Stroke patient body and mind assessment and rehabilitation integrated system based on traditional Chinese medicine rehabilitation theory

Abstract

The invention belongs to the technical field of assessment and rehabilitation, and discloses a stroke patient body and mind assessment and rehabilitation integrated system based on a traditional Chinese medicine rehabilitation theory, which comprises an equipment installation and data acquisition module, a data preprocessing module, a feature set construction module, an assessment module, a maintenance module and a data storage management and transmission module. According to the invention, non-contact data acquisition and automatic evaluation, scoring and question answering are carried out aiming at the whole-body movement function and psychological condition of the stroke patient, so that related work can be conveniently carried out in hospital and community environments; and the accuracy of an evaluation result and the convenience of later maintenance are greatly improved by utilizing an artificial intelligence algorithm and an expert traditional Chinese medicine rehabilitation theory. The rehabilitation physiotherapist can further evaluate the function of the patient according to the score of each action and the total score of evaluation in combination with physical and mental data, so as to formulate targeted intervention measures.

Description

Technical Field

[0001] The present invention belongs to the technical field of assessment and rehabilitation, and in particular relates to an integrated system for physical and mental assessment and rehabilitation of stroke patients based on traditional Chinese medicine rehabilitation theory. Background Art

[0002] Stroke, also known as stroke, is an acute brain injury caused by a sudden interruption of blood supply to the brain or rupture of cerebral blood vessels. According to the type of stroke, it can be divided into ischemic stroke (the majority) and hemorrhagic stroke. Ischemic stroke is caused by blockage or stenosis of cerebral blood vessels, while hemorrhagic stroke is caused by rupture of cerebral blood vessels, resulting in blood seeping into brain tissue. Globally, stroke is the third leading cause of death and disability. With the increasing global population and the increasingly serious aging phenomenon, the prevalence and disability rate of stroke are expected to continue to rise. Stroke can cause great harm to the patient's body and mind. The physical effects of stroke are divided into: 1). Limb dysfunction: The most common physical effect of stroke is that patients with hemiplegia (unilateral limb weakness or paralysis) usually lose control of one side of the body, which manifests as weakness in the arms and legs and difficulty in movement. 2). Balance and coordination problems: Due to damage to the area of ​​the brain that controls balance and movement coordination, patients experience problems such as unsteady gait and dizziness, increasing the risk of falling. 3). Epileptic seizures: Some stroke patients will experience epileptic seizures after the onset of the disease, which is caused by abnormal neural activity after brain damage. The psychological effects of stroke can be divided into: 1). Depression and anxiety: Common psychological reactions after stroke are depression and anxiety, especially for patients who have lost their ability to move and be independent. The loss of physical function coupled with uncertainty about the future can make them feel helpless, isolated, and even lose confidence in life. 2). Decreased cognitive function: Stroke affects the cognitive area of ​​the brain, leading to decreased memory, lack of concentration, confusion, and decreased judgment. These problems can affect patients' daily lives and make it difficult for them to complete tasks that they were previously proficient in. 3). Psychological stress and difficulty in social adaptation: Due to the decline in quality of life and role changes caused by stroke, patients often feel tremendous psychological pressure. Some people feel that they have become a burden to their families and find it difficult to readapt to work and social life, which further exacerbates their psychological distress.

[0003] Physical and mental assessment helps to detect and identify various physical and mental problems after stroke at an early stage, such as movement disorders, sensory disorders, cognitive disorders, psychological disorders, etc. Rehabilitation therapy can effectively help stroke patients recover or improve impaired functions, such as limb movement function, sensory function, cognitive function, etc., and improve patients' daily living ability and quality of life. Physical and mental assessment is the premise and basis of rehabilitation therapy. Only through comprehensive physical and mental assessment can a scientific and reasonable rehabilitation treatment plan be formulated. At the same time, the effect of rehabilitation therapy also needs to be monitored and evaluated through physical and mental assessment. To this end, we need to propose a unified and effective plan for physical and mental assessment and rehabilitation of stroke patients in the existing medical system. Western medicine adopts a systematic and evidence-based medicine-based approach to physical and psychological assessment and rehabilitation of stroke patients. For physical assessment of patients, imaging CT scans and National Institute of Health stroke scale (NIHSS) neurological assessment are often used. For psychological assessment, the Barthel index is often used to assess the independence of patients' daily activities (ADL), such as dressing, eating, and going to the toilet, and the Fugl-Meyer assessment scale (FMA) is used to assess the recovery of motor function, especially the motor ability of the upper and lower limbs. In the assessment of the physical and psychological conditions of stroke patients, Western medicine adopts a cycle of assessment first, rehabilitation training, and finally assessment. Although there are more rigorous quantitative tools and standardized processes, there are still problems such as insufficient attention to mental health assessment, one-sided overall health assessment, and insufficient personalized assessment. In addition, the rehabilitation content of Western medicine mainly includes progressive physical resistance training, walking training, and joint range of motion exercises. For psychological intervention, antidepressant drugs are prescribed. Western medicine usually emphasizes the treatment of the acute phase of stroke, and the time window of the rehabilitation stage is often limited in Western medicine. Western medicine treatment also leads to drug dependence and side effects. Western medicine's local treatment overemphasizes physical function recovery and ignores psychological and overall adjustment, resulting in poor rehabilitation results.

[0004] In contrast, TCM adopts a unique holistic concept and a dialectical treatment approach to assess the physical and psychological conditions of stroke patients, and conducts a comprehensive assessment of patients through the "four examinations" of observation, auscultation, inquiry, and palpation. Compared with Western medicine, TCM emphasizes more on individualization and holistic conditioning, not only focusing on specific symptoms, but also on the overall state, emotions, and social environment of the patient. TCM has a unique set of methods for stroke rehabilitation, focusing on holistic conditioning and dialectical treatment of individual differences, covering comprehensive conditioning of both physical and psychological aspects. Including acupuncture, massage, Chinese medicine conditioning, emotional conditioning, Qigong and Tai Chi, etc., focusing on holistic conditioning and individual differences, covering both physical function recovery and mental health conditioning. Compared with the standardized and localized treatment of Western medicine, TCM's treatment methods are more flexible and natural, with fewer side effects, and have significant advantages in long-term rehabilitation and psychological adjustment. TCM emphasizes holistic view and dialectical treatment, focusing on regulating the body's balance and qi and blood, pursuing the purpose of health preservation while curing diseases. This concept of "preventing disease before it occurs" has attracted many people who focus on long-term health and disease prevention. With the improvement of living standards, people pay more attention to health and their needs become more diversified. However, in terms of stroke physical and mental assessment, TCM lacks the rigorously verified and widely recognized assessment tools and standards like Western medicine, which leads to a strong subjectivity in the assessment results. A comprehensive comparison of the theoretical methods of TCM and Western medicine for stroke physical and mental assessment and rehabilitation shows that the concept of TCM "holistic adjustment and prevention before illness" is consistent with the trend of modern people's pursuit of health management. We need a stroke assessment and rehabilitation system based on TCM rehabilitation theory, which needs to integrate Western medicine quantitative processes and assessment-training-assessment cycle theory with traditional TCM treatment to avoid the defects of unclear mechanism and lack of objectivity.

[0005] With the development of science and technology, artificial intelligence (AI) and large language models (such as GPT-4) are gradually playing an increasingly important role in the physical and psychological assessment and rehabilitation of stroke patients. Through machine learning and deep learning technology, AI can analyze medical images such as CT and MRI to help doctors diagnose the type of stroke, damaged brain areas, and the severity of hemorrhage or infarction more quickly and accurately. AI-driven rehabilitation robots can help stroke patients with limb rehabilitation training, such as upper and lower limb exercise training. The robot can adjust the intensity and method of training in real time according to the patient's specific rehabilitation situation. By analyzing the patient's facial expressions, voice intonation and language content, AI can assess the patient's emotional state, including anxiety, depression, stress response, etc. The large language model can analyze the patient's conversation records or voice logs, identify emotional changes and give corresponding psychological intervention suggestions. Artificial intelligence and large language models currently have the advantages of high efficiency and accuracy, personalization and dynamic adjustment, full process, popularity and low cost, which will greatly help the physical and mental assessment and rehabilitation of stroke patients. Combined with the theory of traditional Chinese medicine rehabilitation, more down-to-earth and more effective rehabilitation can be carried out.

