Scoliosis minimally invasive surgery auxiliary system for identifying vertebral body morphological change based on AI
By using AI to identify minute morphological changes in the vertebral bodies of scoliosis, personalized minimally invasive surgical plans are provided, solving the problems of large trauma and slow recovery in traditional open surgery, and achieving precise minimally invasive correction and improved safety.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- 张嘉庚
- Filing Date
- 2026-03-30
- Publication Date
- 2026-05-12
AI Technical Summary
Existing technologies struggle to accurately identify minute morphological changes in the vertebral bodies during scoliosis, leading to traditional open surgery relying on large steel rods and screws, resulting in significant trauma, slow recovery, and a lack of individualized surgical planning.
By employing AI-based multimodal image input, vertebral body segmentation and 3D reconstruction, identification of subtle morphological changes, comprehensive assessment of scoliosis, and minimally invasive surgery planning and navigation modules, the system achieves accurate identification of subtle vertebral morphological changes and individualized surgical plan design, guiding minimally invasive surgery.
It enables precise filling and correction of asymmetrical vertebral body areas under minimally invasive conditions, reducing surgical trauma, improving the objectivity and consistency of assessment, shortening recovery time, and reducing intraoperative errors and complications.
Smart Images
Figure CN122005084A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the fields of medical devices and artificial intelligence technology, and in particular to a minimally invasive surgical assistance system for scoliosis based on AI recognition of vertebral morphological changes. It is applicable to the planning, navigation and safety assessment of minimally invasive orthopedic surgeries for adolescent idiopathic scoliosis, Hummus' disease and other conditions accompanied by subtle vertebral morphological abnormalities. Background Technology
[0002] The occurrence and progression of scoliosis are often accompanied by subtle morphological differences and asymmetric changes in the vertebral bodies themselves. These subtle structural changes are key determinants of scoliosis progression, biomechanical imbalance, and surgical outcomes. Traditional assessment methods rely on doctors visually reviewing X-rays, which makes it difficult to consistently capture early and subtle morphological changes in the vertebral bodies, easily leading to missed diagnoses or subjective biases.
[0003] Currently, the mainstream surgical procedure for treating scoliosis in clinical practice is traditional open orthopedic surgery, which widely uses large steel rods and screws for fixation and correction. This method involves large incisions, high trauma, and extensive soft tissue damage. Furthermore, surgical planning is highly dependent on the surgeon's experience, making it difficult to design a precise plan based on the subtle individual morphological differences of each patient's vertebrae.
[0004] While there are existing technologies for minimally invasive surgery, there is a general lack of an intelligent auxiliary system that can accurately identify minute morphological changes in the vertebral body. This makes it impossible to downgrade traditional high-trauma open surgery with large steel rods and nails to minimally invasive surgery, and there is also a lack of precise support for minimally invasive vertebral body morphology adjustment. Summary of the Invention
[0005] The main technical problem to be solved by this invention is to accurately identify minute morphological changes in the vertebral body through artificial intelligence, thereby downgrading the traditional high-trauma open surgery that relies on large steel rods and nails to minimally invasive surgery. This enables precise filling, adjustment and correction of asymmetrical parts of the vertebral body under minimally invasive conditions, thus solving the problems of strong subjectivity in assessment, insufficient individualization of surgical planning, large trauma and slow recovery caused by the inability to capture minute morphological differences in the vertebral body in the existing technology.
[0006] To solve the above-mentioned technical problems, the technical solution provided by the present invention is as follows: A minimally invasive surgical assistance system for scoliosis based on AI-based recognition of vertebral morphological changes includes: A multimodal image input module is used to acquire and preprocess spinal image data; The vertebral segmentation and 3D reconstruction module is used to automatically segment the vertebral body and reconstruct a 3D surface model; The AI vertebral body minute morphological change recognition module is used to quantitatively analyze the minute morphological changes of each vertebral body and output quantitative indicators and the degree of asymmetry. The comprehensive scoliosis assessment module combines minute vertebral morphological changes with the overall spinal alignment to automatically measure Cobb angle, vertebral rotation, and trunk offset, and generate a standardized assessment report. The minimally invasive surgery planning and navigation module is used to automatically design a minimally invasive surgical plan based on the individualized vertebral micromorphological characteristics, plan the vertebral body filling site, material usage and distribution, and provide intraoperative navigation support; the minimally invasive surgical plan guides the doctor to use filling materials to fill, adjust and correct asymmetrical parts of the vertebral body under minimally invasive conditions.
