Children orthopedic monitoring method and health management system
By collecting images and video data through portable devices and combining them with image and video recognition technology, convenient and accurate pediatric orthopedic testing can be achieved, solving the convenience and radiation problems of traditional testing, supporting home and telemedicine, and improving testing efficiency and coverage.
Patent Information
- Application Number
- CN202510717143.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-30
- Publication Date
- 2025-10-17
AI Technical Summary
Existing technologies for pediatric orthopedic testing are not convenient enough, making it difficult to achieve dynamic and continuous monitoring. In addition, the high penetration rate and testing costs of traditional equipment limit the feasibility of regular screening.
Portable devices (such as mobile phones and cameras) are used to collect monitoring images and video data. Leg shape feature data is obtained through image recognition, and gait feature data is obtained through video recognition. Combined with preset algorithms and model analysis, accurate assessment of leg shape and gait characteristics can be achieved.
It requires no contact operations, eliminates children's resistance, avoids radiation risks, breaks through time and space limitations, supports home testing and telemedicine, improves testing efficiency and accuracy, and realizes the transformation from passive diagnosis to active prevention.
Smart Images
Figure CN120809182A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of computer-aided medical treatment, and in particular to a pediatric orthopedic monitoring method and a health management system. BACKGROUND
[0002] The adolescent period is a critical stage of human skeletal development. Abnormalities in leg shape (such as knee varus and knee valgus) and gait characteristics (such as foot arch collapse and stride asymmetry) often indicate potential skeletal muscle system diseases or dysfunctions. Studies have shown that early detection and intervention of lower extremity force line deviation can effectively prevent degenerative joint disease in adulthood, while abnormalities in gait patterns can increase the risk of sports injuries and body aesthetics.
[0003] Traditional medical assessments mainly rely on static imaging tests (such as X-rays and CT scans) and dynamic gait laboratory analysis within hospitals. However, these methods have significant limitations: first, the equipment popularity rate and detection cost limit the feasibility of regular screening; second, the spatial and temporal constraints of medical institutions make it difficult to achieve dynamic continuous monitoring during the growth and development period.
[0004] Therefore, there is a need for a more convenient and continuous pediatric orthopedic monitoring method. SUMMARY
[0005] Therefore, the present application provides a pediatric orthopedic monitoring method and a health management system to solve the problem of inconvenient pediatric orthopedic detection in the prior art.
[0006] The present application provides a pediatric orthopedic monitoring method, comprising:
[0007] Obtaining monitoring image data and monitoring video data;
[0008] Performing image recognition on the monitoring image data to obtain leg shape feature data;
[0009] Performing video recognition on the monitoring video data to obtain gait feature data;
[0010] Obtaining a monitoring result based on the leg shape feature data and the gait feature data.
[0011] In a preferred scheme: the leg shape feature data includes leg shape regression data and leg shape classification data; performing image recognition on the monitoring image data to obtain leg shape feature data, comprising:
[0012] Based on a preset image recognition algorithm, performing image recognition on the monitoring image data to obtain leg shape regression data;
[0013] Inputting the leg shape regression data into a preset classification decision model to obtain leg shape classification data.
[0014] In a preferred scheme: the leg shape feature data comprises arch height and foot length; the preset image recognition algorithm comprises a Canny edge detection algorithm; based on the preset image recognition algorithm, the image recognition is performed on the monitoring image data to obtain leg shape regression data, comprising:
[0015] Obtaining monitoring image data;
[0016] Based on the Canny edge detection algorithm, edge detection is performed on the monitoring image data to obtain foot contour data;
[0017] According to the foot contour data, the arch curve is detected based on the Hough transform to obtain the arch height and the foot length.
[0018] In a preferred scheme: the leg shape feature data comprises knee distance; the preset image recognition algorithm comprises a key point algorithm; based on the preset image recognition algorithm, the image recognition is performed on the monitoring image data to obtain leg shape regression data, comprising:
[0019] Obtaining monitoring image data and corresponding shooting pose data;
[0020] Based on the key point detection algorithm, key point detection is performed on the monitoring image data to obtain knee key points;
[0021] According to the shooting pose data, feature transformation is performed on the knee key points to obtain key point actual coordinates;
[0022] According to the distance between the key point actual coordinates, the knee distance is obtained.
