generate a local feature map including the local structure of the lumbar intervertebral disc and a global feature map including the macroscopic structure of the lumbar intervertebral disc;

[0122] S202, extracting structural features of the annulus fibrosus, nucleus pulposus and protrusion of the lumbar intervertebral disc from the local feature maps and the global feature maps at different viewing angles, so as to reconstruct the anatomical structure of the lumbar intervertebral disc in three-dimensional space and obtain a geometric deformation model of the lumbar intervertebral disc;

[0123] S203, constructing a multi-level evaluation model for lumbar disc degeneration based on the geometric deformation model, wherein the multi-level evaluation model is trained through a deep learning network to perform feature classification and quantitative scoring to evaluate the level of prominent degeneration.

[0124] It can be seen that according to the lumbar CT image data, the image data of the three-dimensional lumbar intervertebral disc area is projected onto a multi-view plane to generate a local feature map containing the local structure of the lumbar intervertebral disc and a global feature map containing the macroscopic structure of the lumbar intervertebral disc; the structural features of the lumbar intervertebral disc annulus fibrosus, nucleus pulposus and protrusion are extracted from the local feature maps and global feature maps of different viewpoints to reconstruct the anatomical structure of the lumbar intervertebral disc in three-dimensional space and obtain a geometric deformation model of the lumbar intervertebral disc; according to the geometric deformation model, a multi-level evaluation model of lumbar intervertebral disc degeneration is constructed, wherein the multi-level evaluation model is trained by a deep learning network to perform feature classification and quantitative scoring to evaluate the level of protrusion degeneration, so that the degeneration state of the lumbar intervertebral disc can be comprehensively and accurately analyzed through the combination of three-dimensional reconstruction and deep learning technology.

[0125] The above describes in detail the structure, features and effects of the present invention based on the embodiments shown in the drawings. The above is only a preferred embodiment of the present invention, but the present invention is not limited to the scope of implementation shown in the drawings. Any changes made according to the concept of the present invention, or modifications to equivalent embodiments with equivalent changes, which still do not exceed the spirit covered by the description and drawings, should be within the protection scope of the present invention.

Claims

1. A method for evaluating lumbar disc herniation and degeneration based on CT images, characterized in that: The method comprises: According to the lumbar CT image data, the image data of the three-dimensional lumbar intervertebral disc area is projected onto a multi-view plane to generate a local feature map including the local structure of the lumbar intervertebral disc and a global feature map including the macroscopic structure of the lumbar intervertebral disc; The structural features of the fibrous ring, nucleus pulposus and protrusion of the lumbar intervertebral disc are extracted from the local feature maps and the global feature maps of different viewing angles to reconstruct the anatomical structure of the lumbar intervertebral disc in three-dimensional space and obtain a geometric deformation model of the lumbar intervertebral disc; wherein, the local feature map is processed by applying an image segmentation algorithm to segment the fibrous ring, nucleus pulposus and protrusion structure of the lumbar intervertebral disc from the background, and in the global feature map, morphological operations are used to clarify the boundaries of different segmented structures; key structural features are extracted from the clarified different segmented structures, and the key structural features include: the thickness and shape of the fibrous ring, the volume and shape of the nucleus pulposus, and the size and position of the protrusion; The extracted key structural features are mapped into three-dimensional space to construct a three-dimensional anatomical structure model of the lumbar intervertebral disc. The extracted feature points are smoothed and connected using an interpolation method to form a complete geometric shape, and the appearance of the lumbar intervertebral disc in different degenerative states is reconstructed. The geometric shape of the three-dimensional anatomical structure model is corrected using a numerical optimization algorithm to ensure the consistency of the model with the original CT image data, and the reconstructed three-dimensional anatomical structure model is obtained. Perform geometric deformation analysis on the reconstructed three-dimensional anatomical structure model to identify shape changes and the degree of degeneration, and extract geometric deformation features; integrate the geometric deformation features into the geometric deformation model to represent the structural changes of the lumbar intervertebral disc under different degenerative states, and the geometric deformation model can reflect the relative position of each structure and its degree of change; Based on the geometric deformation model, a multi-level evaluation model of lumbar disc degeneration is constructed. The multi-level evaluation model is trained through a deep learning network to perform feature classification and quantitative scoring to evaluate the level of prominent degeneration.

2. The method according to claim 1, characterized in that The method of projecting the image data of the three-dimensional lumbar intervertebral disc region onto a multi-view plane based on the lumbar CT image data to generate a local feature map including the local structure of the lumbar intervertebral disc and a global feature map including the macroscopic structure of the lumbar intervertebral disc includes: Perform three-dimensional reconstruction based on lumbar CT image data to build a three-dimensional lumbar intervertebral disc model; The three-dimensional lumbar intervertebral disc model is projected onto a preset multi-view plane through a projection algorithm, and each view contains information of three dimensions; In each perspective, a local feature map containing specific key structures of the lumbar intervertebral disc is extracted, and in the same perspective, a corresponding global feature map is generated to capture the macroscopic structure of the entire lumbar intervertebral disc, ensuring that all relevant anatomical structures are included.