[0007] Optionally, it also includes a mechanical analysis and risk warning module, which is used to assess the distribution of orthopedic stress based on the slight morphological asymmetry of the vertebral body, and to warn of the risk of intraoperative vertebral body injury, loss of orthopedic correction or mechanical imbalance.
[0008] Optionally, it also includes an output and interaction module for generating preoperative assessment reports, surgical planning schemes and visual annotation diagrams, and supports integration with minimally invasive surgical equipment and navigation systems.
[0009] Optionally, the AI vertebral micromorphological change recognition module detects and extracts the height difference, tilt, and asymmetry of the vertebral body in the left-right, front-back, and vertical directions.
[0010] Optionally, the minimally invasive surgery planning and navigation module is also used to plan minimally invasive approaches, avoid high-risk areas with abnormal morphology, and generate a three-dimensional visualization path to support real-time intraoperative navigation.
[0011] Optionally, the filler material is a commercially available filler material, either absorbable or non-absorbable.
[0012] Optionally, the multimodal image input module supports importing DICOM data from full-length spinal X-rays, CT scans, and MRI scans, and performs image registration, noise reduction, and standardization preprocessing.
[0013] Optionally, the vertebral segmentation and 3D reconstruction module automatically segments the thoracic and lumbar vertebrae based on a deep learning model, and accurately locates the endplates, marginal and central anatomical structures.
[0014] The technical solution provided by this invention has the following technical effects: 1. By using AI to accurately identify minute morphological changes in the vertebral body, and based on the identification results, guide doctors to use filling materials to fill, adjust and correct asymmetrical parts of the vertebral body under minimally invasive conditions. This truly downgrades the traditional high-trauma open surgery that relies on large steel rods and nails to a safer and minimally invasive surgical method, fundamentally reducing surgical trauma and shortening patient recovery time.
[0015] 2. Focusing on subtle morphological changes in the vertebral body, AI quantitative analysis is used to extract the height difference, tilt, and asymmetry of the vertebral body in the left-right, front-back, and vertical directions, rather than being limited to a single wedge-shaped change. This allows for earlier and more accurate capture of pathological features, significantly improving the objectivity and consistency of the assessment and effectively avoiding the subjective bias of manual image interpretation.
[0016] 3. It supports the use of commercially available absorbable or non-absorbable filling materials for vertebral body adjustment and correction under minimally invasive conditions. All filling materials used are clinically available products, eliminating the need for additional research and development of specialized materials. The technical approach is feasible, in line with current clinical practice, and has good prospects for industrial transformation.
[0017] 4. Based on the unique vertebral body morphology of each patient, the AI recognition results drive the automatic design of minimally invasive surgical plans, realizing individualized minimally invasive planning for each patient. This breaks through the limitations of traditional standardized fixation procedures and makes the surgical plan more in line with the actual anatomical characteristics of the patient.
[0018] 5. AI-assisted identification of subtle anatomical risk points optimizes minimally invasive approaches and operational paths, and provides real-time 3D visualization navigation support during surgery, effectively reducing intraoperative errors and complications and improving orthopedic results; at the same time, through biomechanical analysis and risk warning, potential risks such as intraoperative vertebral body injury and loss of orthopedic correction are identified in advance, further ensuring surgical safety.
