Personalized rehabilitation scheme optimization-oriented evaluation-intervention integrated system and equipment

Abstract

The invention provides an evaluation-intervention integrated system and equipment oriented to personalized rehabilitation scheme optimization. The evaluation-intervention integrated system comprises a personalized scheme recommendation module, a curative effect evaluation module and an intervention scheme optimization and adjustment module. The personalized scheme recommendation module generates a personalized rehabilitation treatment scheme by collecting multi-dimensional data and combining with a machine learning algorithm. And the curative effect evaluation module monitors and comprehensively evaluates the rehabilitation effect of the patient in real time, and generates a detailed curative effect report. And the intervention scheme optimization and adjustment module dynamically adjusts the treatment scheme and optimizes the treatment content, intensity and frequency according to the curative effect evaluation result and patient feedback. Through intelligent analysis and a feedback mechanism, an accurate and customized rehabilitation treatment scheme is provided for each patient, dynamic adjustment is performed according to a real-time evaluation result, and the rehabilitation treatment accuracy and the treatment effect are remarkably improved. The system also supports real-time data sharing and cooperation among patients, doctors and rehabilitation experts, and can perform cross-region and cross-professional personalized treatment management.

Description

Technical Field

[0001] This invention relates to the field of rehabilitation medicine technology, specifically to an integrated assessment-intervention system and device for optimizing personalized rehabilitation programs. Background Technology

[0002] With the continuous development of rehabilitation medicine, personalized treatment plans are playing an increasingly important role in patient rehabilitation. However, existing rehabilitation treatment plans are mostly based on doctors' experience, lacking a precise understanding of individual patient differences, leading to variations in treatment outcomes. Traditional rehabilitation methods typically rely on manual assessment and periodic adjustments, making real-time monitoring and dynamic optimization difficult, and failing to fully utilize modern intelligent technologies for personalized intervention. Furthermore, existing treatment plan recommendation systems, such as the "A Large-Scale Medical Data Knowledge Mining and Treatment Plan Recommendation System" proposed by Nanjing University of Aeronautics and Astronautics (e.g., existing patent document CN201911117826.5), lack intelligent treatment plan recommendation and automatic adjustment mechanisms, resulting in delayed feedback during treatment and an inability to make timely and effective adjustments based on the patient's actual condition. Therefore, how to provide real-time, personalized rehabilitation treatment plans through advanced intelligent means has become a pressing problem for rehabilitation medicine. Summary of the Invention

[0003] To address the shortcomings of the prior art, the present invention aims to provide an integrated assessment-intervention system and device for optimizing personalized rehabilitation programs, thereby solving the problems of traditional rehabilitation treatment programs lacking personalization, real-time feedback, and dynamic adjustment.

[0004] To achieve the above objectives, this invention provides an integrated assessment-intervention system for optimizing personalized rehabilitation programs, comprising a personalized program recommendation module, an efficacy assessment module, and an intervention program optimization and adjustment module. The personalized program recommendation module analyzes multi-dimensional patient data, including basic information, medical history, motor ability, psychological state, and real-time physiological indicators, to generate a tailored rehabilitation treatment plan. The efficacy assessment module monitors the patient's physiological indicators, motor ability data, and psychological state information in real time, assessing the rehabilitation effect and generating a detailed efficacy report during the rehabilitation treatment period provided by the personalized program recommendation module. The intervention program optimization and adjustment module dynamically adjusts and optimizes the rehabilitation treatment plan based on the assessment results of the efficacy assessment module and patient feedback during the aforementioned rehabilitation treatment period, ensuring the real-time adaptability and therapeutic effect of the plan.

[0005] Preferably, the personalized treatment plan recommendation module collects multi-dimensional data on the patient's basic information, medical history, motor ability, and psychological state, and uses a collaborative filtering algorithm combined with a large language model enhanced by RAG retrieval to intelligently generate personalized rehabilitation treatment plans. This module recommends the most suitable treatment content based on the patient's specific needs, considering their physical condition and rehabilitation goals.

