Rehabilitation medical evaluation management system based on dynamic multi-scale adaptation and automatic calculation

Abstract

The invention discloses a rehabilitation medical assessment management system based on dynamic multi-scale adaptation and automatic calculation, and particularly relates to the technical field of medical assessment management. The assessment management system comprises a patient information acquisition module, a dynamic scale adaptation module, an automatic data acquisition and calculation module, a data analysis and report generation module, a rehabilitation management module and a database storage center, and the patient information acquisition module is used for acquiring patient information and storing the acquired information in the database storage center. According to the rehabilitation medical assessment management system based on dynamic multi-scale adaptation and automatic calculation, through dynamic multi-scale adaptation, the most suitable assessment scale combination can be selected according to individual differences and rehabilitation stages of patients, the rehabilitation conditions of the patients are comprehensively and accurately reflected, and the accuracy of assessment results is improved.

Description

Technical Field

[0001] This application relates to the field of medical assessment and management technology, and more specifically, to a rehabilitation medical assessment and management system based on dynamic multi-scale adaptation and automated calculation. Background Technology

[0002] In rehabilitation medicine, accurate assessment is crucial for developing personalized rehabilitation plans and judging rehabilitation effectiveness. Currently, rehabilitation assessments mainly rely on single or a few fixed assessment scales, which are insufficient to comprehensively reflect the complex rehabilitation status of patients. Furthermore, traditional assessment methods often involve manually filling out scales and manually calculating results, which is not only inefficient but also prone to human error, failing to meet the needs of the rapidly developing rehabilitation medicine field. In addition, different patients have varying rehabilitation progress and physical conditions, and existing assessment systems lack the ability to dynamically adjust assessment scales based on the patient's actual situation, resulting in insufficient accuracy and specificity of the assessment results.

[0003] Therefore, a rehabilitation medical assessment and management system based on dynamic multi-scale adaptation and automated calculation is proposed to address the above problems. Summary of the Invention

[0004] In order to overcome the above-mentioned deficiencies of the prior art, this application provides a rehabilitation medical assessment and management system based on dynamic multi-scale adaptation and automated calculation to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, this application provides the following technical solution: a rehabilitation medical assessment and management system based on dynamic multi-scale adaptation and automated calculation. This assessment and management system includes: a patient information collection module, a dynamic scale adaptation module, an automated data collection and calculation module, a data analysis and report generation module, a rehabilitation management module, and a database storage center.

[0006] The patient information collection module is used to collect patient information and store the collected information in the database storage center;

[0007] The dynamic scale adaptation module uses the information obtained by the patient information collection module and a preset adaptation algorithm to dynamically select the most suitable combination of assessment scales for the patient's current rehabilitation stage and condition from the system's built-in dynamic multi-scale library.

[0008] The automated data acquisition and calculation module collects physiological and motor data by connecting to the intelligent medical device, collects dynamic assessment scale response data, and uses the Kalman filter formula for prediction.

[0009] The data analysis and report generation module performs in-depth analysis on the evaluation results obtained by the automated data acquisition and calculation module to obtain the changing trend of the patient's recovery status and generates a recovery evaluation report.

[0010] The rehabilitation management module analyzes and reports the rehabilitation suggestions provided by the evaluation result analysis and report generation module, and formulates a rehabilitation plan for the patient in combination with the rehabilitation scheme knowledge base built in the system. Meanwhile, the rehabilitation management module is also used for tracking and managing the execution of the rehabilitation plan, and adjusting the rehabilitation plan according to the rehabilitation progress of the patient.

[0011] The database storage center is used for storing system-related data, adopts master-slave replication and sub-database and sub-table technology, encrypts data by using an AES algorithm, and optimizes data query by using index coverage optimization and Redis cache technology.

[0012] Further, the output end of the patient information acquisition module is electrically connected to the input end of the dynamic scale adaptation module, and the patient information acquisition module complies with the HL7 standard protocol or FHIR specification when it is connected to the hospital electronic medical record system, and realizes automatic extraction of structured data through an API interface. The patient information includes patient basic information, rehabilitation diagnosis information and current rehabilitation stage information, and the information is obtained through connection with the hospital electronic medical record system, self-entry by the patient, manual input by medical staff, judgment of age validity, setting of a flag bit to judge the entry of mandatory words, and storage in the database storage center.

[0013] Further, the patient information acquisition module judges the validity of the data by executing the following strategies:

[0014] wherein 1 represents valid age data, and 0 represents invalid data, and a flag bit F is set. i If a field is mandatory and has not been entered, F i = 0, and the system triggers a prompt process. If it has been entered, F i = 1.