[0023] In a preferred scheme: video recognition is performed on the monitoring image data to obtain gait feature data, comprising:
[0024] Obtaining monitoring image data;
[0025] Performing feature recognition on each frame of image of the monitoring image data to obtain original feature point coordinates;
[0026] According to the time sequence relationship between the original feature point coordinates corresponding to each frame of image, a gait cycle is obtained;
[0027] According to the gait cycle, a plurality of original feature point coordinates are divided to obtain a plurality of groups of target feature point coordinates;
[0028] According to the plurality of groups of target feature point coordinates, gait feature data is obtained.
[0029] In a preferred scheme: the plurality of groups of target feature point coordinates comprise heel touch point coordinates, foot feature point coordinates, and hip / knee joint feature point coordinates; the gait feature data comprises step symmetry, arch collapse data, and joint activity; according to the plurality of groups of target feature point coordinates, the gait feature data is obtained, comprising:
[0030] According to the heel touch point coordinates, the step symmetry is obtained;
[0031] According to the foot feature point coordinates, the foot arch collapse data is obtained;
[0032] According to the hip / knee joint feature point coordinates, the joint range of motion is obtained.
[0033] The application also provides a child orthopedics health management system, comprising:
[0034] The picture recognition and diagnosis module is used for running the child orthopedics monitoring method in any of the above.
[0035] In a preferred scheme, the system further comprises:
[0036] The medical record unified input and retrieval module is used for storing the monitoring image data, the monitoring video data and the monitoring result, and the medical record unified input and retrieval module is associated with an existing medical record database.
[0037] In a preferred scheme, the system further comprises:
[0038] The labeling module is used for obtaining labeling data of the monitoring image data, the monitoring video data and the monitoring result by doctors, and storing the labeling data in the medical record unified input and retrieval module.
[0039] The application also provides a storage medium for storing computer-readable programs or instructions, which can realize the steps in the child orthopedics monitoring method in any of the above when executed by a processor.
[0040] The beneficial effects of the above embodiments are:
[0041] The application provides a child orthopedics monitoring method and a health management system, which first obtains monitoring image data and monitoring video data, performs image recognition on the monitoring image data to obtain leg shape feature data, performs video recognition on the monitoring video data to obtain gait feature data, and then obtains monitoring results according to the leg shape feature data and the gait feature data. Compared with the prior art, the application replaces traditional X-ray and other radioactive examinations by image and video collection, eliminates the need for contact operation, eliminates the resistance of children to medical detection, avoids radiation exposure risks, is particularly suitable for periodic health screening, and can collect data by using a portable device (mobile phone / camera), break through the time and space limitations of medical institutions, allow parents to detect at any time on demand, provide support for remote medical collaboration, improve the utilization efficiency of medical resources, provide data support for early intervention, realize the mode change from passive diagnosis to active prevention, have high efficiency, economy and universality, and are expected to significantly improve the coverage and accuracy of adolescent orthopedics health management. BRIEF DESCRIPTION OF DRAWINGS
[0042] Figure 1 A flow chart of the pediatric orthopedic monitoring method provided by the present invention;
[0043] Figure 2 for Figure 1 Specific step diagram of step S103 in FIG. DETAILED DESCRIPTION
[0044] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.
[0045] Combine Figure 1 As shown, a specific embodiment of the present invention discloses a pediatric orthopedic monitoring method, comprising:
[0046] S101, acquiring monitoring image data and monitoring video data;
[0047] S102, performing image recognition on the monitoring image data to obtain leg shape feature data;
[0048] S103, performing video recognition on the monitoring image data to obtain gait feature data;
[0049] S104: Obtain monitoring results based on the leg shape feature data and the gait feature data.
[0050] In the above process, the monitoring image data and monitoring video data are pictures of the child's legs, feet, torso and other parts taken by parents, as well as walking videos. In practice, a shooting auxiliary function for each body part can be designed and added at the data acquisition end to help users obtain photos that meet the requirements and remind them to retake them. In addition, in the above process, after obtaining the leg shape feature data and gait feature data, how to analyze and obtain the monitoring results is a process that can be thought of by those skilled in the art and can be obtained through experiments. The specific implementation method is related to the actual specific diagnostic method, so this article does not make too many restrictions.
[0051] Furthermore, in a preferred embodiment, the leg shape feature data includes leg shape regression data and leg shape classification data. On this basis, the above step S102, performing image recognition on the monitoring image data to obtain the leg shape feature data, specifically includes:
[0052] Based on the preset image recognition algorithm, image recognition is performed on the monitoring image data to obtain leg shape regression data;
[0053] The leg type regression data is input into a preset classification decision model to obtain leg type classification data.
