Intelligent guidance system and method for shoulder pain rehabilitation

Abstract

The invention discloses an intelligent guidance system and method for shoulder pain rehabilitation, and belongs to the technical field of health information. According to the method, a continuous image sequence of the upper body of a user is obtained, real-time unmarked motion data of the user is generated based on a human body topology model and an optical flow field analysis method, and the unmarked motion data comprises a personal digital skeleton formed by joint angle estimation values and a force grading estimation value. According to the invention, non-interference and personalized rehabilitation monitoring and guiding are realized, the constraint that a physical sensor must be worn in traditional rehabilitation monitoring is eliminated by adopting a pure visual unmarked motion capture technology, and a patient is allowed to train in a completely natural state, so that the comfort and rehabilitation experience of the user are greatly improved, and the rehabilitation effect of the patient is improved. A high-quality data basis is provided for subsequent personalized guidance, and the willingness and compliance of the patient insisting in rehabilitation for a long time are effectively enhanced.

Description

A smart guidance system and method for shoulder pain rehabilitation Technical Field

[0001] This invention belongs to the field of health information technology, and in particular relates to an intelligent guidance system and method for shoulder pain rehabilitation. Background Technology

[0002] Shoulder pain, especially postoperative or chronic strain-related pain, is a common problem affecting patients' quality of life and functional recovery. Rehabilitation exercises, as a core non-pharmacological treatment, are crucial for relieving pain and restoring joint range of motion and muscle strength. Traditional shoulder rehabilitation typically relies on printed illustrated manuals or in-person demonstrations by therapists, requiring patients to memorize the movements and consistently perform them. With the development of information technology, the use of mobile applications or simple videos to assist rehabilitation training has become increasingly common, providing patients with more convenient guidance resources.

[0003] In existing technological practices, some advanced rehabilitation assistive solutions are beginning to incorporate technological means to improve effectiveness. Some solutions utilize wearable inertial sensors to capture the patient's joint movement angles and transmit the data via Bluetooth to a mobile application for recording and comparison. Other solutions design simple two-dimensional or three-dimensional games, allowing patients to complete repetitive rehabilitation movements in an entertaining way, aiming to increase their enjoyment and initiative in training. Furthermore, telemedicine platforms are also beginning to be used for doctor-patient communication, allowing doctors to provide periodic remote guidance to patients via video calls and other means.

[0004] However, the aforementioned existing technologies still have significant shortcomings in practical applications. While wearable sensor-based solutions can provide quantitative data, the process of wearing the devices is cumbersome and may interfere with the patient's natural activities, affecting the accuracy of the data. At the same time, the cost of the devices also limits their widespread adoption. Simple rehabilitation games often lack linkage with the patient's actual physiological state, and their guidance content is fixed, making it impossible to dynamically adjust according to the user's real-time performance and fatigue level, which can easily lead to mismatched training intensity. Furthermore, existing telemedicine platforms mostly remain at the information exchange level, making it difficult for doctors to obtain objective and continuous data on the patient's daily training. This results in a lack of accurate basis for their remote guidance, and there are still blind spots in the quality monitoring of the rehabilitation process. Summary of the Invention

[0005] The purpose of this invention is to address the problem that existing telemedicine platforms mostly remain at the information exchange level, making it difficult for doctors to obtain objective and continuous data on patients' daily training, resulting in a lack of accurate basis for their remote guidance. Therefore, this invention proposes an intelligent guidance system and method for shoulder pain rehabilitation.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: an intelligent guidance method for shoulder pain rehabilitation, specifically including the following steps: S1, acquiring a continuous image sequence of the user's upper body, and generating real-time label-free motion data of the user based on a human topology model and an optical flow field analysis method, wherein the label-free motion data includes a personal digital skeleton composed of joint angle estimates and force grading estimates; S2, acquiring rehabilitation stage goals, dynamically calculating the real-time coordinates of shoulder acupoints in the rehabilitation stage goals based on the personal digital skeleton to obtain dynamic acupoint coordinates, and generating an initial virtual guidance path based on the dynamic acupoint coordinates and the rehabilitation stage goals; S3, calculating the deviation of the current movement based on the label-free motion data and the initial virtual guidance path, and calculating muscle fatigue based on the change of force grading estimates in the label-free motion data over time, and combining the deviation and the muscle fatigue to adjust the initial virtual guidance path in real time to generate an optimized virtual guidance path.