[0015] When the system receives age data, the input value is compared with the range of age validity. If the age is within the range of 0 to 120 years old, the data is marked as valid, F i = 1. Otherwise, if it is out of the range, it is determined as invalid data, F i = 0, and the system triggers a prompt process and checks the invalid data.

[0016] Further, the dynamic multi-scale library stores a plurality of evaluation scales for different rehabilitation fields, and the evaluation scales include but are not limited to functional evaluation scales, symptom evaluation scales and quality of life scales.

[0017] The adaptation algorithm in the dynamic scale adaptation module adopts a strategy of combining decision trees and Bayesian networks to construct an adaptation algorithm model, and the decision tree takes the disease type of the patient, the rehabilitation stage, the age and the gender as decision nodes to select the scale.

[0018] The Bayesian network is used for calculating the scale applicability probability, and event A represents selecting a certain scale, and event B represents a set of patient characteristics, according to the formula: Wherein, P(A|B) is the probability of selecting scale A under the condition that the patient has characteristics B, P(B|A) is the probability of the patient having characteristics B under the condition of selecting scale A, P(A) is the prior probability of selecting scale A, and P(B) is the probability of the patient having characteristics B, the probability values are calculated through historical data, and the branch weight of the decision tree is dynamically adjusted.

[0019] Further, the Kalman filtering formula is used in the automatic data acquisition and calculation module to remove the noise of the data collected in the intelligent medical device, and the specific steps are:

[0020] Through the filtered data combined with the kinematics principle, the collected physiological signals and action data are preprocessed and feature extracted,

[0021] Prediction step:

[0022] Update step:

[0023]

[0024] P k|k =(I-K k H k )P k|k-1 ;

[0025] Wherein, is the state prediction value based on the state estimation value at k-1 time, F k is a state transition matrix, is the state estimation value at k-1 time, B k is a control input matrix, u k is a control input, P k|k-1 is a prediction error covariance, Q k is a process noise covariance, K k is a Kalman gain, H k is an observation matrix, z k is an observation value, R k is an observation noise covariance, is the updated state estimation value at k time, P k|k is the updated error covariance;

[0026] Calculate the range of joint activity, the formula is: R = max(θ1, θ2,..., θ n )-min(θ1, θ2,..., θ n );

[0027] The assessment results are derived from the data obtained from the scale responses, and the formula is as follows:

[0028] Where θ1,θ2,...,θ n To collect angle data of the joint at different times, R represents the range of motion of the joint, and w i s represents the weight of a certain item in the scale. i The patient was given a score, where S is the overall score of motor function, α is the weighting coefficient of the motor data, and m is the number of items on the scale.

[0029] Furthermore, the output of the automated data acquisition and calculation module is electrically connected to the input of the data analysis and report generation module. The data analysis and report generation module includes a data analysis unit and a report generation unit. The data analysis unit is used to propose an analysis model. Through this model, the evaluation results obtained by the automated data acquisition and calculation module are analyzed in depth and compared with the system's built-in rehabilitation evaluation standards and historical evaluation data to analyze the changing trends of the patient's rehabilitation status. The analysis model uses the ARIMA model, specifically:

[0030] Let time series y t The ARIMA(p,d,q) model expression is: in, For the autoregressive part; It is a difference operator used to stationary time series, Θ(B)=1+θ1B+...+θ q B q For the moving average portion; ∈ t The sequence is a white noise sequence, where p is the autoregressive order, d is the difference order, and q is the moving average order.

[0031] Model parameters are estimated using the least squares method. θ i This allows for the prediction of future rehabilitation indicator values. During comparative analysis, the current assessment indicator x is calculated. cur Compared with historical indicators x his The rate of change r is specifically: The rate of change r reflects the dynamic changes in rehabilitation effect. A positive rate of change r indicates that the rehabilitation effect has improved, and the larger the value, the more significant the improvement. A negative rate of change r indicates that the rehabilitation effect has regressed.

[0032] Furthermore, the report generation unit generates an evaluation report based on the analysis of the data analysis unit using the Freemarker template engine and the ECharts chart library.

