Rehabilitation intelligent assessment system for rheumatism based on ICF concept and classification system

Abstract

The present application relates to the field of medical rehabilitation information technology, in particular to a rheumatism rehabilitation intelligent evaluation system based on ICF concept and classification system, comprising: a multi-source data module for collecting and integrating multi-dimensional data; a dynamic evaluation module based on ICF theory, constructing a dynamic weight distribution model, updating the evaluation results and radar chart according to the multi-dimensional data; a data feature module based on latent profile analysis algorithm, performing high-dimensional feature dimension reduction processing according to the evaluation results and radar chart, and identifying hidden subtypes; a causal intervention module based on the hidden subtypes, wherein the present application, in use, applies the multi-dimensional dynamic coupling evaluation of ICF theory, the subtype identification of latent profile analysis and the causal modeling of structural equation model to the rheumatism rehabilitation decision-making in a nested manner, so as to solve the problems of single evaluation dimension, static and isolated data and difficulty in adapting to individual differences, and improve the accuracy of intervention strategy and the rehabilitation effect of patients.

Description

Technical Field

[0001] This invention relates to the field of medical rehabilitation information technology, specifically to an intelligent assessment system for rheumatoid arthritis rehabilitation based on the ICF concept and classification system. Background Technology

[0002] The International Classification of Functioning, Disability and Health (ICF) is a classification system published by the World Health Organization. It provides a standardized language and framework for describing health status outcomes, comprehensively assessing an individual's health status from three dimensions: physical function and structure, activity, and participation, rather than just focusing on the disease itself. Applying the ICF framework in the rehabilitation assessment of rheumatic diseases can more comprehensively and accurately reflect the degree of functional impairment, the impact on daily living activities, and the problems faced by patients in social participation.

[0003] Patent publication number CN119274800A describes in its specification that "This invention relates to the field of medical rehabilitation technology, specifically to a remote intelligent rehabilitation assessment system. The system includes a rehabilitation monitoring module, a remote service module, an interactive feedback module, and an intelligent assessment module; the rehabilitation monitoring module includes an emotional signal monitoring module, a physiological signal monitoring module, and a gamified performance monitoring module, used to monitor the patient's emotional state, physiological state, and reaction status; the remote service module connects the rehabilitation monitoring module, the interactive feedback module, and the intelligent assessment module; the interactive feedback module is used for information exchange and feedback between the patient, rehabilitation therapist, and the remote service module; the intelligent assessment module is used to evaluate the remote service..." The rehabilitation index data and optimized rehabilitation information obtained from the module are analyzed and evaluated. This invention achieves remote, comprehensive, and refined intelligent assessment of patient rehabilitation by comprehensively monitoring the patient's emotional, physiological, and reactive states. While the aforementioned technology, based on multi-source data fusion and dynamic weight allocation, achieves precise and personalized intelligent rehabilitation management through collaborative assessment of emotional, physiological, and reactive dimensions, combined with gamified testing and adaptive algorithms, traditional assessment systems struggle to balance the dynamic coupling of patients' multidimensional health states and group heterogeneity. In the assessment and decision-making process, the single assessment dimension, static and isolated data, and difficulty in adapting to individual differences lead to crude intervention strategies and low effectiveness.

[0004] In conclusion, the development of an intelligent assessment system for rheumatic disease rehabilitation based on the ICF concept and classification system remains a key issue that urgently needs to be addressed in the field of medical rehabilitation information technology. Summary of the Invention

[0005] The purpose of this invention is to address the problems in existing technologies where traditional assessment systems struggle to balance the dynamic coupling of patients' multidimensional health status with group heterogeneity, and where assessment and decision-making processes suffer from single assessment dimensions, static and isolated data, and difficulty in adapting to individual differences, resulting in crude intervention strategies and low effectiveness.

[0006] To achieve the above objectives, this invention provides an intelligent assessment system for rheumatic disease rehabilitation based on the ICF concept and classification system, comprising:

[0007] A multi-source data module is used to collect and integrate multi-dimensional data;

[0008] The dynamic evaluation module, based on ICF theory, constructs a dynamic weight allocation model and updates the evaluation results and radar chart according to the multi-dimensional data.