[0006] Preferably, the sources of the multi-dimensional data include wearable devices, sensors, medical records, and questionnaires; after data collection is completed, the raw data is preprocessed; the preprocessing includes missing value imputation, outlier handling, and standardization;

[0007] For missing values ​​in multidimensional data, the KNN interpolation method is used to fill in the missing values ​​by finding the K samples that are closest to the missing data and calculating their weighted average.

[0008] For outliers in multi-dimensional data, the Isolation Forest algorithm is used to detect and remove outliers by constructing a random tree model.

[0009] Z-score standardization was used to transform the multidimensional data, after missing value imputation and outlier handling, into a distribution with a mean of 0 and a standard deviation of 1.

[0010] Preferably, the personalized treatment plan recommendation module analyzes the rehabilitation plans of similar patients using a collaborative filtering algorithm to recommend a personalized treatment plan for the current patient; this method utilizes the similarity between patients to find a suitable treatment plan for the current patient, as detailed below:

[0011] First, construct a patient-feature matrix, where each row represents a patient and each column represents a feature;

[0012] Next, cosine similarity is used to calculate the similarity between patients; the formula for calculating cosine similarity can be expressed as:

[0013]

[0014] Where A and B are the feature vectors of the first patient A and the second patient B, respectively; based on the similarity score, the K nearest neighbors algorithm is used to find the K most similar patients to the current patient;

[0015] Finally, to generate personalized and precise treatment plans, a recommendation method combining a Large Language Model (LLM) with Retrieval Augmentation (RAG) and a rule base is introduced. This method first utilizes a large language model (such as GPT-4) to perform semantic understanding and contextual modeling on multi-dimensional data of similar patients. Specifically, the input to the large language model is the current patient's feature data and treatment plan data from K similar patients. The large language model transforms the patient's feature data (such as age, medical history, motor ability, and psychological state) into a high-dimensional semantic representation and captures the complex relationships between data, outputting a semantic understanding vector of the patient's current state. Second, this method employs the RAG framework to enhance the knowledge retrieval capability of the large language model. Specifically, after obtaining the semantic understanding vector from the patient's multi-dimensional data through the large language model, it accesses external knowledge bases (such as medical literature, clinical guidelines, and rehabilitation case databases) to retrieve relevant knowledge fragments matching the patient's features using similarity. These retrieved knowledge fragments are then input into the large language model through a secondary input. Finally, the large language model uses the patient's multi-dimensional data and the secondary input retrieval augmentation data as input to output a personalized plan containing treatment content, intensity, and frequency.

[0016] Preferably, the efficacy assessment module uses a variety of intelligent monitoring devices (including wearable devices, sensors, and imaging equipment) to monitor the patient's physiological indicators (including heart rate, blood pressure, range of motion, and muscle strength) as well as the patient's motor ability and psychological state during the rehabilitation process in real time. These data are then input into a pre-trained Transformer efficacy analysis model to comprehensively evaluate these indicators and generate a quantitative efficacy report that reflects the patient's rehabilitation progress.

[0017] Preferably, the intervention plan optimization and adjustment module uses a feedback mechanism to utilize real-time feedback data generated by the patient during treatment (including pain level, treatment discomfort, and mobility), inputs the patient's feedback data and efficacy evaluation results into a pre-trained reinforcement learning model, and intelligently adjusts the treatment plan to ensure the real-time adaptability and optimization effect of the plan.

[0018] Preferably, the personalized treatment plan recommendation module comprehensively considers the individual differences of patients based on their historical treatment data, current status, and medical knowledge base, and tailors a rehabilitation plan for each patient. This includes in-depth analysis of the patient's health data, treatment records, and rehabilitation history data, and intelligent decision-making in conjunction with medical literature and clinical guidelines.

[0019] Preferably, the system also includes a cloud platform module, which enables real-time synchronization and sharing of treatment data (including rehabilitation treatment plans, efficacy evaluation results, and patient feedback data) from patients, doctors, and rehabilitation specialists. Through cloud computing technology, all data is stored in the cloud, allowing users to access patient rehabilitation records and evaluation results anytime, anywhere via the internet. This enables cross-regional telemedicine services and interdisciplinary collaboration, thereby optimizing treatment plans and improving rehabilitation outcomes.