[0007] As a further description of the above technical solution: the generation of real-time label-free motion data for the user includes: processing the continuous image sequence to extract skin texture feature regions in the images; calculating the motion vector field of the skin texture feature regions between continuous images; mapping the motion vector field to the human body topology model to estimate and generate joint angle estimates in the personal digital skeleton; analyzing the amplitude changes of the motion vector field in the corresponding muscle regions to generate the force grading estimates.

[0008] As a further description of the above technical solution: the generation of the initial virtual guidance path includes determining the expected activity range and target acupoints for this training from the rehabilitation stage goals; calculating the projection position of the target acupoints in the user's physical space based on the current posture of the personal digital skeleton; planning a guidance trajectory that avoids the preset pain-sensitive direction with the projection position as the endpoint, and generating the initial virtual guidance path with the guidance trajectory.

[0009] As a further description of the above technical solution: the initial virtual guidance path is adjusted in real time; the matching error between the user's current posture and the initial virtual guidance path is calculated based on the deviation; the user's current exercise endurance level is assessed based on the muscle fatigue level; based on the matching error and the exercise endurance level, a preset adjustment strategy is invoked to modify the guidance speed or trajectory curvature parameters of the initial virtual guidance path, thereby generating the optimized virtual guidance path.

[0010] As a further description of the above technical solution: it also includes adjusting the virtual task based on the matching error and the sports endurance level, and synchronously converting the modification of the guiding speed or trajectory curvature parameter into a change in the movement speed or position of the target object in the virtual task; when the matching error is lower than a first threshold and the sports endurance level is higher than a second threshold, the challenge level of the virtual task is increased.

[0011] As a further description of the above technical solution: the method also includes the following steps: continuously recording the unmarked motion data, the deviation, and the muscle fatigue to form a training process log; performing time series analysis on the training process log to generate a rehabilitation progress trend curve and a compliance index; obtaining the user's self-reported pain score, and performing correlation analysis between the self-reported pain score and the rehabilitation progress trend curve to generate a personalized adaptation profile.

[0012] As a further description of the above technical solution: the method also includes the following steps: uploading the personalized adaptive profile to a cloud server via the network; obtaining rehabilitation prescription adjustment instructions for the personalized adaptive profile from the doctor's terminal; and updating the rehabilitation stage goals according to the rehabilitation prescription adjustment instructions.

[0013] As a further description of the above technical solution: the method also includes the following steps: comparing the compliance index with a plan for determining training completion, and generating a training reminder notification; when the compliance index is lower than a preset compliance threshold, sending an auxiliary supervision request to an authorized family member terminal; receiving an encouragement message sent from the family member terminal, and presenting the encouragement message to the user.

[0014] As a further description of the above technical solution: the human topology model and a standard for generating the force grading estimate are established through the following steps: acquiring synchronous image sequences and sensor force data of healthy subjects performing standard rehabilitation movements; training the human topology model using the synchronous image sequences and corresponding joint angle data to establish a mapping relationship library between image features and joint angles; calibrating the sensor force data with the motion vector features in the mapping relationship library to establish a standard for the force grading estimate.