[0006] Functional assessment of body movement plays an important role in the rehabilitation process. At present, Western medicine clinics mostly use scales to assess motor dysfunction after stroke, and the commonly used scales are roughly divided into the following categories: ① Based on changes in muscle conditions ② Based on changes in movement patterns ③ Based on changes in upper limb function. Traditional Chinese medicine mainly diagnoses through traditional Chinese medicine techniques such as observation, smell, questioning, and palpation. However, all of the above methods have some limitations: 1) The recording, review, and analysis process requires a lot of manpower and time; 2) The assessment and analysis are highly subjective. Therefore, there is a need for a method that uses high-precision equipment combined with AI intelligent recognition algorithms to collect real-time and refined data on physical conditions (such as the invention patent CN202410333606.0 "A stroke rehabilitation quality assessment system and its assessment method", which collects patient body data through artificial intelligence, but has the defect of not being able to capture patient body data in real time and in a refined manner, so we need a high-precision acquisition solution that can capture patient movement data in real time). At the same time, in the context of clinical rehabilitation and community rehabilitation, the commonly used machine vision analysis devices and inertial and pressure sensors show the singleness of the data, that is, the incompleteness of the evaluation of the overall body motor function assessment results (such as the invention patent CN202210827083.6 "A multi-dimensional lower limb rehabilitation assessment device and assessment method for stroke patients", which analyzes and evaluates lower limb movements, but lacks whole-body movement assessment. Therefore, we need a solution to collect all types of patient whole-body movement data rather than a single type). In addition, some studies have carried out the development of prognostic psychological rehabilitation assessment systems for stroke patients (such as invention patent CN202110792300.8 "Psychological assessment system for stroke patients based on automatic data collection of Internet terminals", which uses "questionnaire scoring + expert intervention" to assess patients' psychology. It has the defect of low intelligence level of rehabilitation assessment and difficult to accurately evaluate. We need an intelligent psychological assessment solution, such as invention patent CN202210188719.7 "A positive psychological nursing intervention system for stroke patients", which analyzes psychological trends through images and videos, but has the defect of insufficient initiative in assessment and treatment. We need a system that gives intelligent rehabilitation response in time after the psychological assessment is completed), but there is still room for improvement: 1) Doctors are still required to be present for real-time question and answer assessment, which is not convenient for data collection and assessment in the patient's home and community environment; 2) The assessment scale is too subjective, and there is a lack of objective human data to assist in verifying the conclusion; 3) The system has not been deeply combined with the theory of traditional Chinese medicine rehabilitation for system development, and only stays on the surface of the disease.At the same time, many studies have integrated algorithms and medical theories for the physical and mental assessment and rehabilitation of stroke patients (such as the invention patent CN202311686157.X "A lightweight visual posture assessment and contactless emotion guidance rehabilitation system", which uses algorithm training to recognize human facial expressions and predict emotions, but the training method cannot form personalized predictions for each user, that is, the algorithm training lacks personalization. We need a personalized training method based on AI), but it still has the following problems: the rigidity and dogmatism of traditional training, and it is impossible to conduct personalized training and master the unique physical movement and psychological expression characteristics of each patient; the training process lacks parameter guidance based on traditional Chinese medicine exercise methods, and some data irrelevant to the evaluation have not been deleted, and there is a lack of medical rigor.

[0007] Through the above analysis, the problems and defects of the prior art are as follows:

[0008] (1) The rigidity and dogmatism of traditional training make it impossible to conduct personalized training and grasp the unique physical movement and psychological expression characteristics of each patient;

[0009] (2) The training process lacked parameter guidance based on traditional Chinese medicine exercise methods and failed to delete some data irrelevant to the evaluation, resulting in a lack of medical rigor. Summary of the invention

[0010] In view of the problems existing in the prior art, the present invention provides an integrated system for physical and mental assessment and rehabilitation of stroke patients based on traditional Chinese medicine rehabilitation theory.

[0011] The present invention is implemented as follows: a stroke patient physical and mental assessment and rehabilitation integrated system based on traditional Chinese medicine rehabilitation theory, comprising:

[0012] Equipment installation and data acquisition module, data preprocessing module, feature set construction module, evaluation module, maintenance module, data storage management and transmission module;

[0013] The equipment installation and data acquisition module is connected with the data preprocessing module and the data storage management and transmission module, and is used to build the assessment and rehabilitation task environment and install the equipment, carry out the corresponding assessment tasks according to the adopted physical assessment scale and psychological question-and-answer system, and identify the required data from the screening;

[0014] The data preprocessing module is connected with the equipment installation and data collection module, the feature set construction module, the evaluation module, and the data storage management and transmission module, and is used to preprocess the above collected data, including data filtering and segmentation, semantic recognition and organization;

[0015] The feature set construction module is connected with the data preprocessing module and the evaluation module to calculate the corresponding kinematic features, classify the corresponding psychological analysis features, and dynamically screen the features to construct a feature set personalized for the patient;

[0016] The evaluation module is connected with the data preprocessing module, the feature set construction module, the maintenance module, and the data storage management and transmission module, and is used to use the selected evaluation scale to perform single automatic scoring and total score calculation for each task on the kinematic feature set constructed by the above modules through the trained model to analyze the physical condition; use the local large language model to intelligently analyze the patient's psychological condition through the knowledge graph based on traditional Chinese medicine theory; and retrieve relevant evaluation results based on the personalized knowledge graph constructed according to the different conditions of each patient;

[0017] A maintenance module, connected to the evaluation module, is used to maintain the constructed evaluation model, update the above-constructed feature set by accumulating the motion data of new patients, and update the parameters of each evaluation model by cross-validation method;

[0018] The data storage management and transmission module is connected with the equipment installation and data acquisition module, the feature set construction module, and the evaluation module to save the patient's score results. The rehabilitation therapist makes a personalized rehabilitation evaluation report based on the training score, and sends the report, training plan and training opinions to the patient through this module, and the patient can review it freely.

[0019] Further, the device installation and data acquisition module method is as follows:

[0020] Demarcate the rehabilitation assessment area and install the equipment required for the test; the patient stands in the planned area, watches a video explaining the movement guidance and psychological questions / questionnaires of the rehabilitation assessment, and uses a computer and a 2K smart wide-angle USB camera device to collect whole-body movement data. The computer psychological questions and answers interact with the patient through intelligent questions and answers;

[0021] Then, a customized motion data acquisition system was run on the computer to identify and filter the required key joint tracking data, and then a psychological question-answering system based on a large language model was run to perform semantic processing on the collected data. After that, a data set was constructed according to the 10 motion tasks included in the adopted stroke rehabilitation motor function scale.

[0022] The data comes from recruited stroke patients of various degrees; all participants were evaluated by senior rehabilitation therapists with rich medical practice experience using the STREAM scale and signed the relevant informed consent; when all assessment tasks are completed, a total of 10 different motion data sets will be generated; these data will be organized into corresponding databases according to the task types for data accumulation and later maintenance work;

[0023] Psychological Q&A / questionnaire data are collected through a computer questionnaire system; Q&A or questionnaire questions are designed by professional therapists based on traditional Chinese medicine rehabilitation theory.

[0024] Further, the data preprocessing module method is as follows:

[0025] Utilize a wide-angle USB camera device to capture real-time video feeds for motor function assessment and input into the assessment system;

[0026] First, the video source is processed by Gaussian filtering, and the feature data is intelligently screened through the self-developed and improved YOLO algorithm;

[0027] Then, real-time video is shot with a wide-angle camera, and data on the corresponding joints is extracted through the self-developed and improved YOLO algorithm, which intelligently analyzes the patient's movement thinking and builds a personalized movement data map. The computer intelligent psychological evaluation system is used to collect the patient's intelligent question and answer data, and the data is organized by intelligent summarization, keyword extraction and other operations based on the question and answer results, thereby building a personalized psychological data map.

[0028] Further, the feature set construction module method is as follows:

[0029] Extract relevant kinematic features from the video data preprocessed in the above steps, such as identifying 17 key points of each person's body. Each human target frame information contains six data: the horizontal coordinates and vertical coordinates of the human anchor point, the width and height of the human prediction frame, the confidence of the human detection frame, and the category confidence;

[0030] Define three main joint angles: elbow angle, shoulder angle, and leg angle; the elbow angle refers to the angle formed by the shoulder, elbow, and wrist in a two-dimensional plane; the shoulder angle refers to the angle formed by the elbow, shoulder, and hip in a two-dimensional plane; the leg angle refers to the angle formed by the hip, knee, and calf in a two-dimensional plane;

[0031] The angles of the elbow, shoulder, and leg are used for rehabilitation movement scoring; the calculation formulas for the three types of angles are as follows:

[0032]

[0033] The shoulder, elbow, wrist, hip, knee and calf are represented by W, M, P, D, T and X respectively; the shoulder-elbow line, elbow-wrist line, shoulder-hip line, hip-knee line and knee-calf line are represented by WM, MP, WD, DT and TX respectively; θ1, θ2 and θ3 represent the angle between the shoulder-elbow line and the elbow-wrist line, the angle between the elbow-shoulder line and the shoulder-hip line, and the angle between the hip-knee line and the knee-calf line.