[0020] Preferably, the efficacy assessment module includes an assessment sub-module based on standardized assessment scales (such as the Barthel Index, Fugl-Meyer Score, and Modified Ashworth Score). The standardized scales are used to assess the patient's functional recovery, motor ability, and daily living abilities. Combined with physiological data collected by intelligent monitoring devices, efficacy assessment results are generated periodically to provide quantitative analysis of rehabilitation progress.

[0021] Preferably, the intervention plan optimization and adjustment module has self-learning capabilities, enabling it to dynamically optimize the treatment plan using adaptive algorithms based on different treatment cycles and patient feedback data. By analyzing the treatment effects and feedback data of patients at different time points, the system can automatically adjust parameters such as the content, intensity, and frequency of the treatment plan, and provide precise intervention based on the individualized needs of patients.

[0022] Preferably, the system further includes a patient terminal module, which can interact with the patient's smart devices (such as smartwatches, motion sensors, etc.) to display personalized rehabilitation plans, efficacy evaluation results, and optimized rehabilitation treatment plans in real time. It also allows patients to record daily rehabilitation training, provide symptom feedback, and evaluate treatment plans, further enhancing patient engagement and rehabilitation effectiveness.

[0023] The above-mentioned method for implementing an integrated assessment-intervention system for optimizing personalized rehabilitation programs includes the following steps:

[0024] (S1) Personalized treatment plan generation: Collect multi-dimensional data of patients, including basic information, medical history, motor ability, psychological state and real-time physiological indicators; generate personalized rehabilitation treatment plans by combining collaborative filtering algorithm with large language model of retrieval enhancement generation (RAG);

[0025] (S2) Real-time efficacy assessment: During the rehabilitation treatment, the patient's physiological indicators, motor ability and psychological state data are acquired in real time through intelligent monitoring equipment, and input into the pre-trained Transformer model for comprehensive assessment to generate a quantitative efficacy report.

[0026] (S3) Dynamic treatment plan optimization: Based on the efficacy evaluation results and patient feedback data, the content, intensity and frequency parameters of the treatment plan are dynamically adjusted using a reinforcement learning model.

[0027] Furthermore, the multi-dimensional data mentioned in step (S1) originates from wearable devices, sensors, medical records, and questionnaires, and the raw data undergoes preprocessing, including:

[0028] (S11) Use KNN interpolation to fill in missing values;

[0029] (S12) Use the isolated forest algorithm to detect and remove outliers;

[0030] (S13) The data is transformed into a distribution with a mean of 0 and a standard deviation of 1 by Z-score standardization.

[0031] Furthermore, the specific steps for combining the collaborative filtering algorithm with the RAG-enhanced large language model in step (S1) include:

[0032] (S14) Construct the patient-feature matrix and calculate the cosine similarity between patients;

[0033] (S15) Similar patients are screened using the K-nearest neighbor algorithm;

[0034] (S16) Input the current patient characteristics and similar patient plans into the large language model to generate semantic understanding vectors;

[0035] (S17) Retrieve external medical knowledge bases based on the RAG framework, and input the matched knowledge fragments into the large language model again;

[0036] (S18) Output a personalized plan that includes treatment content, intensity, and frequency.

[0037] Furthermore, the external medical knowledge base includes clinical guidelines, medical literature, and rehabilitation case databases, and the retrieval and matching uses a cosine similarity algorithm.

[0038] Furthermore, the efficacy assessment described in step (S2) also incorporates standardized scales, including the Barthel Index, Fugl-Meyer Score, or Modified Ashworth Score, to generate a quantitative rehabilitation progress report; the reinforcement learning model described in step (S3) uses the efficacy assessment results and patient feedback as the state space and the treatment plan parameters as the action space, and optimizes the strategy through a reward function.