[0015] A guidance system for an intelligent guidance method in shoulder pain rehabilitation includes: an image acquisition module for acquiring a continuous image sequence of the user's upper body; a data processing center connected to the image acquisition module, which processes the continuous image sequence based on a human topology model and optical flow field analysis method to generate label-free motion data including a personal digital skeleton and force grading estimates; acquiring rehabilitation stage goals, calculating dynamic acupoint coordinates based on the personal digital skeleton, and generating an initial virtual guidance path; calculating deviation and muscle fatigue based on the label-free motion data and the initial virtual guidance path, and generating an optimized virtual guidance path by combining the deviation and muscle fatigue; and a virtual reality interaction module connected to the data processing center, configured to receive the optimized virtual guidance path and generate corresponding virtual task scenarios.

[0016] In summary, due to the adoption of the above technical solutions, the beneficial effects of the present invention are as follows: 1. The present invention achieves non-intrusive and personalized rehabilitation monitoring and guidance. By adopting pure visual markerless motion capture technology, it breaks free from the constraints of wearing physical sensors in traditional rehabilitation monitoring, allowing patients to train in a completely natural state. This not only greatly improves the user's comfort and rehabilitation experience, but also avoids stiff and unnatural movements caused by wearing devices, thereby obtaining more realistic and accurate motion data. This provides a high-quality data foundation for subsequent personalized guidance and effectively enhances the patient's willingness and compliance to persist in rehabilitation in the long term.

[0017] 2. This invention constructs a dynamic closed-loop adaptive rehabilitation guidance system. This system can analyze the user's movement accuracy and physiological fatigue state in real time, and adjust the difficulty and pace of virtual guidance in real time based on the analysis results. This ensures that the intensity of rehabilitation training is always matched with the patient's current ability, which can provide effective stimulation to promote recovery, and avoid fatigue accumulation or secondary injury caused by overtraining. This dynamic adjustment mechanism makes the rehabilitation process safer and more efficient, and significantly improves the overall treatment effect.

[0018] 3. This invention uploads objective, quantifiable rehabilitation data to a cloud platform, enabling doctors to overcome time and space limitations, remotely and accurately monitor patients' home rehabilitation progress, and adjust treatment plans accordingly. It also introduces the care and supervision of family members, transforming the solitary rehabilitation task into an emotionally supportive family interaction. This multi-party collaborative model effectively solves the problems of poor information flow and lack of supervision in traditional rehabilitation, optimizes the allocation of medical resources, and creates a positive rehabilitation atmosphere, thereby comprehensively improving the quality and efficiency of rehabilitation management. Attached Figure Description

[0019] Figure 1 is a flowchart illustrating an intelligent guidance system and method for shoulder pain rehabilitation proposed in this invention. Detailed Implementation

[0020] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0021] Please refer to Figure 1. This invention provides a technical solution: an intelligent guidance method for shoulder pain rehabilitation, specifically including the following steps: S1, acquiring a continuous image sequence of the user's upper body, and generating real-time label-free motion data of the user based on a human topology model and an optical flow field analysis method. The label-free motion data includes a personal digital skeleton composed of joint angle estimates and force grading estimates; generating real-time label-free motion data of the user includes: processing the continuous image sequence to extract skin texture feature regions in the images; calculating the motion vector field of the skin texture feature regions between continuous images; mapping the motion vector field to the human topology model, estimating and... Generate joint angle estimates in the personal digital skeleton; analyze the amplitude changes of the motion vector field in the corresponding muscle region to generate the force grading estimate; the human topology model and a standard for generating the force grading estimate are established through the following steps: acquiring synchronous image sequences and sensor force data of healthy subjects performing standard rehabilitation movements; training the human topology model using the synchronous image sequences and corresponding joint angle data to establish a mapping relationship library between image features and joint angles; calibrating the sensor force data with the motion vector features in the mapping relationship library to establish a standard for the force grading estimate; S2, obtain rehabilitation stage goals, based on The personal digital skeleton dynamically calculates the real-time coordinates of shoulder acupoints in the rehabilitation stage goals to obtain dynamic acupoint coordinates, and generates an initial virtual guidance path based on the dynamic acupoint coordinates and the rehabilitation stage goals. Generating the initial virtual guidance path includes determining the expected activity range and target acupoints for this training from the rehabilitation stage goals; calculating the projection position of the target acupoints in the user's physical space based on the current posture of the personal digital skeleton; planning a guidance trajectory that avoids a preset pain-sensitive direction with the projection position as the endpoint, and generating the initial virtual guidance path based on the guidance trajectory; S3, based on the unmarked motion data and the initial virtual guidance path, calculating the current... The deviation of the previous movement and the change of muscle fatigue based on the force grading estimate in the unmarked motion data over time are used to calculate muscle fatigue. The initial virtual guidance path is then adjusted in real time based on the deviation and muscle fatigue to generate an optimized virtual guidance path. The initial virtual guidance path is adjusted in real time, and the matching error between the user's current posture and the initial virtual guidance path is calculated based on the deviation. The user's current exercise endurance level is assessed based on the muscle fatigue. Based on the matching error and the exercise endurance level, a preset adjustment strategy is invoked to modify the guidance speed or trajectory curvature parameters of the initial virtual guidance path to generate the optimized virtual guidance path.It also includes adjusting the virtual task based on the matching error and the exercise endurance level, synchronously converting modifications to the guide speed or trajectory curvature parameters into changes in the movement speed or position of the target object in the virtual task; when the matching error is below a first threshold and the exercise endurance level is above a second threshold, the challenge level of the virtual task is increased.