[0034] Further, the evaluation module method is as follows:

[0035] A neural network model is used as the core of the physical assessment model; the neural network model consists of three parts: input layer, hidden layer, and output layer; the number of neurons in the input layer is the number of kinematic features, and the number of neurons in the output layer corresponds to several different scores for each task of STREAM. This assessment model can be automatically scored by AI;

[0036] Use the intelligent motor function assessment system to automatically score and sum the patient's exercise. First, use the trained neural network model to automatically score according to the feature set constructed in step 3, then sum the final scores of each task to get the total score of this assessment, and output the physical assessment content corresponding to the scale.

[0037] Knowledge graph and intelligent recommendation algorithm are used as the core of the psychological assessment model; the knowledge graph is constructed and stored based on the disease knowledge about stroke in the fields of traditional Chinese and Western medicine and traditional Chinese medicine exercise methods, including symptom classification, assessment diagnosis, syndrome classification, psychological state, semantic types of intervention plans and the semantic relationships between them. Each semantic type has a corresponding description to illustrate the concept and characteristics of the semantic type.

[0038] Further, the data storage management and transmission module method is as follows:

[0039] Rehabilitation therapists log into the integrated assessment system to obtain patient information, and use the calculated scores, movement data, and psychological assessment results as a reference to assess the physical and mental rehabilitation of stroke patients. They compile a personalized rehabilitation assessment report and send the report and treatment opinions of the training plan to the patient remotely through the data storage management and transmission system, so that the patient can further strengthen his or her exercises in a targeted manner.

[0040] Another object of the present invention is to provide a method for integrating physical and mental assessment and rehabilitation of stroke patients based on TCM rehabilitation theory, including:

[0041] Step 1: Use the equipment installation and data collection module to build the assessment and rehabilitation task environment and install the equipment, carry out the corresponding assessment tasks based on the physical assessment scale and psychological question-and-answer system used, and identify the required data from the screening;

[0042] Step 2: preprocess the collected data through a data preprocessing module, including data filtering and segmentation, semantic recognition and organization;

[0043] Step 3, calculate the corresponding kinematic features through the feature set construction module, classify the corresponding psychological analysis features, and dynamically screen the features to construct a feature set personalized for the patient;

[0044] Step 4: The evaluation module uses the selected evaluation scale to automatically score each task and calculate the total score of the kinematic feature set constructed by the above module through the trained model to analyze the physical condition; the local large language model is used to intelligently analyze the patient's psychological condition through the knowledge graph based on traditional Chinese medicine theory; at the same time, the personalized knowledge graph constructed according to the different conditions of each patient retrieves relevant evaluation results;

[0045] Step 5, maintaining the constructed evaluation model through the maintenance module, updating the above constructed feature set by accumulating the motion data of new patients, and updating the parameters of each evaluation model through the cross-validation method;

[0046] Step 6: The patient's score results are saved through the data storage management and transmission module. The rehabilitation therapist makes a personalized rehabilitation assessment report based on the training score and sends the report, training plan and training suggestions to the patient through this module. The patient can review them freely.

[0047] Another object of the present invention is to provide a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the integrated method for physical and mental assessment and rehabilitation of stroke patients based on traditional Chinese medicine rehabilitation theory.

[0048] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, enables the processor to execute the steps of the integrated method for physical and mental assessment and rehabilitation of stroke patients based on traditional Chinese medicine rehabilitation theory.

[0049] Another object of the present invention is to provide an information data processing terminal, which is used to implement the integrated system for physical and mental assessment and rehabilitation of stroke patients based on traditional Chinese medicine rehabilitation theory.

[0050] In combination with the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solutions to be protected by the present invention are analyzed from the following aspects:

[0051] First, in view of the technical problems existing in the above-mentioned prior art and the difficulty of solving the problems, the technical solutions to be protected by the present invention and the results and data during the research and development process are closely combined to analyze in detail and deeply how the technical solutions of the present invention solve the technical problems, and some creative technical effects brought about after solving the problems. The specific description is as follows:

[0052] The present invention proposes an integrated system for physical and mental assessment and rehabilitation of stroke patients based on TCM rehabilitation theory to carry out rehabilitation assessment and recuperation of stroke patients. Body, that is, the patient's body: real-time body data collection and analysis based on high-precision wide-angle cameras and deep learning algorithms, redundant data is processed by the self-developed improved YOLO algorithm to summarize the patient's unique database, and multimodal fusion artificial intelligence algorithms are used to evaluate the movement function. Mind, that is, the patient's psychology: quick psychological testing and data collection based on TCM theory and large language models, using the "TCM theory + intelligent question and answer + AI analysis" three-in-one method to form a comprehensive psychological map of the patient, and judge the patient's psychological state through data analysis and abnormal feature screening and artificial intelligence-based psychological health assessment. Finally, the system integrates the physical and mental assessment data, combines artificial intelligence algorithms, absorbs the advantages of traditional Chinese medicine rehabilitation theory and avoids the defects of Western medicine rehabilitation theory, and proposes a comprehensive rehabilitation plan with medical basis. According to the principle of "adjusting the mind, regulating the breath, and regulating the body in one", Tai Chi, Ba Duan Jin, Wu Qin Xi, and Yi Jin Jing are intelligently selected from traditional Chinese medicine exercise methods as rehabilitation means. They help patients restore neurological function and relieve anxiety by regulating body functions, promoting blood circulation, and enhancing muscle strength. Compared with similar technologies at home and abroad, the technology included in this digital physical and mental assessment intelligent processing platform is more adaptable to scenarios (applicable to family, community, and hospital environments), the output results are more accurate and faster, and the hardware it relies on is cheaper, making it more suitable for large-scale promotion. At the same time, it is more in line with the national policy of vigorously promoting traditional Chinese medicine exercise health-preserving exercises, meets the needs of social development, and fills the gap in the research field of integrated physical and mental assessment and rehabilitation system for stroke patients based on traditional Chinese medicine rehabilitation theory.

[0053] Step 1: A wide-angle USB camera (preferred but not limited to DS-UVC-U64 Pro device for capture and collection) was used to collect motion data in real time. Compared with traditional wearable sensor-based data acquisition devices (such as whole-body motion capture devices, etc.), the use of non-contact camera equipment as the hardware device for data acquisition avoids the defects of discomfort, size incompatibility, and inability of patients with severe stroke to wear the device normally. During the data collection process, due to the large image collection range of the wide-angle camera and the high accuracy of the collected data, it is conducive to accurately understanding the patient's movement status and ensuring the effective evaluation of motor function. On the other hand, due to its simple use, no debugging is required, and it can be used immediately after powering on, it avoids the early calibration and coordinate calibration steps similar to the hand and leg data acquisition systems, thereby further realizing and facilitating data collection in the community environment.

[0054] Step 2 and Step 3: The present invention has a data preprocessing module with good performance and a patient-personalized feature set construction module. The self-developed improved YOLO algorithm used in the present invention can accurately extract the designated coordinate data of the body of the stroke patient with high accuracy, light weight and real-time. On the other hand, the self-developed algorithm is used to dynamically screen and extract features, forming a personalized feature library for the patient, thereby further improving the accuracy of the evaluation.

[0055] Step 4 and Step 5: Compared with the traditional stroke patient whole body motion capture and physical and mental assessment methods, the intelligent assessment system of the present invention has the advantages of low cost, good stability, strong objectivity, etc., and can be easily applied in hospitals and community environments. The physical assessment model used is developed based on neural networks and has excellent classification and pattern recognition capabilities. The psychological assessment model used is based on expert guidance and a traditional Chinese medicine database, and has excellent case analysis and psychological problem-solving capabilities. At the same time, a post-maintenance module is constructed to regularly update the internal parameters of the model through the patient motion data accumulated in actual use, thereby improving its accuracy in actual use and continuously improving the adaptability of the local population.

[0056] Step 6: The data storage, management and transmission system of the present invention enables physical therapists to obtain the patient's movement data for evaluation, improves work efficiency, and reduces the cost of patient and social rehabilitation services. Data storage, management and transmission can more conveniently manage the patient's data and avoid data loss that affects the rehabilitation progress. At the same time, physical therapists can use the system to promptly feedback the rehabilitation evaluation report to the patient, which increases the pertinence of rehabilitation training and improves the effectiveness of rehabilitation effects.

[0057] The intelligent rehabilitation system for stroke patients based on the traditional Chinese medicine rehabilitation theory of the present invention is a set of rehabilitation equipment with low cost, good stability and rich usage scenarios. In the future, the equipment is expected to be applied to different rehabilitation environments such as medical clinics and communities to assist in the physical and mental rehabilitation treatment of stroke patients. The prototype operation process of this system is standardized, and the output evaluation results are objective and standardized. It aims to make it easy for attending physicians to view and use it, so that one rehabilitation therapist can serve multiple patients, improve work efficiency and reduce workload, and correspondingly reduce the financial burden of rehabilitation services on patients and society. Physiotherapists can also use accumulated patient data to regularly update and maintain the evaluation model to improve the performance of the evaluation system.