[0039] The present invention also provides an integrated assessment-intervention device for optimizing personalized rehabilitation programs, including a memory, a processor, and a computer program stored in the memory and executable on the processor, and further including the integrated assessment-intervention system for optimizing personalized rehabilitation programs; and the processor implements the functions of the integrated assessment-intervention system for optimizing personalized rehabilitation programs when executing the computer program.

[0040] Compared with the prior art, the beneficial effects of the present invention are reflected in:

[0041] (1) This invention provides a personalized rehabilitation treatment plan based on intelligent data analysis and real-time feedback. Compared with traditional treatment methods that rely on doctors' experience, it can intelligently generate and adjust personalized treatment plans according to the patient's specific condition, physical condition and rehabilitation needs, thereby improving the accuracy and effectiveness of treatment.

[0042] (2) The efficacy assessment module can reflect the patient's rehabilitation progress in a timely manner by monitoring the patient's physiological indicators and motor ability in real time, and overcome the problems of delayed assessment and untimely adjustment in traditional treatment by dynamically optimizing and adjusting the treatment plan.

[0043] (3) The intervention program optimization and adjustment module can automatically adjust the treatment content, intensity and frequency based on patient feedback and real-time efficacy data, ensuring the continuous adaptability and personalized optimization of the treatment program, significantly improving the rehabilitation effect and shortening the rehabilitation cycle. Attached Figure Description

[0044] Figure 1 This is a system architecture diagram of an integrated assessment-intervention system for optimizing personalized rehabilitation programs, as shown in the embodiment.

[0045] Figure 2 This is a flowchart of the multimodal rehabilitation data preprocessing process in the embodiment.

[0046] Figure 3 Flowchart for generating personalized rehabilitation plans enhanced by RAG. Detailed Implementation

[0047] To further understand the structure, features, and other objectives of the present invention, a detailed description is provided below with reference to the accompanying drawings. The embodiments illustrated in these drawings are for illustrative purposes only and are not intended to limit the scope of the invention.

[0048] Figure 1This is a system architecture diagram of an integrated assessment-intervention system for personalized rehabilitation program optimization according to the present invention. The integrated assessment-intervention system for personalized rehabilitation program optimization includes a personalized program recommendation module, an efficacy evaluation module, and an intervention program optimization and adjustment module. The personalized program recommendation module generates a customized rehabilitation treatment plan by analyzing multi-dimensional patient data, including the patient's basic information, medical history, motor ability, psychological state, and real-time physiological indicators. The efficacy evaluation module assesses the rehabilitation effect and generates a detailed efficacy report during the rehabilitation treatment period provided by the personalized program recommendation module by real-time monitoring of the patient's physiological indicators, motor ability data, and psychological state information. The intervention program optimization and adjustment module dynamically adjusts and optimizes the rehabilitation treatment plan based on the evaluation results of the efficacy evaluation module and patient feedback during the patient's rehabilitation treatment to ensure the real-time adaptability and treatment effect of the plan.

[0049] The personalized treatment plan recommendation module collects multi-dimensional data on the patient's basic information, medical history, motor ability, and psychological state. It then uses a collaborative filtering algorithm combined with a large language model enhanced by RAG retrieval to intelligently generate personalized rehabilitation treatment plans. Based on the patient's specific needs, considering their physical condition and rehabilitation goals, this module recommends the most suitable treatment content.

[0050] The sources of the multi-dimensional data include wearable devices, sensors, medical records, and questionnaires; after the data collection is completed, the raw data is preprocessed; the preprocessing includes missing value imputation, outlier handling, and standardization;

[0051] like Figure 2 For missing values ​​in multidimensional data, the KNN interpolation method is used to fill in the missing values ​​by finding the K samples that are closest to the missing data and calculating their weighted average.

[0052] For outliers in multi-dimensional data, the Isolation Forest algorithm is used to detect and remove outliers by constructing a random tree model.

[0053] Z-score standardization was used to transform the multidimensional data, after missing value imputation and outlier handling, into a distribution with a mean of 0 and a standard deviation of 1.