[0009] The data feature module, based on the latent profile analysis algorithm, performs high-dimensional feature dimensionality reduction processing based on the evaluation results and radar chart to identify the hidden subtype;

[0010] The causal intervention module, based on the aforementioned hidden subtype, utilizes structural equation modeling to construct a causal path network and combines it with knowledge graph technology to generate a targeted intervention strategy library.

[0011] The closed-loop feedback module is used to establish a two-way feedback mechanism between evaluation and decision-making, and to input the feedback into the dynamic evaluation module to optimize the weight allocation and iteratively update the causal path network.

[0012] Beneficial effects

[0013] Compared with known public technologies, the technical solution provided by this invention has the following beneficial effects:

[0014] When used, this invention nests the multidimensional dynamic coupling assessment of ICF theory with the subtype identification of potential profile analysis and the causal modeling of structural equation modeling, and applies it to the decision-making of rheumatic disease rehabilitation. This helps to solve the problems of traditional assessment having a single dimension, static and isolated data, and difficulty in adapting to individual differences, and is conducive to improving the accuracy of intervention strategies and the rehabilitation effect of patients.

[0015] When in use, this invention learns and optimizes causal intervention pathways for different patient subtypes, making the pushed rehabilitation plans more personalized and effective. Patients upload their daily status at home through smart terminals, and the system updates the weights and causal models in real time. Doctors adjust their treatment strategies accordingly, forming an adaptive and intelligent rehabilitation closed loop, which helps improve rehabilitation outcomes and patient satisfaction, and reduces the waste of medical resources. Attached Figure Description

[0016] Figure 1This is a system diagram of the intelligent assessment system for rheumatic disease rehabilitation based on the ICF concept and classification system of this invention.

[0017] Figure 2 This is a diagram of the software interface of the intelligent assessment system for rheumatic disease rehabilitation based on the ICF concept and classification system of this invention.

[0018] Figure 3 This is a diagram showing the completion rate of accurate assessments for ICF patients.

[0019] Figure 4 This is a status interface diagram based on ICF patient precision assessment. Detailed Implementation

[0020] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0021] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0022] See Figure 1-4 This application will describe the invention in further detail with reference to the accompanying drawings:

[0023] Example:

[0024] like Figure 1 As shown, this invention provides an intelligent assessment system for rheumatic disease rehabilitation based on the ICF concept and classification system, including:

[0025] A multi-source data module is used to collect and integrate multi-dimensional data;

[0026] Furthermore, the operation process of the multi-source data module includes:

[0027] The system is used to collect and integrate multi-dimensional data. Through an API interface protocol, it connects with hospital information systems, image archiving, and communication systems to collect patients' biochemical test data, imaging data, and medication records in real time. Simultaneously, a patient-side interactive interface is developed to support self-assessment data entry based on the five dimensions of the ICF framework: disease activity, physical function, quality of life, social support, and work efficiency. The patient-side app interface supports the completion of the ICF five-dimensional questionnaire, generating the following self-assessment data vector:

[0028]

[0029] In the formula, This represents a five-dimensional evaluation vector within the ICF framework. This represents a set with five elements. Represents any number of Evaluation scores in each dimension, The score range for each dimension is from 0 to 1, where 0 represents very poor and 1 represents excellent. All scales are standardized and mapped to the [0,1] interval. A data fusion algorithm is constructed using DS evidence theory. For data conflict scenarios, a manual review process is initiated through outlier cross-validation rules. Trust assignments are provided from both data sources, and the synthesized evidence is as follows:

[0030]

[0031] In the formula, Indicates the proposition The combined trust distribution function value, Indicates the first source of evidence pair subset The trust distribution function value, Indicates the second source of evidence pair subset The trust distribution function value, Indicates all In a combination, the intersection equals All pairs Summation above, This represents the normalization factor over all non-conflicting parts. Represents all combinations of subsets when the intersection is an empty set. Summation, Representing a subset and The intersection is empty, expand to When there are multiple data sources, construct the fusion process:

[0032]

[0033] In the formula, This represents the trust distribution function after fusion. This refers to a method of fusing multiple pieces of evidence using the combination rules of the Dempster-Shafer evidence theory. Indicates the first to the second The original evidence trust distribution function.