[0058] Compared with the prior art, the present invention has three essential differences and advantages: 1) The high-precision recognition algorithm based on the deep learning model and attention mechanism of the present invention has high accuracy, fast response and real-time recognition compared with general algorithms. Based on the improved YOLO algorithm, a unique patient database can also be formed; 2) Based on AI+big language, the present invention uses artificial intelligence algorithms and local big prediction models to intelligently and automatically score, evaluate, and conduct question-and-answer surveys, thereby obtaining evaluation data that is more recognized by rehabilitation therapists, rather than evaluating the patient's motor function and psychological condition through subjective judgment; 3) The evaluation and rehabilitation system based on TCM exercise exercises proposed by the present invention forms a medically based and different from Western medicine diagnosis and treatment stroke patient evaluation and rehabilitation system through the patient data accumulated in actual clinical practice and the knowledge map formed by TCM exercise exercises. The TCM exercise exercises body and mind evaluation and rehabilitation system is the product of the deep integration of traditional Chinese medicine wisdom and modern technology, and is a new innovation in the direction of combining TCM with AI to evaluate rehabilitation.

[0059] In summary, the present invention performs contactless data collection and automated assessment and scoring of stroke patients' whole-body motor function and psychological status, making it convenient to carry out related work in hospitals and community environments, and greatly improves the accuracy of the assessment results and the convenience of later maintenance by using artificial intelligence algorithms and expert TCM rehabilitation theories. Rehabilitation therapists can conduct further functional assessments of patients based on the scores of each action and the total score of the assessment, combined with physical and mental data, to formulate targeted intervention measures.

[0060] Second, as auxiliary evidence of the inventiveness of the claims of the present invention, it is also reflected in the following important aspects:

[0061] (1) The expected benefits and commercial value of the technical solution of the present invention after transformation are:

[0062] The commercial value of the present invention lies in its innovative integration of TCM rehabilitation theory and modern artificial intelligence technology, providing a comprehensive solution for the rehabilitation treatment of stroke patients. The system realizes real-time evaluation of the patient's body movement function through high-precision cameras and deep learning algorithms, and combines TCM theory with large language models to accurately analyze the patient's psychological state. This integrated evaluation method not only improves the scientificity and personalization of the rehabilitation program, but also broadens its application scenarios, making it suitable for a variety of environments such as families, communities and hospitals. In addition, the system has obvious advantages in hardware cost and ease of operation, making it easier to promote and popularize. It can not only provide accurate and fast evaluation results, but also intelligently recommend suitable TCM exercise methods such as Tai Chi and Ba Duan Jin based on the evaluation data to help patients recover physically and mentally. Globally, with the advent of an aging society and the increase in rehabilitation needs, the market potential of the present invention is huge. More importantly, the system is in line with the national policy orientation of promoting TCM exercise health-preserving exercises, conforms to the trend of social development, and fills the research gap in the field of TCM rehabilitation. Therefore, the present invention not only has significant social value, but also has huge commercial potential, and is expected to become an innovative leader in the field of stroke rehabilitation.

[0063] (2) The technical solution of the present invention fills the technical gap in the industry at home and abroad:

[0064] TCM rehabilitation has received extensive attention and development in China, especially in the TCM departments, orthopedics and other departments of some TCM hospitals and general hospitals, where the ideas and technical methods of TCM rehabilitation have been widely used. However, the current assessment and treatment methods of TCM rehabilitation are relatively scattered, lacking a systematic and integrated physical and mental assessment rehabilitation system. Western medicine has relatively mature technologies and methods in the field of rehabilitation, but often focuses on the functional recovery of a single organ or system, lacking the guidance of a holistic view. Although there is a certain understanding and interest in TCM rehabilitation theory and methods abroad, its application and development still face many challenges due to differences in cultural background and medical system. The present invention can fill the technical gaps in the field of systematic TCM rehabilitation system and the field of holistic stroke patient rehabilitation assessment. The system can combine the holistic functional view of TCM to conduct a comprehensive and systematic assessment of the individual's physical and mental state. This includes a comprehensive evaluation of multiple aspects such as physical function, psychological state, and quality of life, thereby more accurately reflecting the individual's rehabilitation needs. Based on the results of the integrated physical and mental assessment, the system can formulate personalized rehabilitation plans. These plans not only focus on the recovery of physical functions, but also focus on the adjustment of psychological state and the improvement of quality of life, thereby achieving comprehensive physical and mental rehabilitation.

[0065] (3) The technical solution of the present invention solves the technical problems that people have been eager to solve but have never been able to solve successfully:

[0066] 1. Traditional physical and mental assessment methods often focus on a single aspect of physical function or psychological state, lacking comprehensiveness and accuracy.

[0067] 2. The mutual influence and connection between the body and mind are often overlooked, resulting in incomplete and in-depth evaluation results.

[0068] 3. There are many types of traditional Chinese medicine rehabilitation techniques, but they often lack systematicity and integration.

[0069] 4. Traditional rehabilitation process management often relies on manual recording and monitoring, lacking intelligence and automation. The integrated physical and mental assessment and rehabilitation system based on TCM rehabilitation theory can combine the holistic view of TCM and the principle of dialectical treatment to conduct a comprehensive and systematic assessment of the individual's physical and mental state. The system can comprehensively consider multiple aspects such as physical function, psychological state, living habits, environmental factors, etc., so as to more accurately reflect the individual's rehabilitation needs and potential problems.

[0070] (4) The technical solution of the present invention overcomes technical prejudice:

[0071] TCM rehabilitation theory emphasizes concepts such as yin and yang balance, viscera and meridians, qi, blood and body fluids, which are considered vague or difficult to quantify in the modern scientific system, leading some people to doubt the scientific nature of TCM rehabilitation theory. Through scientific methods and strategies such as scientific verification, interdisciplinary cooperation, standardized evaluation, intelligent evaluation, and personalized treatment, the system can overcome these technical biases.

[0072] Third, the technical solution of the present invention is based on the theory of traditional Chinese medicine rehabilitation, integrating the physical and mental assessment of stroke patients with the rehabilitation process, and solves the following key problems existing in the prior art:

[0073] 1. Inefficiency of data collection and processing: In the prior art, rehabilitation assessment for stroke patients mainly relies on manual operation, with low efficiency of data collection and processing, and problems of incomplete or erroneous data. The present invention realizes automatic collection and intelligent filtering of assessment data through equipment installation and data collection module 1 and data preprocessing module 2, greatly improving the efficiency and accuracy of data processing.

[0074] 2. Lack of personalized assessment and rehabilitation guidance: Traditional rehabilitation assessment methods usually use fixed scales, which cannot effectively assess and provide rehabilitation guidance for individual differences of different patients. The present invention uses the feature set construction module 3 to dynamically construct a personalized feature set based on the patient's kinematic and psychological characteristics, and combines traditional Chinese medicine rehabilitation theory and knowledge graphs to intelligently provide personalized rehabilitation assessment and plans for each patient, greatly enhancing the accuracy and applicability of the assessment.

[0075] 3. Evaluation results and rehabilitation programs are not updated in time: The evaluation results of traditional rehabilitation methods are usually difficult to update and adjust in time according to the patient's rehabilitation progress. The present invention uses the maintenance module 5 to continuously optimize and update the evaluation model using the continuously collected new data, thereby achieving dynamic adjustment of the rehabilitation program, ensuring that the rehabilitation process matches the patient's actual condition, and thus improving the rehabilitation effect.

[0076] 4. Inconvenient communication between patients and physical therapists: In the prior art, rehabilitation assessment results usually need to be transmitted manually, and there are delays and inconveniences in the interaction between patients and physical therapists. The present invention uses the data storage management and transmission module 6 to timely save and transmit assessment results and rehabilitation plans. Patients can check their rehabilitation progress at any time and obtain personalized rehabilitation training suggestions, thereby greatly improving rehabilitation efficiency and patient participation.

[0077] The technical solution of the present invention solves the problems of low evaluation efficiency, insufficient personalization, and delayed feedback in the prior art through automated and intelligent data collection and processing, combined with personalized rehabilitation evaluation and dynamic program updates, and significantly improves the rehabilitation effect of stroke patients and the technical level of industrial applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1 It is a structural block diagram of an integrated system for physical and mental assessment and rehabilitation of stroke patients based on TCM rehabilitation theory provided by an embodiment of the present invention.

[0079] Figure 2 It is a flow chart of an integrated method for physical and mental assessment and rehabilitation of stroke patients based on TCM rehabilitation theory provided by an embodiment of the present invention.

[0080] Figure 3 It is a technical roadmap provided by the embodiments of the present invention.

[0081] Figure 4 It is a schematic diagram of system motion analysis provided by an embodiment of the present invention.