[0054] The personalized treatment plan recommendation module analyzes the rehabilitation plans of similar patients using a collaborative filtering algorithm to recommend a personalized treatment plan for the current patient. This method utilizes the similarity between patients to find a suitable treatment plan for the current patient, as detailed below:

[0055] First, construct a patient-feature matrix, where each row represents a patient and each column represents a feature;

[0056] Next, cosine similarity is used to calculate the similarity between patients; the formula for calculating cosine similarity can be expressed as:

[0057]

[0058] Where A and B are the feature vectors of the first patient A and the second patient B, respectively; based on the similarity score, the K nearest neighbors algorithm is used to find the K most similar patients to the current patient;

[0059] Finally, as Figure 3 To generate personalized and precise treatment plans, this method introduces a Large Language Model (LLM) combined with Retrieval Augmentation (RAG) and rule-based recommendation. First, the method utilizes a LLM (such as GPT-4) to perform semantic understanding and contextual modeling on multi-dimensional data of similar patients. Specifically, the LLM is input to the current patient's feature data and treatment plans from K similar patients. The LLM transforms the patient's feature data (such as age, medical history, motor ability, and psychological state) into a high-dimensional semantic representation and captures the complex relationships between data, outputting a semantic understanding vector of the patient's current state. Second, the method employs the RAG framework to enhance the knowledge retrieval capabilities of the LLM. Specifically, after obtaining the semantic understanding vector from the patient's multi-dimensional data through the LLM, it accesses external knowledge bases (such as medical literature, clinical guidelines, and rehabilitation case databases) to retrieve relevant knowledge fragments matching the patient's features using similarity. These retrieved knowledge fragments are then input into the LLM through a secondary input method. Finally, the LLM takes the patient's multi-dimensional data and the secondary input retrieval augmentation data as input and outputs a personalized treatment plan containing treatment content, intensity, and frequency.

[0060] The efficacy assessment module uses a variety of intelligent monitoring devices (including wearable devices, sensors, and imaging equipment) to monitor the patient's physiological indicators (including heart rate, blood pressure, range of motion, and muscle strength) as well as the patient's motor ability and psychological state during the rehabilitation process in real time. These data are then input into a pre-trained Transformer efficacy analysis model to comprehensively evaluate these indicators and generate a quantitative efficacy report that reflects the patient's rehabilitation progress.

[0061] The intervention plan optimization and adjustment module uses a feedback mechanism to utilize real-time feedback data generated by the patient during treatment (including pain level, treatment discomfort, and mobility). The patient's feedback data and efficacy evaluation results are input into a pre-trained reinforcement learning model to intelligently adjust the treatment plan, ensuring the real-time adaptability and optimization effect of the plan.

[0062] The personalized treatment plan recommendation module takes into account the patient's historical treatment data, current status, and medical knowledge base, and comprehensively considers the individual differences of the patient to tailor a rehabilitation plan for each patient. This includes in-depth analysis of the patient's health data, treatment records, and rehabilitation history data, and intelligent decision-making in conjunction with medical literature and clinical guidelines.

[0063] It also includes a cloud platform module, which enables real-time synchronization and sharing of treatment data (including rehabilitation treatment plans, efficacy evaluation results, and patient feedback data) among patients, doctors, and rehabilitation specialists. Through cloud computing technology, all data is stored in the cloud, allowing users to access patient rehabilitation records and evaluation results anytime, anywhere via the internet. This facilitates cross-regional telemedicine services and interdisciplinary collaboration, thereby optimizing treatment plans and improving rehabilitation outcomes.

[0064] The efficacy assessment module includes an assessment sub-module based on standardized assessment scales (such as the Barthel Index, Fugl-Meyer Score, and Modified Ashworth Score). The standardized scales are used to assess the patient's functional recovery, motor ability, and daily living abilities. Combined with physiological data collected by intelligent monitoring devices, efficacy assessment results are generated periodically, providing quantitative analysis of rehabilitation progress.