[0034] Specifically, the multi-source data module of this system collects patients' biochemical test data, medical imaging data, and medication records in real time through API interfaces with medical information platforms such as hospital information systems, image archiving, and communication systems, ensuring the comprehensiveness and timeliness of clinical data. At the same time, a patient-side app was developed, and a self-assessment questionnaire with five dimensions (disease activity, physical function, quality of life, social support, and work efficiency) was designed based on the ICF theoretical framework. Patients can directly enter their subjective feelings, generating standardized five-dimensional assessment vectors. The scale scores are uniformly mapped to the 0 to 1 interval, facilitating unified analysis and enhancing the completeness and multi-dimensional understanding of information. Based on the standardized patient self-assessment scale, subjective and objective data are effectively combined. The introduction of evidence theory improves the reliability and robustness of data fusion. Combined with a manual review mechanism, it facilitates the timely identification and processing of abnormal data.

[0035] The dynamic evaluation module, based on ICF theory, constructs a dynamic weight allocation model and updates the evaluation results and radar chart according to the multi-dimensional data.

[0036] Furthermore, the operational process of the dynamic evaluation module includes:

[0037] Based on ICF theory, the five dimensions are decomposed into a set of quantifiable sub-indicators, and a dynamic weight allocation model is constructed. An initial weight is set for one of the dimensions, and a non-linear update is triggered when fluctuations in a related sub-indicator are detected.

[0038]

[0039] In the formula, Indicates at a point in time At that time, the first Dynamic weight values ​​for each dimension Represents the natural constant. This indicates an adjustable sensitivity coefficient. Indicates time At that time, the first Changes in each dimension Five core dimensions of ICF To sum, This represents the exponential weighted value of each dimension. Based on the multi-dimensional data, a Bayesian network algorithm is used to establish causal relationships between dimensions, constructing a directed acyclic graph. In the formula, It is a directed acyclic graph, with nodes For ICF dimensions, edges express yes The cause, for any node Define a conditional probability distribution, expressed as:

[0040]

[0041] In the formula, Represents a probability distribution. Indicates the first One evaluation dimension express The set of parent nodes, This indicates that the parent node is known. The conditional probability distribution, Let the parameters of a function used to model this conditional probability be: Input variables are , To represent a probability model, Indicates the first The parameters of a function.

[0042] Furthermore, the operational process of the dynamic evaluation module includes:

[0043] Calculate the final weighted evaluation score. Each dimension The value can be obtained by summing the sub-indicators and taking the expected value, as shown in the expression:

[0044]

[0045] In the formula, Indicates the first The overall evaluation score at any given moment Indicates from the first Dimensions accumulated to the th Dimension Indicates the first Each dimension at time The weight, Indicates the first Each dimension at time The specific score, Indicates from the first Sub-indicators added to the first There are a total of [number] sub-indicators. Sub-indicators, Indicates the first The first dimension The weights of each sub-indicator satisfy the following: , Indicates the first In the dimension, the first Sub-indicators at time The rating is based on scores across five dimensions. Automatically generate a five-dimensional radar chart and store the historical evaluation sequence after each round of evaluation. In the formula, Indicates time Historical assessment data set, Indicates time The system uses historical assessment data sets and visualizes assessment trends to update assessment results and radar charts.

[0046] Specifically, this system, targeting neurorehabilitation patients, continuously monitors five dimensions: physical function, activity level, participation, environmental factors, and psychological state. It dynamically adjusts the weights of each indicator to accurately reflect subtle changes in the patient's rehabilitation process. Bayesian networks reveal causal paths between different dimensions, helping to identify key influencing factors and enabling targeted rehabilitation interventions. The dynamic weight allocation mechanism improves the sensitivity and adaptability of the assessment, while Bayesian network causal modeling enhances the scientific rigor and reliability of the assessment.