[0054] In the above process, the leg type regression data refers to data capable of representing leg type characteristics obtained by regression operation, such as determination of continuous values such as arch index and knee distance. The leg type classification data refers to data capable of representing leg type characteristics obtained by classification operation, such as X-shaped leg and arch abnormality. The preset classification decision model can be any existing model capable of decision making, such as decision tree.
[0055] In this embodiment, by combining leg type regression data and classification data, the preset image recognition algorithm is used to extract continuous quantitative indicators (regression data) such as arch index and knee distance, to realize accurate geometric shape analysis of leg type. At the same time, the regression data is input into the classification decision model to divide the leg type into different risk levels (classification data), forming a "quantitative + qualitative" dual-track evaluation system. This mechanism not only guarantees the professionalism and refinement of diagnosis, but also improves the detection efficiency and standardization, supports dynamic tracking of development trend and graded intervention guidance, and significantly optimizes the precision and practicality of adolescent orthopedic health management.
[0056] Specifically, in a preferred embodiment, the leg type feature data includes arch height and foot length, the preset image recognition algorithm includes a Canny edge detection algorithm, and the step S102 of performing image recognition on the monitoring image data to obtain leg type feature data includes:
[0057] Obtaining monitoring image data;
[0058] Performing edge detection on the monitoring image data based on the Canny edge detection algorithm to obtain foot contour data;
[0059] Detecting an arch curve based on Hough transform according to the foot contour data to obtain the arch height and the foot length.
[0060] This embodiment uses the Canny edge detection algorithm to extract the foot contour, and combines Hough transform to accurately calculate the arch height and the foot length, which has high precision and high efficiency. Through mature algorithm modular design, it is suitable for mobile devices and supports dynamic expansion. The quantitative parameters can be directly used for personalized evaluation of flat foot and other foot deformities, improving the diagnosis efficiency and user experience.
[0061] In addition, in another embodiment, the leg type feature data further includes knee distance, and the preset image recognition algorithm further includes a key point algorithm. On this basis, the step S102 of performing image recognition on the monitoring image data based on the preset image recognition algorithm to obtain leg type regression data further includes:
[0062] Obtaining monitoring image data and corresponding shooting pose data;
[0063] perform key point detection on the monitoring image data based on a key point detection algorithm to obtain knee key points;
[0064] perform feature transformation on the knee key points according to the shooting pose data to obtain actual coordinates of the key points;
[0065] obtain the knee distance according to the distance between the actual coordinates of the key points.
[0066] The embodiment automatically identifies knee joint points in combination with the key point detection algorithm, and corrects spatial deviation by using the shooting pose data, so as to realize accurate quantification of the knee distance. Not only is the leg type multi-dimensional evaluation system improved, but also the shooting error is eliminated through pose correction, the detection robustness is improved, complex scene application is supported, and the data linkage analysis and diagnosis capability is enhanced.
[0067] Further, in combination with Figure 2 as shown, in one preferred embodiment, S103, video recognition is performed on the monitoring image data to obtain gait feature data, specifically including:
[0068] S201, monitoring image data is obtained;
[0069] S202, feature recognition is performed on each frame of image of the monitoring image data to obtain original feature point coordinates;
[0070] S203, a gait cycle is obtained according to the time sequence relationship between the original feature point coordinates corresponding to each frame of image;
[0071] S204, a plurality of original feature point coordinates are divided according to the gait cycle to obtain a plurality of groups of target feature point coordinates;
[0072] S205, gait feature data is obtained according to the plurality of groups of target feature point coordinates.
[0073] The embodiment analyzes the monitoring image data frame by frame, extracts time sequence features, and divides the gait cycle, which has the following advantages. First, the original feature point coordinates are identified frame by frame, and the gait cycle is determined based on the time sequence relationship, so that the lower limb movement in the walking process is accurately captured. Second, the method does not need to rely on complex external equipment, and can complete the analysis only by using video data, which is suitable for home scenes and improves the detection convenience. In addition, the algorithm models the time sequence of the feature points to effectively filter noise interference and enhance the data reliability. Finally, the generated gait feature data can be directly associated with the motion function evaluation and abnormal risk early warning, and provides dynamic and continuous intelligent analysis support for child orthopedic health monitoring.