[0065] The intervention plan optimization and adjustment module has self-learning capabilities, enabling it to dynamically optimize treatment plans based on different treatment cycles and patient feedback data through adaptive algorithms. By analyzing the treatment effects and feedback data of patients at different time points, the system can automatically adjust parameters such as the content, intensity, and frequency of the treatment plan, and provide precise intervention based on the individualized needs of patients.

[0066] The system also includes a patient terminal module, which can interact with the patient's smart devices (such as smartwatches, motion sensors, etc.) to display personalized rehabilitation plans, efficacy evaluation results, and optimized rehabilitation treatment plans in real time. It also allows patients to record daily rehabilitation training, provide symptom feedback, and evaluate treatment plans, further enhancing patient engagement and rehabilitation effectiveness.

[0067] This invention provides precise and customized rehabilitation treatment plans for each patient through intelligent analysis and feedback mechanisms, and dynamically adjusts them based on real-time assessment results, significantly improving the accuracy and effectiveness of rehabilitation treatment. The system also supports real-time data sharing and collaboration among patients, doctors, and rehabilitation experts, enabling personalized treatment management across regions and specialties, and has broad application prospects and practical value.

Claims

1. An integrated assessment-intervention system for optimizing personalized rehabilitation programs, characterized in that, It includes a personalized treatment plan recommendation module, a treatment efficacy assessment module, and an intervention plan optimization and adjustment module; The personalized treatment plan recommendation module analyzes multi-dimensional patient data, including basic information, medical history, motor ability, psychological state, and real-time physiological indicators, to generate a customized rehabilitation treatment plan. The efficacy assessment module monitors patients’ physiological indicators, motor ability data and psychological state information in real time. During the rehabilitation treatment period when patients adopt the rehabilitation treatment plan given by the personalized plan recommendation module, the module assesses the rehabilitation effect and generates a detailed efficacy report. The intervention program optimization and adjustment module dynamically adjusts and optimizes the rehabilitation treatment plan based on the evaluation results of the efficacy assessment module and patient feedback during the patient's rehabilitation treatment, so as to ensure the real-time adaptability of the plan and the treatment effect.

2. The integrated assessment-intervention system for optimizing personalized rehabilitation programs according to claim 1, characterized in that, The personalized treatment plan recommendation module collects multi-dimensional data on the patient's basic information, medical history, motor ability, and psychological state. It then uses a collaborative filtering algorithm combined with a large language model enhanced by RAG retrieval to intelligently generate personalized rehabilitation treatment plans. Based on the patient's specific needs, considering their physical condition and rehabilitation goals, it recommends the most suitable treatment content.

3. The integrated assessment-intervention system for optimizing personalized rehabilitation programs according to claim 1, characterized in that, The sources of the multi-dimensional data include wearable devices, sensors, medical records, and questionnaires; after data collection, the raw data is preprocessed. Preprocessing includes missing value imputation, outlier handling, and standardization; For missing values ​​in multidimensional data, the KNN interpolation method is used to fill in the missing values ​​by finding the K samples that are closest to the missing data and calculating their weighted average. For outliers in multi-dimensional data, the Isolation Forest algorithm is used to detect and remove outliers by constructing a random tree model. Z-score standardization was used to transform the multidimensional data, after missing value imputation and outlier handling, into a distribution with a mean of 0 and a standard deviation of 1.