[0047] The data feature module, based on the latent profile analysis algorithm, performs high-dimensional feature dimensionality reduction processing based on the evaluation results and radar chart to identify the hidden subtype;

[0048] Furthermore, the operation process of the data feature module includes:

[0049] The algorithm based on the potential profile analysis performs high-dimensional feature dimensionality reduction processing based on the evaluation results and radar image, expressed as follows:

[0050]

[0051] In the formula, Indicates the first Feature vector of each sample The probability density value, Indicates the first The feature vector of each sample Indicates all Sum of the number of hidden subtypes, This represents the number of hidden subtypes in a Gaussian mixture model. Indicates the first The weights of each Gaussian component, Indicates the first A Gaussian distribution pairs of samples The probability density function value, Indicates the first A vector of mean values ​​from a Gaussian distribution. Indicates the first The covariance matrices of a Gaussian distribution are used to determine the optimal number of patient subtype classifications using the minimum information entropy criterion. The expression is:

[0052]

[0053] In the formula, The entropy regularization objective function measures the number of hidden subtypes. The model is divided into a comprehensive index of clarity and complexity. This indicates that for all samples arrive Summation, This indicates the numbering of all cryptic subtypes. arrive Summation, Indicates the first The sample belongs to the first Membership degree of each hidden subtype Indicates the first The sample belongs to the first The membership logarithm function of a hidden subtype Indicates hyperparameters, This represents the number of feature dimensions for each sample. This represents the number of parameters in the covariance matrix of each Gaussian component. The number of parameters representing a single cryptic subtype. Indicates all The total number of parameters required for each hidden subtype The logarithm of the sample size. This represents the optimal number of hidden subtypes. This indicates taking the option that minimizes the objective function. value, Indicates the number of cryptic subtypes in arrive Perform a traversal search within the specified range.

[0054] Furthermore, the operation process of the data feature module includes:

[0055] Through iterative clustering, a multi-round iterative strategy based on latent response weight adjustment is adopted, expressed as:

[0056]

[0057] In the formula, Indicates the first The sample at the th The feature representation vector during round iteration. Indicates the first The sample at the th The new feature representation after round of iterations, This represents the optimal number of hidden subtypes. Indicates the first The sample at the th In the round of iteration, belonging to the 1st Membership degree of each hidden subtype Indicates the first The first cryptic subtype in the... The center representation vector during round iteration. Indicates the first The sample to the first One direction for updating the stealth subtype. Using the step size, the elusive subtype is identified.

[0058] Specifically, this system, based on latent profile analysis algorithms, performs dimensionality reduction and hidden subtype identification on high-dimensional features of patients' multidimensional assessment results and radar chart data. For rheumatic disease rehabilitation management, it processes multidimensional data including disease indicators, quality of life scores, and functional assessments to finely segment patients and identify different pathological subtypes, such as inflammation-dominant and functional impairment types. This provides a scientific basis for subsequent personalized treatment and intervention, achieving effective dimensionality reduction and reasonable grouping of high-dimensional data. The multi-round iterative clustering strategy enhances the system's stability and identification accuracy. Identifying hidden subtypes helps reveal the heterogeneity of patient groups, supports the development of personalized medical plans, and improves clinical diagnosis and treatment outcomes and patient management levels.

[0059] The causal intervention module, based on the aforementioned hidden subtype, utilizes structural equation modeling to construct a causal path network and combines it with knowledge graph technology to generate a targeted intervention strategy library.

[0060] Furthermore, the operational process of the causal intervention module includes:

[0061] Based on the aforementioned hidden subtype, a causal path network is constructed using structural equation modeling, with the latent variable vector set as follows. The observed variable vector is The causal path network is expressed as:

[0062]

[0063] In the formula, Represents the vector of observed variables. Represents a vector of latent variables. This indicates the strength of the influence of each latent variable on the observed variable. Indicates observation error. This represents the influence of one latent variable on another latent variable. This represents the matrix representing the influence of exogenous variables on latent variables. It represents unexplained random fluctuations.