[0082] Figure 5 It is a schematic diagram of the interactive interface of the evaluation and analysis system provided in an embodiment of the present invention.

[0083] Figure 1 In: 1. Equipment installation and data acquisition module; 2. Data preprocessing module; 3. Feature set construction module; 4. Evaluation module; 5. Maintenance module; 6. Data storage management and transmission module. DETAILED DESCRIPTION

[0084] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0085] like Figure 1 As shown, an embodiment of the present invention provides an integrated system for physical and mental assessment and rehabilitation of stroke patients based on TCM rehabilitation theory, comprising:

[0086] Equipment installation and data collection module 1, data preprocessing module 2, feature set construction module 3, evaluation module 4, maintenance module 5, data storage management and transmission module 6.

[0087] The equipment installation and data acquisition module 1 is connected with the data preprocessing module 2 and the data storage management and transmission module 6, and is used to build the assessment and rehabilitation task environment and install the equipment, carry out the corresponding assessment tasks according to the adopted physical assessment scale and psychological question-and-answer system, and identify the required data from the screening;

[0088] The data preprocessing module 2 is connected to the equipment installation and data collection module 1, the feature set construction module 3, the evaluation module 4, and the data storage management and transmission module 6, and is used to preprocess the above collected data, including data filtering and segmentation, semantic recognition and arrangement;

[0089] The feature set construction module 3 is connected with the data preprocessing module 2 and the evaluation module 4, and is used to calculate the corresponding kinematic features, classify the corresponding psychological analysis features, and dynamically screen the features to construct a feature set personalized for the patient;

[0090] The evaluation module 4 is connected with the data preprocessing module 2, the feature set construction module 3, the maintenance module 5, and the data storage management and transmission module 6, and is used to use the selected evaluation scale to perform single automatic scoring and total score calculation for each task on the kinematic feature set constructed by the above modules through the trained model to analyze the physical condition; use the local large language model to intelligently analyze the patient's psychological condition through the knowledge graph based on traditional Chinese medicine theory; and retrieve relevant evaluation results according to the personalized knowledge graph constructed according to the different conditions of each patient;

[0091] The maintenance module 5 is connected to the evaluation module 4 and is used to maintain the constructed evaluation model, update the above-constructed feature set by accumulating the motion data of new patients, and update the parameters of each evaluation model by cross-validation method;

[0092] The data storage management and transmission module 6 is connected with the equipment installation and data collection module 1, the feature set construction module 3, and the evaluation module 4, and is used to save the patient's score results. The rehabilitation therapist makes a personalized rehabilitation evaluation report based on the training score, and sends the report, training plan and training opinions to the patient through this module, and the patient can freely review it.

[0093] The integrated system for physical and mental assessment and rehabilitation of stroke patients based on TCM rehabilitation theory provided by the embodiment of the present invention first builds a complete assessment and rehabilitation task environment through the equipment installation and data acquisition module 1. This module installs the required equipment according to the physical assessment scale and psychological question-and-answer system used, and collects the patient's physical movement data and psychological assessment data. After the data is screened, the data that needs to be focused on is identified for analysis by subsequent modules.

[0094] Subsequently, the data preprocessing module 2 processes the collected raw data. This process includes data filtering, data segmentation, and semantic recognition and organization, thereby generating a high-quality, analyzable data set. The processed data can be further used to construct a personalized feature set for the patient, providing a basis for evaluation and analysis. This module works closely with the subsequent feature set construction module and evaluation module to ensure the integrity and consistency of the data.

[0095] Next, the feature set construction module 3 calculates the corresponding kinematic features based on the processed data, and classifies and organizes the psychological analysis features. Through dynamic screening, a personalized feature set of the patient is constructed. These feature sets include various information such as the patient's motor ability and psychological state, which are analyzed and scored by the evaluation module 4. The evaluation module uses the trained model to automatically score each task of the kinematic feature set and comprehensively calculates the patient's physical condition. At the same time, the module also uses a knowledge graph based on traditional Chinese medicine theory to intelligently analyze the patient's psychological condition, and outputs the evaluation results in combination with a personalized knowledge graph.

[0096] Finally, the maintenance module 5 is responsible for continuously updating the evaluation model, optimizing the model using the data of new patients, and updating and optimizing the parameters through the cross-validation method. The evaluation results are saved by the data storage management and transmission module 6 and transmitted to the rehabilitation therapist. The therapist generates a personalized rehabilitation evaluation report based on the results and sends the report and training plan to the patient. Patients can freely check their own rehabilitation progress and suggestions through this module, thereby providing scientific basis and guidance for subsequent rehabilitation treatment.

[0097] The device installation and data acquisition module method provided by the embodiment of the present invention is as follows:

[0098] Demarcate the rehabilitation assessment area and install the equipment required for the test; the patient stands in the planned area, watches a video explaining the movement guidance and psychological questions / questionnaires of the rehabilitation assessment, and uses a computer and a 2K smart wide-angle USB camera device to collect whole-body movement data. The computer psychological questions and answers interact with the patient through intelligent questions and answers;

[0099] Then, a customized motion data acquisition system was run on the computer to identify and filter the required key joint tracking data, and then a psychological question-answering system based on a large language model was run to perform semantic processing on the collected data. After that, a data set was constructed according to the 10 motion tasks included in the adopted stroke rehabilitation motor function scale.

[0100] The data comes from recruited stroke patients of various degrees; all participants were evaluated by senior rehabilitation therapists with rich medical practice experience using the STREAM scale and signed the relevant informed consent; when all assessment tasks are completed, a total of 10 different motion data sets will be generated; these data will be organized into corresponding databases according to the task types for data accumulation and later maintenance work;

[0101] Psychological Q&A / questionnaire data are collected through a computer questionnaire system; Q&A or questionnaire questions are designed by professional therapists based on traditional Chinese medicine rehabilitation theory.

[0102] The data preprocessing module method provided by the embodiment of the present invention is as follows:

[0103] Utilize a wide-angle USB camera device to capture real-time video feeds for motor function assessment and input into the assessment system;

[0104] First, the video source is processed by Gaussian filtering, and the feature data is intelligently screened through the self-developed and improved YOLO algorithm;

[0105] Then, real-time video is shot with a wide-angle camera, and data on the corresponding joints is extracted through the self-developed and improved YOLO algorithm, which intelligently analyzes the patient's movement thinking and builds a personalized movement data map. The computer intelligent psychological evaluation system is used to collect the patient's intelligent question and answer data, and the data is organized by intelligent summarization, keyword extraction and other operations based on the question and answer results, thereby building a personalized psychological data map.

[0106] The feature set construction module method provided by the embodiment of the present invention is as follows:

[0107] Extract relevant kinematic features from the video data preprocessed in the above steps, such as identifying 17 key points of each person's body. Each human target frame information contains six data: the horizontal coordinates and vertical coordinates of the human anchor point, the width and height of the human prediction frame, the confidence of the human detection frame, and the category confidence;

[0108] Define three main joint angles: elbow angle, shoulder angle, and leg angle; the elbow angle refers to the angle formed by the shoulder, elbow, and wrist in a two-dimensional plane; the shoulder angle refers to the angle formed by the elbow, shoulder, and hip in a two-dimensional plane; the leg angle refers to the angle formed by the hip, knee, and calf in a two-dimensional plane;

[0109] The angles of the elbow, shoulder, and leg are used for rehabilitation movement scoring; the calculation formulas for the three types of angles are as follows:

[0110]

[0111] The shoulder, elbow, wrist, hip, knee and calf are represented by W, M, P, D, T and X respectively; the shoulder-elbow line, elbow-wrist line, shoulder-hip line, hip-knee line and knee-calf line are represented by WM, MP, WD, DT and TX respectively; θ1, θ2 and θ3 represent the angle between the shoulder-elbow line and the elbow-wrist line, the angle between the elbow-shoulder line and the shoulder-hip line, and the angle between the hip-knee line and the knee-calf line.

[0112] The evaluation module method provided by the embodiment of the present invention is as follows:

[0113] A neural network model is used as the core of the physical assessment model; the neural network model consists of three parts: input layer, hidden layer, and output layer; the number of neurons in the input layer is the number of kinematic features, and the number of neurons in the output layer corresponds to several different scores for each task of STREAM. This assessment model can be automatically scored by AI;

[0114] Use the intelligent motor function assessment system to automatically score and sum the patient's exercise. First, use the trained neural network model to automatically score according to the feature set constructed in step 3, then sum the final scores of each task to get the total score of this assessment, and output the physical assessment content corresponding to the scale.

[0115] Knowledge graph and intelligent recommendation algorithm are used as the core of the psychological assessment model; the knowledge graph is constructed and stored based on the disease knowledge about stroke in the fields of traditional Chinese and Western medicine and traditional Chinese medicine exercise methods, including symptom classification, assessment diagnosis, syndrome classification, psychological state, semantic types of intervention plans and the semantic relationships between them. Each semantic type has a corresponding description to illustrate the concept and characteristics of the semantic type.