3. The method according to claim 2, characterized in that The multi-level evaluation model of lumbar disc degeneration is constructed based on the geometric deformation model, including: On the basis of the geometric deformation model, the tensor field of the annulus fibrosus and nucleus pulposus is constructed by high-order tensor analysis to provide deformation information of each region under different stress levels. Based on the tensor field, the local deformation characteristics of the lumbar intervertebral disc are analyzed through the regional deformation tensor, and the herniated displacement of the nucleus pulposus, the stress change of the annulus fibrosus, and the tension distribution between the nucleus pulposus and the intervertebral space are quantified to generate a multidimensional tensor feature map. According to the multi-dimensional tensor feature map and the stress and deformation information it contains, the boundary area between the nucleus pulposus and the annulus fibrosus is segmented by adaptive texture features to obtain the pathological texture feature distribution of the lumbar intervertebral disc tissue; A multi-level evaluation model for lumbar disc degeneration is constructed by using pathological texture features. The multi-level evaluation model is trained through a deep learning network to perform feature classification and quantitative scoring to evaluate the level of prominent degeneration.

4. A lumbar disc herniation degeneration assessment system based on CT images, characterized in that: The system comprises: A projection module is used to project the image data of the three-dimensional lumbar intervertebral disc region onto a multi-view plane based on the lumbar CT image data, so as to generate a local feature map including the local structure of the lumbar intervertebral disc and a global feature map including the macroscopic structure of the lumbar intervertebral disc; The extraction module is used to extract the structural features of the fibrous ring, nucleus pulposus and protrusion of the lumbar intervertebral disc from the local feature maps and the global feature maps of different viewing angles, so as to reconstruct the anatomical structure of the lumbar intervertebral disc in three-dimensional space and obtain a geometric deformation model of the lumbar intervertebral disc; wherein, the image segmentation algorithm is applied to process the local feature map, the fibrous ring, nucleus pulposus and protrusion structure of the lumbar intervertebral disc are segmented from the background, and in the global feature map, morphological operations are used to clarify the boundaries of different segmented structures; key structural features are extracted from the clarified different segmented structures, and the key structural features include: the thickness and shape of the fibrous ring, the volume and shape of the nucleus pulposus, and the size and position of the protrusion; The extracted key structural features are mapped into three-dimensional space to construct a three-dimensional anatomical structure model of the lumbar intervertebral disc. The extracted feature points are smoothed and connected using an interpolation method to form a complete geometric shape, and the appearance of the lumbar intervertebral disc in different degenerative states is reconstructed. The geometric shape of the three-dimensional anatomical structure model is corrected using a numerical optimization algorithm to ensure the consistency of the model with the original CT image data, and the reconstructed three-dimensional anatomical structure model is obtained. Perform geometric deformation analysis on the reconstructed three-dimensional anatomical structure model to identify shape changes and the degree of degeneration, and extract geometric deformation features; integrate the geometric deformation features into the geometric deformation model to represent the structural changes of the lumbar intervertebral disc under different degenerative states, and the geometric deformation model can reflect the relative position of each structure and its degree of change; The evaluation module is used to construct a multi-level evaluation model for lumbar disc degeneration based on the geometric deformation model, wherein the multi-level evaluation model is trained through a deep learning network to perform feature classification and quantitative scoring to evaluate the level of prominent degeneration.

5. The system according to claim 4, characterized in that The projection module is specifically used for: Perform three-dimensional reconstruction based on lumbar CT image data to build a three-dimensional lumbar intervertebral disc model; The three-dimensional lumbar intervertebral disc model is projected onto a preset multi-view plane through a projection algorithm, and each view contains information of three dimensions; In each perspective, a local feature map containing specific key structures of the lumbar intervertebral disc is extracted, and in the same perspective, a corresponding global feature map is generated to capture the macroscopic structure of the entire lumbar intervertebral disc, ensuring that all relevant anatomical structures are included.

6. The system according to claim 5, characterized in that The evaluation module is specifically used for: On the basis of the geometric deformation model, the tensor field of the annulus fibrosus and nucleus pulposus is constructed by high-order tensor analysis to provide deformation information of each region under different stress levels. Based on the tensor field, the local deformation characteristics of the lumbar intervertebral disc are analyzed through the regional deformation tensor, and the herniated displacement of the nucleus pulposus, the stress change of the annulus fibrosus, and the tension distribution between the nucleus pulposus and the intervertebral space are quantified to generate a multidimensional tensor feature map. According to the multi-dimensional tensor feature map and the stress and deformation information it contains, the boundary area between the nucleus pulposus and the annulus fibrosus is segmented by adaptive texture features to obtain the pathological texture feature distribution of the lumbar intervertebral disc tissue; A multi-level evaluation model for lumbar disc degeneration is constructed by using pathological texture features. The multi-level evaluation model is trained through a deep learning network to perform feature classification and quantitative scoring to evaluate the level of prominent degeneration.

7. An electronic device comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to run the computer program to perform the method according to any one of claims 1 to 3.

Images