[0019] 6. This invention integrates seven modules: multimodal image input, vertebral segmentation and three-dimensional reconstruction, AI micromorphological recognition, comprehensive assessment of scoliosis, minimally invasive surgical planning and navigation, biomechanical analysis and risk warning, and output and interaction. It realizes integrated support for the entire process from preoperative assessment and surgical planning to intraoperative navigation, truly serving the clinical closed loop of minimally invasive surgery and making up for the shortcomings of existing technologies in the field of minimally invasive orthopedics. Attached Figure Description
[0020] The above and other features and advantages of the present invention will become more apparent from a detailed description of exemplary embodiments thereof with reference to the accompanying drawings.
[0021] Figure 1 This is a schematic diagram of a minimally invasive scoliosis surgery assistance system based on AI to identify vertebral morphological changes, provided in an embodiment of the present invention. Detailed Implementation
[0022] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the embodiments set forth herein; rather, they are provided so that the invention will be thorough and complete, and the concept of the exemplary embodiments will be fully conveyed to those skilled in the art. The same reference numerals in the drawings denote the same or similar structures, and therefore their detailed description will be omitted.
[0023] The terms “a,” “one,” “the,” and “the” are used to indicate the existence of one or more elements / components / etc.; the terms “including” and “having” are used to indicate an open-ended meaning of inclusion and that other elements / components / etc. may exist in addition to the listed elements / components / etc.
[0024] like Figure 1 As shown, this embodiment provides a minimally invasive scoliosis surgery assistance system based on AI recognition of vertebral body morphological changes. The system includes a multimodal image input module, a vertebral body segmentation and three-dimensional reconstruction module, an AI vertebral body minute morphological change recognition module, a scoliosis comprehensive assessment module, a minimally invasive surgery planning and navigation module, as well as optional biomechanical analysis and risk warning modules and output and interaction modules.
[0025] The multimodal image input module is used to acquire and preprocess spinal imaging data. Operators import DICOM format images such as full-length spinal X-rays, computed tomography (CT) scans, and magnetic resonance imaging (MRI) of the patient into this module. The module automatically identifies the image type and performs image registration for different modalities, eliminating image shifts caused by patient positioning or equipment differences. Subsequent image denoising and standardization preprocessing, including filtering, grayscale normalization, and image enhancement, provides high-quality input data for subsequent vertebral segmentation. For example, for CT images, the module automatically sets the window width and level to optimize the display of skeletal structures.
[0026] Preprocessed image data is fed into the vertebral segmentation and 3D reconstruction module. This module uses a deep learning model to automatically segment each segment of the thoracic and lumbar vertebrae. Specifically, the model is trained on a large number of labeled spinal images, enabling it to accurately identify and segment the boundaries of each vertebra, including key anatomical structures such as the vertebral endplate, vertebral margins, and vertebral center. After segmentation, the module reconstructs a 3D surface model of the vertebra based on the segmentation results, generating a high-precision 3D visualization image, providing a geometric basis for subsequent morphological analysis.
[0027] The 3D reconstructed vertebral model is input into an AI-powered vertebral micromorphological change recognition module. This module performs high-precision quantitative analysis on each vertebra, detecting and extracting the height differences, tilt, and asymmetry of the vertebral body in terms of left-right, front-back, and vertical dimensions. For example, the module quantifies the degree of left-right asymmetry by calculating the ratio of the left to right vertebral height and assesses the degree of wedge-shaped changes by measuring the difference between the height of the anterior and posterior edges of the vertebral body. The module is not limited to a single wedge-shaped change but covers various subtle morphological abnormalities, ultimately outputting quantitative indicators, asymmetry degree, and risk classification. For example, for adolescent idiopathic scoliosis patients, the module can automatically mark the segment with the most significant vertebral rotation and quantify its rotation angle and percentage of asymmetry.
[0028] The AI recognition results, along with the original image data, are input into the comprehensive scoliosis assessment module. This module combines subtle vertebral morphological changes with the overall spinal alignment to automatically measure the Cobb angle, vertebral rotation angle, and trunk offset distance. Specifically, the module automatically calculates the Cobb angle by identifying the upper and lower vertebrae of the spine; assesses the degree of vertebral rotation by analyzing the relative position of the posterior wall of the vertebra and the spinous process; and assesses trunk balance by measuring the offset between the trunk midline and the pelvic midline. The module combines these measurement results with the subtle morphological changes identified by AI to generate a standardized assessment report, including the scoliosis type, progression risk prediction, and recommendations for surgical necessity.