[0116] The data storage management and transmission module method provided by the embodiment of the present invention is as follows:

[0117] Rehabilitation therapists log into the integrated assessment system to obtain patient information, and use the calculated scores, movement data, and psychological assessment results as a reference to assess the physical and mental rehabilitation of stroke patients. They compile a personalized rehabilitation assessment report and send the report and treatment opinions of the training plan to the patient remotely through the data storage management and transmission system, so that the patient can further strengthen his or her exercises in a targeted manner.

[0118] like Figure 2As shown, an embodiment of the present invention provides a method for integrating physical and mental assessment and rehabilitation of stroke patients based on TCM rehabilitation theory, including:

[0119] S101, build the assessment and rehabilitation task environment and install the equipment through the equipment installation and data collection module, carry out the corresponding assessment tasks according to the adopted physical assessment scale and psychological question-and-answer system, and identify the required data from the screening;

[0120] S102, preprocessing the collected data through a data preprocessing module, including data filtering and segmentation, semantic recognition and organization;

[0121] S103, calculating corresponding kinematic features, classifying corresponding psychological analysis features, and dynamically screening the features to construct a feature set personalized for the patient through a feature set construction module;

[0122] S104, using the selected evaluation scale through the evaluation module, and using the trained model to perform single automatic scoring and total score calculation for each task on the kinematic feature set constructed by the above module to analyze the physical condition; using the local large language model, intelligently analyzing the patient's psychological condition through the knowledge graph based on traditional Chinese medicine theory; and retrieving relevant evaluation results based on the personalized knowledge graph constructed according to the different conditions of each patient;

[0123] S105, maintaining the constructed evaluation model through a maintenance module, updating the constructed feature set by accumulating motion data of new patients, and updating the parameters of each evaluation model through a cross-validation method;

[0124] S106, the patient's score results are saved through the data storage management and transmission module. The rehabilitation therapist makes a personalized rehabilitation assessment report based on the training score, and sends the report, training plan and training opinions to the patient through the module, and the patient can freely review it.

[0125] Another object of the present invention is to provide a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the integrated method for physical and mental assessment and rehabilitation of stroke patients based on traditional Chinese medicine rehabilitation theory.

[0126] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, enables the processor to execute the steps of the integrated method for physical and mental assessment and rehabilitation of stroke patients based on traditional Chinese medicine rehabilitation theory.

[0127] Another object of the present invention is to provide an information data processing terminal, which is used to implement the integrated system for physical and mental assessment and rehabilitation of stroke patients based on traditional Chinese medicine rehabilitation theory.

[0128] The present invention is specifically implemented:

[0129] The present invention provides a personalized physical and mental health service system based on "unity of form and spirit". It adopts human skeleton detection technology, multi-source psychological analysis technology, multi-mode data fusion technology, intelligent digital assessment technology, and personalized configuration technology of TCM exercise therapy recuperation plan to carry out technical innovation research such as portable body data collection, multi-mode psychological data collection, digital physical and mental assessment, and personalized recuperation module configuration; integrates the above technologies to develop a personalized physical and mental health service system and prototype for stroke patients based on TCM exercise therapy.

[0130] Portable "shape" data collection: Use high-precision wide-angle cameras and other peripheral equipment to conveniently and contactlessly collect patients' designated behavioral movements during physical function assessment, and obtain real-time video of related human movements; introduce a convolutional neural network human key point detection and tracking system based on target recognition algorithm and re-identification algorithm to achieve high-precision real-time human skeleton coordinate extraction; use cutting-edge artificial intelligence algorithms (such as attention mechanism, time series-based exercise process correction method) to perform precision correction on the extracted human skeleton data, and provide accurate "shape" data for the next step of intelligent assessment.

[0131] Multi-modal "God" data collection: Use the retrospective personality development scale based on the field of traditional Chinese medicine to conduct a comprehensive psychological assessment of patients, including seven dimensions: courage quotient, self, willpower, thinking style, interpersonal relationships, sexual quotient and world view, life events, and parenting style; self-developed psychological data collection solution based on the llama large language model, collects and summarizes the results of patients' psychological questions and answers through AI, and provides accurate "God" data for the next step of intelligent assessment.

[0132] Intelligent digital physical and mental assessment: study the focus of each clinical assessment scale (such as the motor function screening FMS scale focuses on changes in whole-body motor function, and the hospital anxiety and depression scale focuses on changes in the subject's mood) and conduct corresponding feature value extraction algorithm research; develop an intelligent assessment model based on deep learning to achieve digital reconstruction of clinical assessment scales, thereby improving the efficiency and objectivity of the assessment. Based on the self-developed algorithm, the results of physical motor function assessment and psychological data are intelligently collected, the local knowledge map of traditional Chinese medicine exercise exercises is read, and personalized assessment results are output to build a personalized stroke map for patients.

[0133] Personalized TCM physical and mental rehabilitation: Combining the results of intelligent digital physical and mental assessment, a personalized recommendation configuration algorithm based on deep learning and a knowledge map of TCM exercise methods are adopted. According to the specific physical evaluation results, the TCM exercise methods covered by the knowledge map such as "Shi-style Twelve-character Health-preserving Exercise" and "Three Steps and Nine Methods for Neck Correction" are intelligently screened. At the same time, the specific psychological evaluation results are analyzed and screened for corresponding rehabilitation plans in the map, and intelligent rehabilitation plans are output. According to the principle of "combining mind, breath and body adjustment", a personalized stroke rehabilitation map for patients is drawn.

[0134] The present invention comprises six modules in total, and the functions of each module are as follows:

[0135] Equipment installation and data collection module: Build the assessment and rehabilitation task environment and install the equipment, carry out the corresponding assessment tasks based on the physical assessment scale and psychological question-and-answer system used, and identify the required data through screening.

[0136] Data preprocessing module: preprocess the collected data, including data filtering and segmentation, semantic recognition and organization.

[0137] Feature set construction module: calculate the corresponding kinematic features, classify the corresponding psychological analysis features, and dynamically screen the features to construct a patient-personalized feature set.

[0138] Evaluation module: Utilize the selected evaluation scale and the trained model to automatically score each task and calculate the total score for the kinematic feature set constructed by the above module to analyze the physical condition; utilize the local large language model to intelligently analyze the patient's psychological condition through the knowledge graph based on TCM theory; and retrieve relevant evaluation results based on the personalized knowledge graph constructed according to the different conditions of each patient.

[0139] Maintenance module: Maintain the constructed evaluation model, update the above constructed feature set by accumulating the motion data of new patients, and update the parameters of each evaluation model through cross-validation method.

[0140] Data storage, management and transmission module: saves the patient's score results. The rehabilitation therapist makes a personalized rehabilitation assessment report based on the training scores and sends the report, training plan and training opinions to the patient through this module. The patient can review them freely.

[0141] like Figure 3 The following is a further explanation of the above steps:

[0142] Step 1 (Equipment Installation and Data Collection)

[0143] Prior to this, a rehabilitation assessment area was divided and the equipment required for the test was installed. The patient stood in the planned area, watched a video explaining the action guidance and psychological questions / questionnaires of the rehabilitation assessment, and used a computer and a 2K intelligent wide-angle USB camera device to collect whole-body motion data. The computer psychological questions and answers interacted with the patient intelligently. Then, a customized motion data acquisition system (such as Figure 5 As shown in the figure, the required key joint tracking data is identified and screened, and then the psychological question-and-answer system based on the large language model is run to perform semantic processing on the collected data. After that, the data set is constructed according to the 10 movement tasks contained in the adopted Stroke Rehabilitation Assessment of Movement (STREAM) scale (taking the assessment task of STREAM as an example, the physiotherapist can change it according to the selected rating scale). Wide-angle USB camera, computer, table, and chair are necessary equipment for the experimental environment; pencils, cans, and baseballs are used as assessment aids (the aids are replaced according to the selected scale). When collecting data, it is necessary to ensure that most of the subject's movements can be captured by the lens, and the conversion between the world coordinate system and the camera coordinate system is realized, and the psychological assessment question-and-answer data is collected and useless information is screened by AI to ensure the reliability of the data. Therefore, the camera needs to be fixed in a suitable position and calibrated, and the camera needs to be set up on a tripod to ensure its viewing angle. The final report can be sent to the patient via the Internet, and the patient can choose to freely view the assessment results in daily life and undergo professional rehabilitation training in the hospital.