[0064] Furthermore, the operational process of the causal intervention module includes:

[0065] The transmission paths between dimensions are verified using the maximum likelihood estimation method. A covariance structure and a log-likelihood function are constructed. Then, the negative log-likelihood is minimized using gradient descent to obtain the optimal parameters. The expression is:

[0066]

[0067] In the formula, Represents the parameter vector The function, This indicates the strength of the influence of each latent variable on the observed variable. Represents the identity matrix. This represents the influence of one latent variable on another latent variable. The propagation matrix representing the causal effects in the system. Represents the covariance matrix between latent variables. This represents the transpose inverse of the propagation matrix of causal effects in the system. yes transpose, Represents the measurement error covariance matrix. Indicates in the parameter The model fit is goodness of fit. Represents the parameter vector The logarithmic function of the absolute value, Indicates the first The number of samples for each cryptic subtype. Indicates the scaling factor. Indicates the first The sample covariance matrix of each hidden subtype. Denotes the inverse matrix of the covariance matrix. This represents the sum of all elements on the main diagonal. Represents the optimal parameter vector. This represents finding the parameters that minimize the objective function. Simultaneously, by combining knowledge graph technology, a targeted intervention strategy library is generated, and a triplet knowledge graph is constructed, with the expression:

[0068]

[0069] In the formula, A collection representing a knowledge graph. This represents a knowledge edge in the graph, which automatically matches personalized decision-making solutions for patients with different occult subtypes.

[0070] Specifically, this system constructs a causal path network based on a structural equation model of latent subtypes, accurately characterizing the relationship between latent variables and observed variables, as well as the causal transmission mechanism between latent variables. For patients recovering from rheumatic diseases, it identifies different latent subtypes in the patient population, such as immune-responsive type and inflammation-dominant type. It uses structural equation modeling to reveal the causal relationship between various pathological indicators, and combines it with clinical knowledge graphs to generate personalized drug and rehabilitation intervention plans. This enables the construction of a knowledge-driven targeted intervention strategy library, enhancing the scientific nature and practicality of decision-making, automatically matching personalized plans, improving treatment effects and patient satisfaction, and promoting the development of precision medicine.

[0071] A closed-loop feedback module is used to establish a two-way feedback mechanism between evaluation and decision-making, providing feedback to the dynamic evaluation module to optimize weight allocation and iteratively update the causal path network.

[0072] Furthermore, the operation process of the closed-loop feedback module includes:

[0073] The mechanism for establishing a two-way feedback between assessment and decision-making is input back into the dynamic assessment module. Based on the intervention effect data reported by patients after the implementation of the decision-making plan, a dynamic loss function is introduced and input back into the dynamic assessment module. The dynamic weights are iteratively optimized using a mirror descent method with entropy constraints. The expression is:

[0074]

[0075] In the formula, Indicates the first Feedback loss function at each time point The squared Euclidean distance is used to measure the difference between the predicted value and the actual feedback value. Indicates the first A vector of real patient feedback at each time point. Indicates the first The system predicts the output vector at each time point. This indicates that the hyperparameter is used to adjust the weight of the entropy regularization term in the total loss. Indicates the current weight distribution Information entropy Indicates at a point in time Time Dynamic weight values ​​for each dimension Represents the entropy function. Indicates the first The latest update Weight values ​​for each dimension, This indicates that the right side represents the value before normalization. This represents the natural exponential function. Indicates the learning rate. This indicates that the loss function is related to the current i-th... The partial derivatives of the weights of each dimension.

[0076] Furthermore, the operation process of the closed-loop feedback module includes:

[0077] The causal path network is iteratively updated by using a policy gradient-based reinforcement learning algorithm to obtain path selection strategies based on the patient's long-term evaluation results and radar chart. ,expression:

[0078]

[0079] In the formula, Indicates the first A vector of real patient feedback at each time point. Indicates the first The system predicts the output vector at each time point. Representation Strategy Total expected return Indicating in strategy The expected value under, Indicates from arrive The summation symbol, Discount factor Power of 1 Indicates the value of the discount factor , This means that the smaller the difference between the assessed value and the actual feedback, the higher the reward. This represents the coefficient of the penalty term. The Kullback-Leibler divergence measures the difference between two probability distributions. Indicates the current time The constructed causal path network Indicates the previous moment causal path network This indicates the degree of deviation between the current path network and the previous path structure; the greater the deviation, the higher the penalty.