[0022] The method further includes the following steps: continuously recording the unmarked motion data, the deviation, and the muscle fatigue to form a training process log; performing time series analysis on the training process log to generate a rehabilitation progress trend curve and a compliance index; obtaining the user's self-reported pain score and performing correlation analysis between the self-reported pain score and the rehabilitation progress trend curve to generate a personalized adaptation profile.

[0023] The method further includes the following steps: uploading the personalized adaptation profile to a cloud server via the network; obtaining rehabilitation prescription adjustment instructions for the personalized adaptation profile from the doctor's terminal; updating the rehabilitation stage goals according to the rehabilitation prescription adjustment instructions; the method further includes the following steps: comparing the compliance index with a plan for determining training completion and generating a training reminder notification; when the compliance index is lower than a preset compliance threshold, sending an assistance and supervision request to an authorized family member's terminal; receiving an encouragement message sent from the family member's terminal and presenting the encouragement message to the user.

[0024] A guidance system for an intelligent guidance method for shoulder pain rehabilitation includes: an image acquisition module for acquiring a continuous image sequence of the user's upper body; a data processing center connected to the image acquisition module for processing the continuous image sequence based on a human topology model and an optical flow field analysis method; acquiring rehabilitation stage goals; calculating dynamic acupoint coordinates based on the personal digital skeleton; and generating an initial virtual guidance path.

[0025] In this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance; the term "multiple" refers to two or more unless otherwise explicitly defined. The terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; "linking" can be a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0026] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. An intelligent guidance method for shoulder pain rehabilitation, characterized in that, Specifically, the following steps are included: S1. Acquire a continuous image sequence of the user's upper body, and generate real-time label-free motion data of the user based on a human body topology model and an optical flow field analysis method. The label-free motion data includes a personal digital skeleton composed of joint angle estimates and force grading estimates. S2. Obtain the rehabilitation stage goal, dynamically calculate the real-time coordinates of shoulder acupoints in the rehabilitation stage goal based on the personal digital skeleton, obtain dynamic acupoint coordinates, and generate an initial virtual guidance path based on the dynamic acupoint coordinates and the rehabilitation stage goal. S3. Based on the unmarked motion data and the initial virtual guidance path, calculate the deviation of the current action, and calculate the muscle fatigue based on the change of the force grading estimate in the unmarked motion data over time. Combine the deviation and the muscle fatigue, adjust the initial virtual guidance path in real time to generate an optimized virtual guidance path.