[0144] The Stroke Rehabilitation Assessment of Movement (STREAM) is a rating scale used to evaluate the recovery of voluntary movement and basic mobility in supine, sitting and standing positions after stroke. There are 30 sub-items, including 10 upper limb movements, 10 lower limb movements and 10 basic movements. Upper and lower limb movements are scored on a three-point scale, with each score ranging from 0 (unable to complete) to 2 (normal completion); basic movement is scored on a four-point scale, with each score ranging from 0 (unable to complete) to 3 (independent completion). The total score is 70 points. It is sensitive to the rehabilitation changes of stroke patients and has good intra-rater and inter-rater reliability and internal consistency.

[0145] The whole body motion data is captured by a wide-angle USB camera device, which is a 4-megapixel intelligent wide-angle USB camera developed by Hikvision. It has a fixed focal length of 2.8mm, a built-in omnidirectional microphone with a pickup distance of 5 meters, a resolution of 2560×1440, supports wide dynamic range, a range greater than 120dB, is suitable for backlight environment monitoring, and supports automatic focus. This device is the preferred device for high-precision input of real-time video data for patient motion assessment.

[0146] Prioritize the construction of the motion data set module. The data comes from the recruited stroke patients of various degrees. All participants were evaluated by senior rehabilitation therapists with rich medical practice experience using the STREAM scale and signed the relevant informed consent. When all the assessment tasks are completed, a total of 10 different motion data sets will be generated. These data will be organized into corresponding databases according to the task type for data accumulation and later maintenance work.

[0147] Psychological Q&A / questionnaire data were collected through a computerized questionnaire system. The Q&A or questionnaire questions were designed by professional therapists based on TCM rehabilitation theory to ensure the validity of data collection.

[0148] Step 2 (Data Preprocessing)

[0149] Using a wide-angle USB camera device, real-time video sources are collected for motor function assessment and input into the assessment system. First, the video source is processed by Gaussian filtering, and the feature data is intelligently screened through the self-developed improved YOLO algorithm. Then, real-time video is shot with a wide-angle camera, and the data of the corresponding required joints is extracted through the self-developed improved YOLO algorithm. The patient's motor thinking is intelligently analyzed to build a personalized sports data map. Using a computer intelligent psychological evaluation system, the patient's intelligent question and answer data is collected, and the data is intelligently summarized, keyword extracted, and other operations are performed according to the question and answer results to organize the data, and then a personalized psychological data map is built.

[0150] Step 3 (Feature set construction)

[0151] Preferably, relevant kinematic features are extracted from the video data preprocessed in the above steps, such as identifying 17 key points of each person's body. Each human target frame information contains six data: the horizontal coordinate of the human anchor point, the vertical coordinate, the width and height of the human prediction frame, the confidence of the human detection frame, and the category confidence. The inventors defined three main joint angles: elbow angle, shoulder angle, and leg angle. The elbow angle refers to the angle formed by the shoulder, elbow, and wrist in a two-dimensional plane; the shoulder angle is the angle formed by the elbow, shoulder, and hip in a two-dimensional plane; the leg angle is the angle formed by the hip, knee, and calf in a two-dimensional plane. The inventors proposed to extract the three types of angles (θ1, θ2, θ3) and the range of motion of the joints as calculation data and assist in subsequent evaluation. The inventors have developed a customized intelligent evaluation system for motion function (such as Figure 4 As shown), the system can complete feature extraction and subsequent motion evaluation.

[0152] Prior to this, the angles of the elbow, shoulder, and leg are used for rehabilitation movement scoring. The calculation formulas for the three types of angles are as follows:

[0153]

[0154] The geometric figures of the three types of angles are as follows Figure 4 As shown. In order to simplify the figure, the shoulder, elbow, wrist, hip, knee and calf are represented by W, M, P, D, T and X respectively. The shoulder-elbow line, elbow-wrist line, shoulder-hip line, hip-knee line and knee-calf line are represented by WM, MP, WD, DT and TX respectively. θ1, θ2 and θ3 represent the angle between the shoulder-elbow line and the elbow-wrist line, the angle between the elbow-shoulder line and the shoulder-hip line, and the angle between the hip-knee line and the knee-calf line.

[0155] Preferably, when constructing psychological feature values, it is necessary to use the llama large language model to extract the keywords involved in the local TCM knowledge graph, match the corresponding evaluation results and rehabilitation plans according to the key information, and mark the keywords that appear multiple times during the intelligent collection process, and finally obtain a database with the patient's personalized psychological characteristics.

[0156] Step 4 (Evaluation)

[0157] Prior to this, a neural network model is used as the core of the body assessment model. The neural network model consists of three parts: input layer, hidden layer, and output layer. The number of neurons in the input layer is the number of kinematic features, and the number of neurons in the output layer corresponds to several different scores for each task of STREAM. The assessment model can be automatically scored by AI.

[0158] Preferably, use the intelligent motor function assessment system ( Figure 4 ) Automatically score and sum the patient's exercise. First, the trained neural network model (one model for each exercise task) is used to automatically score according to the feature set constructed in step 3. Then, the final score of each task is summed to obtain the total score of this assessment, and the physical assessment content corresponding to the scale is output at the same time.

[0159] Preferably, a knowledge graph and an intelligent recommendation algorithm are used as the core of the psychological assessment model. The knowledge graph is constructed and stored based on the disease knowledge about stroke in the fields of traditional Chinese and Western medicine and the exercise methods of traditional Chinese medicine, including the semantic types of symptom classification, evaluation diagnosis, syndrome classification, psychological state, and intervention plans and the semantic relationships between them. Each semantic type has a corresponding description to illustrate the concept and characteristics of the semantic type. The intelligent psychological data collection system organizes data, implements keyword extraction based on natural language processing (NLP), performs text analysis tasks on the text data input by the user, thereby extracting entities, relationships, and semantics in the text, outputting keywords and summary content, matching the corresponding keyword results in the local traditional Chinese medicine knowledge graph, and outputting the final results through the intelligent recommendation algorithm.

[0160] Step 5 (Rehabilitation Program)

[0161] Preferably, the knowledge graph and intelligent recommendation algorithm are used as the core of the comprehensive physical and mental rehabilitation program formulation model. The final evaluation results are obtained by inputting the movement evaluation system (such as the score according to the scale, the symptoms and movement status description corresponding to the score) and the psychological evaluation system (the psychological state predicted by the keywords and intelligent recommendation algorithm), and the corresponding diagnosis and treatment plan is matched with the local knowledge graph through semantic analysis and keyword induction. The personalized physical and psychological map of the patient is constructed through the intelligent recommendation algorithm, and finally the rehabilitation plan integrating physical and mental rehabilitation is output, and the patient can freely view the report.

[0162] Step 6 (Maintenance)

[0163] Preferably, in order to improve the accuracy of the evaluation model in actual use, the evaluation system is configured to accumulate the patient's motion data, and the parameters of each evaluation model are regularly trained and updated using cross-validation.

[0164] Step 7 (Storage, Management and Transfer)

[0165] Preferably, rehabilitation therapists can log into an integrated assessment system (such as Figure 5 The system obtains patient information and uses the calculated scores, exercise data, and psychological assessment results as a reference (the system only displays the total score; if you want to view a single score, you need to click on a single record) to assess the physical and mental rehabilitation of stroke patients, compile a personalized rehabilitation assessment report, and remotely send the report and treatment opinions of the training plan to the patient through the data storage management and transmission system, so that the patient can further strengthen his or her exercises in a targeted manner.

[0166] The present invention is applicable to the sequelae and prognosis of neurological diseases such as stroke. TCM rehabilitation theory emphasizes promoting the recovery and reconstruction of nerve function and improving the patient's motor ability and self-care ability through acupuncture, massage, Chinese medicine and other methods. TCM rehabilitation exercises are used to guide the patient's psychology.

[0167] Table 1 STREAM stroke rehabilitation scoring system

[0168]

[0169]

[0170] It should be noted that the embodiments of the present invention can be implemented by hardware, software, or a combination of software and hardware. The hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. It can be understood by a person of ordinary skill in the art that the above-mentioned devices and methods can be implemented using computer executable instructions and / or contained in a processor control code, such as a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. Such code is provided on the carrier medium. The device and its modules of the present invention can be implemented by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., can also be implemented by software executed by various types of processors, and can also be implemented by a combination of the above-mentioned hardware circuits and software, such as firmware.

[0171] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with the technical field within the technical scope disclosed by the present invention and within the spirit and principle of the present invention should be covered by the protection scope of the present invention.