[0080] Specifically, this system constructs a two-way feedback mechanism for assessment and decision-making to dynamically optimize personalized rehabilitation plans for patients. In the rehabilitation management of rheumatic patients, through the implementation of multiple intervention programs and the dynamic input of patient feedback data, the system continuously adjusts the weights of assessment dimensions, such as social support and physical function, to reflect the patient's latest health status. At the same time, the system learns and optimizes causal intervention paths for different patient subtypes, making the pushed rehabilitation plans more personalized and effective. Patients upload their daily status at home through smart terminals, and the system updates the weights and causal models in real time. Doctors adjust their treatment strategies accordingly, forming an adaptive and intelligent rehabilitation closed loop, which is conducive to improving rehabilitation effects and patient satisfaction, and reducing the waste of medical resources.

[0081] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A rheumatic disease rehabilitation intelligent assessment system based on the ICF concept and classification system, characterized in that, include: A multi-source data module is used to collect and integrate multi-dimensional data; The dynamic evaluation module, based on ICF theory, constructs a dynamic weight allocation model and updates the evaluation results and radar chart according to the multi-dimensional data. The data feature module, based on the latent profile analysis algorithm, performs high-dimensional feature dimensionality reduction processing based on the evaluation results and radar chart to identify the hidden subtype; The causal intervention module, based on the aforementioned hidden subtype, utilizes structural equation modeling to construct a causal path network and combines it with knowledge graph technology to generate a targeted intervention strategy library. A closed-loop feedback module is used to establish a two-way feedback mechanism between evaluation and decision-making, providing feedback to the dynamic evaluation module to optimize weight allocation and iteratively update the causal path network. The operation process of the multi-source data module includes: The system is used to collect and integrate multi-dimensional data. Through an API interface protocol, it connects with hospital information systems, image archiving, and communication systems to collect patients' biochemical test data, imaging data, and medication records in real time. Simultaneously, a patient-side interactive interface is developed to support self-assessment data entry based on the five dimensions of the ICF framework: disease activity, physical function, quality of life, social support, and work efficiency. The patient-side interactive interface supports the completion of the ICF five-dimensional questionnaire, generating the following self-assessment data vector: In the formula, This represents a five-dimensional evaluation vector within the ICF framework. This represents a set with five elements. Represents any number of Evaluation scores in each dimension, The score range for each dimension is from 0 to 1, where 0 represents very poor and 1 represents excellent. All scales are standardized and mapped to the [0,1] interval. A data fusion algorithm is constructed using DS evidence theory. For data conflict scenarios, a manual review process is initiated through outlier cross-validation rules. Trust assignments are provided from both data sources, and the synthesized evidence is as follows: In the formula, Indicates the proposition The combined trust distribution function value, Indicates the first source of evidence pair subset The trust distribution function value, Indicates the second source of evidence pair subset The trust distribution function value, Indicates all In a combination, the intersection equals All pairs Summation above, This represents the normalization factor over all non-conflicting parts. Represents all combinations of subsets when the intersection is an empty set. Summation, Representing a subset and The intersection is empty, expand to When there are multiple data sources, construct the fusion process: In the formula, This represents the trust distribution function after fusion. This refers to a method of fusing multiple pieces of evidence using the combination rules of the Dempster-Shafer evidence theory. Indicates the first to the second The original evidence trust distribution function; The operation process of the dynamic evaluation module includes: Based on ICF theory, the five dimensions are broken down into a set of quantifiable sub-indicators, and a dynamic weight allocation model is constructed. When fluctuations in relevant sub-indicators are detected, a non-linear update is triggered. In the formula, Indicates a point in time At that time, the first Dynamic weight values ​​for each dimension Represents the natural constant. This indicates an adjustable sensitivity coefficient. Indicates time At that time, the first Changes in each dimension Five core dimensions of ICF To sum, This represents the exponential weighted value of each dimension. Based on the multi-dimensional data, a Bayesian network algorithm is used to establish causal relationships between dimensions, constructing a directed acyclic graph. In the formula, It is a directed acyclic graph, with nodes For ICF dimensions, edges express yes The cause, for any node Define a conditional probability distribution, expressed as: In the formula, Represents a probability distribution. Indicates the first One evaluation dimension express The set of parent nodes, This indicates that the parent node is known. The conditional probability distribution, Let the parameters of a function used to model this conditional probability be: Input variables are , To represent a probability model, Indicates the first The parameters of the function; Calculate the final weighted evaluation score. Each dimension The value can be obtained by summing the sub-indicators and taking the expected value, as shown in the expression: In the formula, Indicates the first The overall evaluation score at any given moment Indicates from the first Dimensions accumulated to the th Dimension Indicates the first Each dimension at time The weight, Indicates the first Each dimension at time The specific score, Indicates from the first Sub-indicators added to the first There are a total of [number] sub-indicators. Sub-indicators, Indicates the first The first dimension The weights of each sub-indicator satisfy the following: , Indicates the first In the dimension, the first Sub-indicators at time The rating is based on scores across five dimensions. Automatically generate a five-dimensional radar chart and store the historical evaluation sequence after each round of evaluation. In the formula, Indicates time Historical assessment data set, Indicates time The system uses historical assessment data sets and visualizes assessment trends to update assessment results and radar charts.