4. The integrated assessment-intervention system for optimizing personalized rehabilitation programs according to claim 1, characterized in that, The personalized treatment plan recommendation module analyzes the rehabilitation plans of similar patients using a collaborative filtering algorithm to recommend a personalized treatment plan for the current patient; it utilizes the similarity between patients to find a suitable treatment plan for the current patient, as detailed below: First, construct a patient-feature matrix, where each row represents a patient and each column represents a feature; Next, cosine similarity is used to calculate the similarity between patients; the formula for calculating cosine similarity can be expressed as: Where A and B are the feature vectors of the first patient A and the second patient B, respectively; based on the similarity score, the K nearest neighbors algorithm is used to find the K most similar patients to the current patient; Finally, to generate personalized and precise treatment plans, a recommendation method combining a Large Language Model (LLM) with Retrieval Augmentation Generation (RAG) and a rule base is introduced: First, the LLM is used to perform semantic understanding and contextual modeling on multi-dimensional data of similar patients. The input to the LLM is the current patient's feature data and the treatment plan data of K similar patients. The LLM transforms the patient's feature data into a high-dimensional semantic representation and captures the complex relationships between the data, outputting a semantic understanding vector of the patient's current state. Second, the RAG framework is used to enhance the knowledge retrieval capability of the LLM. After obtaining the semantic understanding vector from the patient's multi-dimensional data through the LLM, relevant knowledge fragments matching the patient's features are retrieved by accessing an external knowledge base using similarity retrieval. These retrieved knowledge fragments are then input into the LLM through a secondary input method. Finally, the LLM takes the patient's multi-dimensional data and the secondary input retrieval augmentation data as input and outputs a personalized plan containing treatment content, intensity, and frequency.

5. The integrated assessment-intervention system for optimizing personalized rehabilitation programs according to claim 1, characterized in that, The efficacy assessment module uses various intelligent monitoring devices to monitor the patient's physiological indicators, as well as their motor abilities and psychological state during the rehabilitation process in real time. This data is then input into a pre-trained Transformer efficacy analysis model to comprehensively evaluate these indicators and generate a quantitative efficacy report that reflects the patient's rehabilitation progress.

6. The integrated assessment-intervention system for optimizing personalized rehabilitation programs according to claim 1, characterized in that, The intervention plan optimization and adjustment module has self-learning capabilities. Based on different treatment cycles and patient feedback data, it dynamically optimizes the treatment plan through adaptive algorithms. By analyzing the treatment effects and feedback data of patients at different time points, it automatically adjusts the content, intensity, and frequency parameters of the treatment plan and provides precise intervention based on the individual needs of patients. The intervention plan optimization and adjustment module inputs patient feedback data and efficacy evaluation results into a pre-trained reinforcement learning model to intelligently adjust the treatment plan, ensuring the real-time adaptability and optimization effect of the plan.

7. The integrated assessment-intervention system for optimizing personalized rehabilitation programs according to claim 1, characterized in that, The personalized treatment plan recommendation module takes into account the patient's historical treatment data, current status, and medical knowledge base, and comprehensively considers the individual differences of the patient to tailor a rehabilitation plan for each patient. This includes in-depth analysis of the patient's health data, treatment records, and rehabilitation history data, and intelligent decision-making in conjunction with medical literature and clinical guidelines.

8. The integrated assessment-intervention system for optimizing personalized rehabilitation programs according to claim 1, characterized in that, The efficacy assessment module includes an assessment sub-module based on standardized assessment scales. The standardized scales are used to assess the patient's functional recovery, motor ability, and daily living ability. Combined with physiological data collected by intelligent monitoring devices, efficacy assessment results are generated periodically, providing quantitative analysis of rehabilitation progress.

9. The integrated assessment-intervention system for optimizing personalized rehabilitation programs according to claim 1, characterized in that, It also includes a cloud platform module, which can synchronize and share treatment data of patients, doctors and rehabilitation experts in real time. Through cloud computing technology, all data is stored in the cloud, and users can access patients' rehabilitation records and assessment results anytime and anywhere via the Internet, and realize cross-regional telemedicine services and interdisciplinary collaboration, thereby optimizing treatment plans and improving rehabilitation outcomes. It also includes a patient terminal module, which can interact with the patient's smart device to display personalized rehabilitation plans, efficacy evaluation results, and optimized rehabilitation treatment plans in real time. It also allows patients to record daily rehabilitation training, provide symptom feedback, and evaluate treatment plans, further enhancing patient engagement and rehabilitation outcomes.

10. An integrated assessment-intervention device for optimizing personalized rehabilitation programs, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, It also includes the integrated assessment-intervention system for optimizing personalized rehabilitation programs; and the processor, when executing the computer program, implements the functions of the integrated assessment-intervention system for optimizing personalized rehabilitation programs as described in any one of claims 1 to 9.

Images