Claims

1. An integrated system for physical and mental assessment and rehabilitation of stroke patients based on TCM rehabilitation theory, characterized in that: include: Equipment installation and data acquisition module, data preprocessing module, feature set construction module, evaluation module, maintenance module, data storage management and transmission module; The equipment installation and data acquisition module is connected with the data preprocessing module and the data storage management and transmission module, and is used to build the assessment and rehabilitation task environment and install the equipment, carry out the corresponding assessment tasks according to the adopted physical assessment scale and psychological question-and-answer system, and identify the required data from the screening; The data preprocessing module is connected with the equipment installation and data collection module, the feature set construction module, the evaluation module, and the data storage management and transmission module, and is used to preprocess the above collected data, including data filtering and segmentation, semantic recognition and organization; The feature set construction module is connected with the data preprocessing module and the evaluation module to calculate the corresponding kinematic features, classify the corresponding psychological analysis features, and dynamically screen the features to construct a feature set personalized for the patient; The evaluation module is connected with the data preprocessing module, the feature set construction module, the maintenance module, and the data storage management and transmission module, and is used to use the selected evaluation scale to perform single automatic scoring and total score calculation for each task on the kinematic feature set constructed by the above modules through the trained model to analyze the physical condition; and to use the local large language model to intelligently analyze the patient's psychological condition through the knowledge graph based on traditional Chinese medicine theory; At the same time, the personalized knowledge graph constructed according to each patient's different conditions retrieves relevant evaluation results; A maintenance module, connected to the evaluation module, is used to maintain the constructed evaluation model, update the above-constructed feature set by accumulating the motion data of new patients, and update the parameters of each evaluation model by cross-validation method; The data storage management and transmission module is connected with the equipment installation and data acquisition module, the feature set construction module, and the evaluation module to save the patient's score results. The rehabilitation therapist makes a personalized rehabilitation evaluation report based on the training score, and sends the report, training plan and training opinions to the patient through this module, and the patient can review it freely.

2. The integrated system for physical and mental assessment and rehabilitation of stroke patients based on TCM rehabilitation theory as claimed in claim 1, characterized in that: The device installation and data acquisition module method is as follows: Demarcate the rehabilitation assessment area and install the equipment required for the test; the patient stands in the planned area, watches a video explaining the movement guidance and psychological questions / questionnaires of the rehabilitation assessment, and uses a computer and a 2K smart wide-angle USB camera device to collect whole-body movement data. The computer psychological questions and answers interact with the patient through intelligent questions and answers; Then, a customized motion data acquisition system was run on the computer to identify and filter the required key joint tracking data, and then a psychological question-answering system based on a large language model was run to perform semantic processing on the collected data. After that, a data set was constructed according to the 10 motion tasks included in the adopted stroke rehabilitation motor function scale. The data comes from recruited stroke patients of various degrees; all participants were evaluated by senior rehabilitation therapists with rich medical practice experience using the STREAM scale and signed the relevant informed consent; when all assessment tasks are completed, a total of 10 different motion data sets will be generated; these data will be organized into corresponding databases according to the task types for data accumulation and later maintenance work; Psychological Q&A / questionnaire data are collected through a computer questionnaire system; Q&A or questionnaire questions are designed by professional therapists based on traditional Chinese medicine rehabilitation theory.

3. The integrated system for physical and mental assessment and rehabilitation of stroke patients based on TCM rehabilitation theory as claimed in claim 1, characterized in that: The data preprocessing module method is as follows: Utilize a wide-angle USB camera device to capture real-time video feeds for motor function assessment and input into the assessment system; First, the video source is processed by Gaussian filtering, and the feature data is intelligently screened through the self-developed and improved YOLO algorithm; Then, real-time video is shot with a wide-angle camera, and data on the corresponding joints is extracted through the self-developed and improved YOLO algorithm, which intelligently analyzes the patient's movement thinking and builds a personalized movement data map. The computer intelligent psychological evaluation system is used to collect the patient's intelligent question and answer data, and the data is organized by intelligent summarization, keyword extraction and other operations based on the question and answer results, thereby building a personalized psychological data map.

4. The integrated system for physical and mental assessment and rehabilitation of stroke patients based on TCM rehabilitation theory as claimed in claim 1, characterized in that: The feature set construction module method is as follows: Extract relevant kinematic features from the video data preprocessed in the above steps, such as identifying 17 key points of each person's body. Each human target frame information contains six data: the horizontal coordinates and vertical coordinates of the human anchor point, the width and height of the human prediction frame, the confidence of the human detection frame, and the category confidence; Define three main joint angles: elbow angle, shoulder angle, and leg angle; the elbow angle refers to the angle formed by the shoulder, elbow, and wrist in a two-dimensional plane; the shoulder angle refers to the angle formed by the elbow, shoulder, and hip in a two-dimensional plane; the leg angle refers to the angle formed by the hip, knee, and calf in a two-dimensional plane; The angles of the elbow, shoulder, and leg are used for rehabilitation movement scoring; the calculation formulas for the three types of angles are as follows: The shoulder, elbow, wrist, hip, knee and calf are represented by W, M, P, D, T and X respectively; the shoulder-elbow line, elbow-wrist line, shoulder-hip line, hip-knee line and knee-calf line are represented by WM, MP, WD, DT and TX respectively; θ1, θ2 and θ3 represent the angle between the shoulder-elbow line and the elbow-wrist line, the angle between the elbow-shoulder line and the shoulder-hip line, and the angle between the hip-knee line and the knee-calf line.

5. The integrated system for physical and mental assessment and rehabilitation of stroke patients based on TCM rehabilitation theory as claimed in claim 1, characterized in that: The evaluation module method is as follows: A neural network model is used as the core of the physical assessment model; the neural network model consists of three parts: input layer, hidden layer, and output layer; the number of neurons in the input layer is the number of kinematic features, and the number of neurons in the output layer corresponds to several different scores for each task of STREAM. This assessment model can be automatically scored by AI; Use the intelligent motor function assessment system to automatically score and sum the patient's exercise. First, use the trained neural network model to automatically score according to the feature set constructed in step 3, then sum the final scores of each task to get the total score of this assessment, and output the physical assessment content corresponding to the scale. Knowledge graph and intelligent recommendation algorithm are used as the core of the psychological assessment model; the knowledge graph is constructed and stored based on the disease knowledge about stroke in the fields of traditional Chinese and Western medicine and traditional Chinese medicine exercise methods, including symptom classification, assessment diagnosis, syndrome classification, psychological state, semantic types of intervention plans and the semantic relationships between them. Each semantic type has a corresponding description to illustrate the concept and characteristics of the semantic type.

6. The integrated system for physical and mental assessment and rehabilitation of stroke patients based on TCM rehabilitation theory as claimed in claim 1, characterized in that: The data storage management and transmission module method is as follows: Rehabilitation therapists log into the integrated assessment system to obtain patient information, and use the calculated scores, movement data, and psychological assessment results as a reference to assess the physical and mental rehabilitation of stroke patients. They compile a personalized rehabilitation assessment report and send the report and treatment opinions of the training plan to the patient remotely through the data storage management and transmission system, so that the patient can further strengthen his or her exercises in a targeted manner.

7. A method for integrating physical and mental assessment and rehabilitation of stroke patients based on TCM rehabilitation theory, which implements the integrated physical and mental assessment and rehabilitation system for stroke patients based on TCM rehabilitation theory as claimed in any one of claims 1 to 6, characterized in that: The integrated method of physical and mental assessment and rehabilitation for stroke patients based on TCM rehabilitation theory includes: Step 1: Use the equipment installation and data collection module to build the assessment and rehabilitation task environment and install the equipment, carry out the corresponding assessment tasks based on the physical assessment scale and psychological question-and-answer system used, and identify the required data from the screening; Step 2: preprocess the collected data through a data preprocessing module, including data filtering and segmentation, semantic recognition and organization; Step 3, calculate the corresponding kinematic features through the feature set construction module, classify the corresponding psychological analysis features, and dynamically screen the features to construct a feature set personalized for the patient; Step 4: The evaluation module uses the selected evaluation scale to automatically score each task and calculate the total score of the kinematic feature set constructed by the above module through the trained model to analyze the physical condition; the local large language model is used to intelligently analyze the patient's psychological condition through the knowledge graph based on traditional Chinese medicine theory; at the same time, the personalized knowledge graph constructed according to the different conditions of each patient retrieves relevant evaluation results; Step 5, maintaining the constructed evaluation model through the maintenance module, updating the above constructed feature set by accumulating the motion data of new patients, and updating the parameters of each evaluation model through the cross-validation method; Step 6: The patient's score results are saved through the data storage management and transmission module. The rehabilitation therapist makes a personalized rehabilitation assessment report based on the training score and sends the report, training plan and training suggestions to the patient through this module. The patient can review them freely.

8. A computer device, characterized in that: The computer device includes a memory and a processor, the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the integrated method for physical and mental assessment and rehabilitation of stroke patients based on traditional Chinese medicine rehabilitation theory as described in claim 7.

9. A computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the steps of the integrated method for physical and mental assessment and rehabilitation of stroke patients based on traditional Chinese medicine rehabilitation theory as claimed in claim 7.

10. An information data processing terminal, characterized in that: The information data processing terminal is used to implement the integrated system for physical and mental assessment and rehabilitation of stroke patients based on traditional Chinese medicine rehabilitation theory as described in any one of claims 1-6.

Images