2. The intelligent assessment system for rheumatoid arthritis rehabilitation based on the ICF concept and classification system according to claim 1, characterized in that, The operation process of the data feature module includes: The algorithm based on the potential profile analysis performs high-dimensional feature dimensionality reduction processing based on the evaluation results and radar image, expressed as follows: In the formula, Indicates the first Feature vector of each sample The probability density value, Indicates the first The feature vector of each sample Indicates all Sum of the number of hidden subtypes, This represents the number of hidden subtypes in a Gaussian mixture model. Indicates the first The weights of each Gaussian component, Indicates the first A Gaussian distribution on the sample The probability density function value, Indicates the first A vector of mean values ​​from a Gaussian distribution. Indicates the first The covariance matrices of a Gaussian distribution are used to determine the optimal number of patient subtype classifications using the minimum information entropy criterion. The expression is: In the formula, The entropy regularization objective function measures the number of hidden subtypes. The model is divided into a comprehensive index of clarity and complexity. This indicates that for all samples arrive Summation, This indicates the numbering of all cryptic subtypes. arrive Summation, Indicates the first The sample belongs to the first Membership degree of each hidden subtype Indicates the first The sample belongs to the first The membership logarithm function of a hidden subtype Indicates hyperparameters, This represents the number of feature dimensions for each sample. This represents the number of parameters in the covariance matrix of each Gaussian component. The number of parameters representing a single cryptic subtype. Indicates all The total number of parameters required for each hidden subtype The logarithm of the sample size. This represents the optimal number of hidden subtypes. This indicates taking the option that minimizes the objective function. value, Indicates the number of cryptic subtypes in arrive Perform a traversal search within the specified range.

3. The intelligent assessment system for rheumatoid arthritis rehabilitation based on the ICF concept and classification system according to claim 2, characterized in that, The operation process of the data feature module includes: Through iterative clustering, a multi-round iterative strategy based on latent response weight adjustment is adopted, expressed as: In the formula, Indicates the first The sample at the th The feature representation vector during round iteration. Indicates the first The sample at the th The new feature representation after round of iterations, This represents the optimal number of hidden subtypes. Indicates the first The sample at the th In the round of iteration, belonging to the 1st Membership degree of each hidden subtype Indicates the first The first cryptic subtype in the... The center representation vector during round iteration. Indicates the first The sample to the first One direction for updating the stealth subtype. Using the step size, the elusive subtype is identified.