[0074] Specifically, in a preferred embodiment, in the above process, the multiple sets of target feature point coordinates include heel touch point coordinates, foot feature point coordinates, hip / knee joint feature point coordinates, and the gait feature data includes step symmetry, foot arch collapse data, and joint range of motion. Correspondingly, the above step S305, obtaining gait feature data according to the multiple sets of target feature point coordinates, specifically includes:
[0075] obtaining step symmetry according to the heel touch point coordinates;
[0076] obtaining foot arch collapse data according to the foot feature point coordinates;
[0077] obtaining joint range of motion according to the hip / knee joint feature point coordinates.
[0078] The above process provides some preferred gait feature analysis methods. Specifically, the step length of the left and right feet can be calculated according to the heel touch point coordinates, and then the step symmetry can be analyzed. The trajectory of the key points of the foot during walking can be calculated through the foot feature point coordinates, and then the foot arch collapse degree can be analyzed. For example, the foot feature point coordinates include calcaneus coordinates, first metatarsal head coordinates, and fifth metatarsal head coordinates, a first vector is established according to the calcaneus coordinates and the first metatarsal head coordinates, a second vector is established according to the calcaneus coordinates and the fifth metatarsal head coordinates, and then the cosine angle of the first vector and the second vector is analyzed as the foot arch collapse data. The length of the joint at rest, the maximum range of motion of the joint, the hip flexion angle, the hip extension angle, and other data can be calculated through the hip / knee joint feature point coordinates, and then the flexion amplitude is obtained as the joint range of motion.
[0079] The application also provides a pediatric orthopedic monitoring system, comprising:
[0080] a picture recognition and diagnosis module for running the pediatric orthopedic monitoring method of any one of the above.
[0081] Specifically, the picture recognition and diagnosis module can be set on the parent's mobile phone or computer, and can be used to submit the child's monitoring image data and monitoring video data and other related information, including classification selection data and secondary classification selection data of core parts such as foot bottom, lower limbs, upper limbs, and spine, and related medical record information, corresponding video and picture data, and establishing a personal information center. At the same time, the child's personal information, contact information, follow-up consultation records, and other information can also be submitted.
[0082] Further, in a preferred embodiment, the health management system further comprises:
[0083] a medical record unified input and retrieval module for storing monitoring image data, monitoring video data, and monitoring results, and the medical record unified input and retrieval module is associated with an existing medical record database.
[0084] The medical record unified input and retrieval module is built on the basis of the existing medical record system in the hospital, and is a faster, more complete and more convenient database system. It is mainly used for data migration and integration, ensures that the database contains the existing medical records of the hospital, and ensures the integrity and accuracy of the data, and optimizes the database structure and index to improve the speed and efficiency of data query, so that doctors can view the multi-dimensional data of patients at any time and anywhere through mobile phones and computers, and quickly locate and find the records of any patient.
[0085] Further, in a preferred embodiment, the health management system described above further comprises:
[0086] The labeling module is used for obtaining the labeling data of the doctors on the monitoring image data, the monitoring video data and the monitoring results, and storing the labeling data in the medical record unified input and retrieval module.
[0087] The labeling module can be set on the mobile phone and computer of the doctor, and the diagnosis result information of the child can be added, deleted, modified and inquired, including uploading electronic medical records and scanned copies (picture pg format or pdf format, supporting c recognition of scanned copies), the parents can see the information added by the doctor in synchronization, and any update can prompt the parents and the doctor. The introduction of the labeling module can help the doctor to label the photos, mark the abnormal areas or key features. The identification result and the doctor's labeling information and the picture can be updated to the case unified input and retrieval system in real time, so that the parents can obtain the diagnosis update notification in the first time.
[0088] The embodiment also provides a computer readable storage medium, which stores a child orthopedics monitoring program, and the child orthopedics monitoring program can realize the steps in the above embodiment when executed by a processor.
[0089] The present application provides a kind of child orthopedics monitoring method and health management system, it first obtains monitoring image data and monitoring video data, carries out image recognition to monitoring image data, obtains leg type feature data, carries out video identification to monitoring video data, obtains gait feature data, then according to leg type feature data and gait feature data, obtains monitoring result. Compared with prior art, the present application replaces traditional X-ray and other radioactive examination by image and video acquisition, without contact operation, eliminates the resistance psychology of children to medical detection, while avoiding the risk of radiation exposure, especially suitable for periodic health screening. And, the present application can collect data using portable device (mobile phone / camera), break through the time and space limit of medical institution, parents can detect at any time as needed, and provide support for telemedicine collaboration, improve the utilization efficiency of medical resources, provide data support for early intervention, the present application realizes the mode change from passive diagnosis to active prevention, with high efficiency, economy and universality, which is expected to significantly improve the coverage and accuracy of adolescent orthopedics health management.