2. The intelligent guidance method for shoulder pain rehabilitation according to claim 1, characterized in that, The process of generating real-time label-free motion data for the user includes: processing the continuous image sequence to extract skin texture feature regions from the images; calculating the motion vector field of the skin texture feature regions between the continuous images; mapping the motion vector field to the human body topology model to estimate and generate joint angle estimates in the personal digital skeleton; and analyzing the amplitude changes of the motion vector field in the corresponding muscle regions to generate the force grading estimates.

3. The intelligent guidance method for shoulder pain rehabilitation according to claim 1, characterized in that, The generation of the initial virtual guidance path includes determining the expected activity range and target acupoints for this training from the rehabilitation stage goals; and calculating the projection position of the target acupoints in the user's physical space based on the current posture of the personal digital skeleton. Using the projection position as the endpoint, a guide trajectory that avoids the preset pain-sensitive direction is planned, and the initial virtual guide path is generated based on the guide trajectory.

4. The intelligent guidance method for shoulder pain rehabilitation according to claim 1, characterized in that, The initial virtual guidance path is adjusted in real time, and the matching error between the user's current posture and the initial virtual guidance path is calculated based on the deviation. The user's current exercise endurance level is assessed based on the muscle fatigue level. Based on the matching error and the athletic endurance level, a preset adjustment strategy is invoked to modify the guiding speed or trajectory curvature parameters of the initial virtual guiding path, thereby generating the optimized virtual guiding path.

5. The intelligent guidance method for shoulder pain rehabilitation according to claim 4, characterized in that, It also includes adjusting the virtual task based on the matching error and the exercise endurance level, and synchronously converting the modification of the guide speed or trajectory curvature parameter into a change in the movement speed or position of the target object in the virtual task; when the matching error is lower than a first threshold and the exercise endurance level is higher than a second threshold, the challenge level of the virtual task is increased.

6. The intelligent guidance method for shoulder pain rehabilitation according to claim 1, characterized in that, The method further includes the following steps: continuously recording the unmarked motion data, the deviation, and the muscle fatigue to form a training process log; performing time series analysis on the training process log to generate a rehabilitation progress trend curve and a compliance index; obtaining the user's self-reported pain score and performing correlation analysis between the self-reported pain score and the rehabilitation progress trend curve to generate a personalized adaptation profile.

7. The intelligent guidance method for shoulder pain rehabilitation according to claim 6, characterized in that, The method further includes the following steps: uploading the personalized adaptation profile to a cloud server via the network; obtaining rehabilitation prescription adjustment instructions for the personalized adaptation profile from the doctor's terminal; and updating the rehabilitation stage goals according to the rehabilitation prescription adjustment instructions.

8. The intelligent guidance method for shoulder pain rehabilitation according to claim 7, characterized in that, The method further includes the following steps: comparing the compliance index with a plan for determining training completion and generating a training reminder notification; when the compliance index is lower than a preset compliance threshold, sending an assistance and supervision request to an authorized family member terminal; receiving an encouragement message sent from the family member terminal and presenting the encouragement message to the user.

9. The intelligent guidance method for shoulder pain rehabilitation according to claim 1, characterized in that, The human topology model and a standard for generating the force grading estimate are established through the following steps: acquiring synchronous image sequences and sensor force data of healthy subjects performing standard rehabilitation movements; training the human topology model using the synchronous image sequences and corresponding joint angle data to establish a mapping relationship library between image features and joint angles; calibrating the sensor force data with the motion vector features in the mapping relationship library to establish a standard for the force grading estimate.

10. A guidance system for an intelligent guidance method for shoulder pain rehabilitation according to any one of claims 1-9, characterized in that, include: The image acquisition module is used to acquire a continuous sequence of images of the user's upper body; The data processing center, connected to the image acquisition module, processes the continuous image sequence based on a human body topology model and optical flow field analysis method; obtains rehabilitation stage goals, calculates dynamic acupoint coordinates based on the personal digital skeleton, and generates an initial virtual guidance path.

Images