4. The intelligent assessment system for rheumatoid arthritis rehabilitation based on the ICF concept and classification system according to claim 3, characterized in that, The operational process of the causal intervention module includes: Based on the aforementioned hidden subtype, a causal path network is constructed using structural equation modeling, with the latent variable vector set as follows. The observed variable vector is The causal path network is expressed as: In the formula, Represents the vector of observed variables. Represents a vector of latent variables. This indicates the strength of the influence of each latent variable on the observed variable. Indicates observation error. This represents the influence of one latent variable on another latent variable. This represents the matrix representing the influence of exogenous variables on latent variables. It represents unexplained random fluctuations.

5. The intelligent assessment system for rheumatoid arthritis rehabilitation based on the ICF concept and classification system according to claim 4, characterized in that, The operational process of the causal intervention module includes: The transmission paths between dimensions are verified using the maximum likelihood estimation method. A covariance structure and a log-likelihood function are constructed. Then, the negative log-likelihood is minimized using gradient descent to obtain the optimal parameters. The expression is: In the formula, Represents the parameter vector The function, This indicates the strength of the influence of each latent variable on the observed variable. Represents the identity matrix. This represents the influence of one latent variable on another latent variable. The propagation matrix representing the causal effects in the system. Represents the covariance matrix between latent variables. This represents the transpose inverse of the propagation matrix of causal effects in the system. yes transpose, Represents the measurement error covariance matrix. Indicates in the parameter The model fit is goodness of fit. Represents the parameter vector The logarithmic function of the absolute value, Indicates the first The number of samples for each cryptic subtype. Indicates the scaling factor. Indicates the first The sample covariance matrix of each hidden subtype. Denotes the inverse matrix of the covariance matrix. This represents the sum of all elements on the main diagonal. Represents the optimal parameter vector. This represents finding the parameters that minimize the objective function. Simultaneously, by combining knowledge graph technology, a targeted intervention strategy library is generated, and a triplet knowledge graph is constructed, with the expression: In the formula, A collection representing a knowledge graph. This represents a knowledge edge in the graph, which automatically matches personalized decision-making solutions for patients with different occult subtypes.

6. The intelligent assessment system for rheumatoid arthritis rehabilitation based on the ICF concept and classification system according to claim 5, characterized in that, The operation process of the closed-loop feedback module includes: The mechanism for establishing a two-way feedback between assessment and decision-making is input back into the dynamic assessment module. Based on the intervention effect data reported by patients after the implementation of the decision-making plan, a dynamic loss function is introduced and input back into the dynamic assessment module. The dynamic weights are iteratively optimized using a mirror descent method with entropy constraints. The expression is: In the formula, Indicates the first The feedback loss function at each time point The squared Euclidean distance is used to measure the difference between the predicted value and the actual feedback value. Indicates the first A vector of real patient feedback at each time point Indicates the first The system predicts the output vector at each time point. This indicates that the hyperparameter is used to adjust the weight of the entropy regularization term in the total loss. Indicates the current weight distribution Information entropy Indicates a point in time Time Dynamic weight values ​​for each dimension Represents the entropy function. Indicates the first The latest update Weight values ​​for each dimension, This indicates that the right side represents the value before normalization. This represents the natural exponential function. Indicates the learning rate. This indicates that the loss function is related to the current i-th... The partial derivatives of the weights of each dimension.

7. The intelligent assessment system for rheumatoid arthritis rehabilitation based on the ICF concept and classification system according to claim 6, characterized in that, The operation process of the closed-loop feedback module includes: The causal path network is iteratively updated by using a policy gradient-based reinforcement learning algorithm to obtain path selection strategies based on the patient's long-term evaluation results and radar chart. ,expression: In the formula, Indicates the first A vector of real patient feedback at each time point Indicates the first The system predicts the output vector at each time point. Representation strategy Total expected return Indicating in strategy The expected value under, Indicates from arrive The summation symbol, Discount factor Power of 1 Indicates the value of the discount factor , This means that the smaller the difference between the assessed value and the actual feedback, the higher the reward. This represents the coefficient of the penalty term. The Kullback-Leibler divergence measures the difference between two probability distributions. Indicates the current time The constructed causal path network Indicates the previous moment causal path network This indicates the degree of deviation between the current path network and the previous path structure; the greater the deviation, the higher the penalty.

Images