[0029] Based on comprehensive evaluation results, the minimally invasive surgery planning and navigation module automatically designs a minimally invasive surgical plan according to the patient's individualized vertebral body micromorphological characteristics. Specifically, the module plans the vertebral body filling site, material dosage, and distribution. In this embodiment, commercially available absorbable or non-absorbable filling materials commonly used in clinical practice are selected, such as calcium phosphate bone cement and polyetheretherketone (PEEK). The module marks the asymmetrical areas of the vertebral body that need filling and adjustment through a 3D visualization interface and calculates the required volume and shape of the filling material. At the same time, the module plans a minimally invasive approach, avoiding nerve roots, blood vessels, and high-risk areas with morphological abnormalities, generating the optimal surgical path. During the operation, the module provides real-time 3D navigation support, aligning the planned path with the actual instrument positions to assist the surgeon in precise operation. The minimally invasive surgical plan guides the surgeon to use filling materials to fill, adjust, and correct asymmetrical areas of the vertebral body under minimally invasive conditions, achieving a downgrade from traditional open surgery to minimally invasive surgery.
[0030] In a preferred embodiment, the system further includes a biomechanical analysis and risk warning module. This module, based on the subtle morphological asymmetry of the vertebral body, uses finite element analysis to assess the stress distribution after correction. For example, the module simulates the stress on the vertebral body, intervertebral disc, and ligaments under different corrective forces, predicting the risk of intraoperative vertebral injury, loss of correction, or biomechanical imbalance. When the risk exceeds a preset threshold, the module issues a warning, prompting the surgeon to adjust the surgical plan, such as reducing the amount of filler, changing the filler location, or increasing internal fixation assistance.
[0031] In another preferred embodiment, the system further includes an output and interaction module. This module generates preoperative assessment reports, surgical planning schemes, and visual annotation diagrams, supporting export in formats such as PDF and DICOM. Simultaneously, the module provides a standardized application programming interface (API) to support integration with minimally invasive surgical equipment and navigation systems, enabling seamless data transfer. During the surgery, doctors can view three-dimensional anatomical structures, planned paths, and navigation information in real time through the interactive interface, improving the intuitiveness and accuracy of the operation.
[0032] The workflow of this system is as follows: 1. Import the patient's full-length spinal X-ray, CT, and MRI DICOM data, and complete image registration, noise reduction, and standardization preprocessing through the multimodal image input module; 2. The vertebral body segmentation and 3D reconstruction module automatically segments the thoracic and lumbar vertebrae based on a deep learning model, reconstructs a 3D surface model, and accurately locates the endplates, edges, and central anatomical structures. 3. The AI vertebral micromorphological change recognition module performs quantitative analysis on each vertebra and detects that there is left-right height asymmetry in the T7-T9 vertebrae and a slight wedge-shaped deformation in the L2 vertebra where the anterior edge is slightly lower than the posterior edge. 4. The comprehensive assessment module for scoliosis combines subtle morphological changes with overall force lines, measuring a Cobb angle of 42°, grade II vertebral rotation, and a trunk offset of 1.8cm, and generating a standardized assessment report. 5. The minimally invasive surgery planning and navigation module automatically designs a minimally invasive surgery plan based on the above characteristics: it plans to fill 2ml, 2.5ml, and 2ml of absorbable bone cement into the asymmetrical areas on the right side of the T7, T8, and T9 vertebral bodies, respectively, plans a percutaneous minimally invasive approach to avoid the nerve root, and generates a three-dimensional visual navigation path. 6. The mechanical analysis and risk warning module assesses the stress distribution after orthopedic surgery to confirm that the risk is within a safe range; 7. The output and interaction module generates preoperative assessment reports and surgical plans, and connects with the navigation system during the operation to assist doctors in accurately completing vertebral body filling and adjustment, achieving minimally invasive correction.