[0090] It should be noted that each of the above-described embodiments of the present application are described in the context of progressing from one embodiment to the next, and each embodiment highlights the differences from the other embodiments. The same and similar parts between the various embodiments are therefore inter-changeable and can be referred to each other.
[0091] The foregoing description of the disclosed embodiments enables a person skilled in the art to make or use the application. Modifications of these embodiments will occur to persons of skill in the art, and are within the scope of this application, while the general principles defined herein can be applied to other embodiments without departing from the spirit or scope of the application. The present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A pediatric orthopedic monitoring method, characterized in that: include: Acquiring monitoring image data and monitoring video data; Perform image recognition on the monitoring image data to obtain leg shape feature data; Perform video recognition on monitoring image data to obtain gait feature data; The monitoring results are obtained based on the leg shape feature data and gait feature data.
2. The pediatric orthopedic monitoring method according to claim 1, characterized in that: The leg shape feature data includes leg shape regression data and leg shape classification data; Perform image recognition on the monitoring image data to obtain leg shape feature data, including: Based on the preset image recognition algorithm, image recognition is performed on the monitoring image data to obtain leg shape regression data; The leg shape regression data is input into the preset classification decision model to obtain the leg shape classification data.
3. The pediatric orthopedic monitoring method according to claim 2, characterized in that: Leg shape feature data includes arch height and foot length; the preset image recognition algorithm includes the canny edge detection algorithm; Based on the preset image recognition algorithm, image recognition is performed on the monitoring image data to obtain leg shape regression data, including: Acquire monitoring image data; Perform edge detection on the monitoring image data based on the canny edge detection algorithm to obtain the foot contour data; According to the foot contour data, the arch curve is detected based on Hough transform to obtain the arch height and foot length.
4. The pediatric orthopedic monitoring method according to claim 2, characterized in that: Leg shape feature data includes knee distance; preset image recognition algorithm includes key point algorithm; Based on the preset image recognition algorithm, image recognition is performed on the monitoring image data to obtain leg shape regression data, including: Obtain monitoring image data and corresponding shooting posture data; Perform key point detection on the monitoring image data based on the key point detection algorithm to obtain the knee key points; Perform feature transformation on the knee key points according to the shooting posture data to obtain the actual coordinates of the key points; The knee spacing is obtained based on the spacing of the actual coordinates of the key points.
5. The pediatric orthopedic monitoring method according to claim 1, characterized in that: Perform video recognition on the monitoring image data to obtain gait feature data, including: Acquire monitoring image data; Perform feature recognition on each frame of monitoring image data to obtain the original feature point coordinates; The gait cycle is obtained according to the temporal relationship between the coordinates of the original feature points corresponding to each frame of the image; Divide the coordinates of multiple original feature points according to the gait cycle to obtain multiple groups of target feature point coordinates; Gait feature data is obtained based on multiple sets of target feature point coordinates.
6. The pediatric orthopedic monitoring method according to claim 5, characterized in that: Multiple sets of target feature point coordinates include heel touchdown coordinates, foot feature point coordinates, and hip / knee joint feature point coordinates; gait feature data include step length symmetry, arch collapse data, and joint range of motion; According to multiple sets of target feature point coordinates, gait feature data is obtained, including: According to the coordinates of the heel contact point, the step length symmetry is obtained; According to the coordinates of the foot feature points, the arch collapse data is obtained; The range of motion of the joint is obtained based on the coordinates of the hip / knee joint feature points.
7. A children's orthopedic health management system, characterized in that: include: An image recognition and diagnosis module is used to run the pediatric orthopedic monitoring method according to any one of claims 1 to 6.
8. The health management system according to claim 7, characterized in that: Also includes: The unified medical record entry and retrieval module is used to store monitoring image data, monitoring imaging data and monitoring results. The unified medical record entry and retrieval module is associated with the existing medical record database.
9. The health management system according to claim 7, characterized in that: Also includes: The annotation module is used to obtain the doctor's annotation data of monitoring image data, monitoring imaging data and monitoring results, and store the annotation data in the unified medical record entry and retrieval module.
10. A storage medium, characterized in that: Used to store computer-readable programs or instructions, which, when executed by a processor, can implement the steps of any one of the pediatric orthopedic monitoring methods in claims 1-6.