[0033] As can be seen from the above embodiments, the system provided by the present invention accurately identifies minute morphological changes in the vertebral body through AI, realizing the downgrade from open surgery to minimally invasive surgery. It has the advantages of precise identification, strong individualization, high safety, and integrated process, and can effectively solve the problems of subjective assessment, insufficient planning, and large trauma in the existing technology.
[0034] In this embodiment of the invention, the term "multiple" refers to two or more, unless otherwise explicitly defined. The terms "install," "connect," and "fix" should be interpreted broadly. For example, "connect" can mean a fixed connection, a detachable connection, or an integral connection. Those skilled in the art can understand the specific meaning of the above terms in this embodiment of the invention based on the specific circumstances.
[0035] In the description of the embodiments of the present invention, it should be understood that the terms "upper" and "lower" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing the embodiments of the present invention and simplifying the description, and do not indicate or imply that the device or unit referred to must have a specific orientation or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the embodiments of the present invention.
[0036] In the description of this specification, the terms "an embodiment," "a preferred embodiment," etc., refer to a specific feature, structure, material, or characteristic described in connection with that embodiment or example, which is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.
[0037] The above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. For those skilled in the art, the embodiments of the present invention can have various modifications and variations. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the embodiments of the present invention should be included within the protection scope of the embodiments of the present invention.
Claims
1. A minimally invasive surgical assistance system for scoliosis based on AI-based recognition of vertebral morphological changes, characterized in that, include: A multimodal image input module is used to acquire and preprocess spinal image data; The vertebral segmentation and 3D reconstruction module is used to automatically segment the vertebral body and reconstruct a 3D surface model; The AI vertebral body minute morphological change recognition module is used to quantitatively analyze the minute morphological changes of each vertebral body and output quantitative indicators and the degree of asymmetry. The comprehensive scoliosis assessment module combines minute vertebral morphological changes with the overall spinal alignment to automatically measure Cobb angle, vertebral rotation, and trunk offset, and generate a standardized assessment report. The minimally invasive surgery planning and navigation module is used to automatically design minimally invasive surgical plans based on individualized vertebral micromorphological characteristics, plan the vertebral body filling sites, material usage and distribution, and provide intraoperative navigation support; The minimally invasive surgical protocol guides doctors to use filling materials to fill, adjust, and correct asymmetrical areas of the vertebral body under minimally invasive conditions.
2. The system according to claim 1, characterized in that, It also includes a mechanical analysis and risk warning module, which is used to assess the distribution of orthopedic stress based on the slight morphological asymmetry of the vertebral body, and to warn of the risk of intraoperative vertebral body injury, loss of orthopedic correction or mechanical imbalance.
3. The system according to claim 1 or 2, characterized in that, It also includes an output and interaction module for generating preoperative assessment reports, surgical planning schemes and visual annotation diagrams, and supports integration with minimally invasive surgical equipment and navigation systems.
4. The system according to claim 1, characterized in that, The AI vertebral micromorphological change recognition module detects and extracts the height difference, tilt, and asymmetry of the vertebral body in the left-right, front-back, and up-down directions.
5. The system according to claim 1, characterized in that, The minimally invasive surgery planning and navigation module is also used to plan minimally invasive approaches, avoid high-risk areas with abnormal morphology, and generate three-dimensional visualization paths to support real-time intraoperative navigation.
6. The system according to claim 1, characterized in that, The filling material is a commercially available filling material that can be absorbed or non-absorbable.
7. The system according to claim 1, characterized in that, The multimodal image input module supports importing DICOM data from full-length spinal X-rays, CT scans, and MRI scans, and performs image registration, noise reduction, and standardization preprocessing.
8. The system according to claim 1, characterized in that, The vertebral segmentation and 3D reconstruction module automatically segments the thoracic and lumbar vertebrae based on a deep learning model, and accurately locates the endplates, marginal and central anatomical structures.