[0033] Further, the input end of the rehabilitation management module is electrically connected with the output end of the data analysis and report generation module, and the rehabilitation management module comprises a scheme making unit and a scheme adjustment, the scheme making unit is used for constructing a scheme making model, and a rehabilitation scheme is made for a patient by combining the scheme making model with a rehabilitation scheme knowledge base built in the system, the rehabilitation scheme knowledge base stores historical cases, and the scheme making model executes the following strategies:

[0034] Based on case-based reasoning (CBR), the similarity sim(X, Y n ) of the current patient feature vector X = [x1, x2,..., x j and the historical case feature vector Y j1 = [y j2 , y jn ,..., y j ] is calculated, the historical case with high similarity is selected as a reference, and the scheme is adjusted by combining a rule engine;

[0035]

[0036] Suppose that a rule in the rule set is: if the muscle strength of a patient is lower than a threshold T, then the training intensity I is set to I min , that is:

[0037] wherein n is the number of dimensions of the features, I min is the minimum training intensity, and I default is the default training intensity.

[0038] Further, the scheme adjustment is used for adjusting the rehabilitation scheme according to the rehabilitation progress of the patient, and the Q-learning algorithm in reinforcement learning is adopted in the scheme adjustment, specifically:

[0039] Suppose that the state space is S and the action space is A, the action a is executed in the state s to transfer to the state s' and obtain the reward r', wherein the Q value updating formula is:

[0040]

[0041] wherein a is a learning rate, and a [0, 1], and g is a discount factor, and g [0, 1];

[0042] By continuously updating the Q value, the rehabilitation scheme adjustment strategy is optimized, and finally the rehabilitation scheme adjustment strategy that maximizes the long-term cumulative reward is formed, so that the treatment path dynamic planning is realized.

[0043] ​Further, the system-related data in the database storage center includes basic information of patients, rehabilitation diagnosis information, dynamic assessment scale data, assessment result data, and rehabilitation program data, all of which are related to rehabilitation medical assessment management.

[0044] Technical effects and advantages of the present application: Compared with the prior art, by dynamic multi-scale adaptation, the most suitable assessment scale combination can be selected according to the individual differences and rehabilitation stages of patients, the rehabilitation status of patients can be comprehensively and accurately reflected, and the accuracy of assessment results can be improved.

[0045] The automatic collection and calculation of assessment data are realized, manual operation is greatly reduced, assessment efficiency is improved, human calculation errors are avoided, in addition, the automation and intelligentization of the system make rehabilitation assessment and management more convenient, and medical staff can quickly obtain assessment results and develop rehabilitation programs.

[0046] Based on accurate assessment results, personalized rehabilitation programs are provided for patients, and the programs can be dynamically adjusted according to the rehabilitation progress of patients to meet the personalized rehabilitation needs of patients and improve the rehabilitation treatment effect.

[0047] The powerful database module uniformly manages rehabilitation medical related data, facilitates data query, statistics and analysis, and through analysis of a large amount of data, rehabilitation treatment rules can be summarized to provide strong support for rehabilitation medical research and clinical practice. BRIEF DESCRIPTION OF DRAWINGS

[0048] Figure 1 The figure is a system flowchart of the present application.

[0049] In the figure: 1, patient information collection module; 2, dynamic scale adaptation module; 3, automatic data collection and calculation module; 4, data analysis and report generation module; 401, data analysis unit; 402, report generation unit; 5, rehabilitation management module; 501, program development unit; 502, program adjustment unit; 6, database storage center. DETAILED DESCRIPTION

[0050] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0051] EMBODIMENT

[0052] As Figure 1The illustrated rehabilitation medical evaluation management system based on dynamic multiple scale adaptation and automatic calculation, the monitoring system The evaluation management system includes: patient information acquisition module 1, dynamic scale adaptation module 2, automatic data acquisition and calculation module 3, data analysis and report generation module 4, rehabilitation management module 5 and database storage center 6,

[0053] The patient information acquisition module 1 is used to collect patient information and store the collected information in the database storage center 6;

[0054] The output end of the patient information acquisition module 1 is electrically connected to the input end of the dynamic scale adaptation module 2, and when the patient information acquisition module 1 is connected with the hospital electronic medical record system, it follows the HL7 standard protocol or FHIR specification, realizes automatic extraction of structured data through API interface, the patient information includes patient basic information, rehabilitation diagnosis information and current rehabilitation stage information, through the connection with the hospital electronic medical record system, the patient self-input, the medical staff manually input information, judge the age validity, set the flag bit to judge the mandatory character input condition, and store the information to the database storage center 6.

[0055] The patient information acquisition module 1 judges the age validity by the following strategy to judge the validity of the data:

[0056] Among them, 1 represents that the age data is valid, 0 represents invalid, and the flag bit F is set i If a field is mandatory and has not been entered, F i =0, the system triggers the prompt process, if it has been entered, F i =1;

[0057] When the system receives the age data, the input value is compared with the range of age validity, if the age is in the interval of 0 to 120 years old, the data is marked as valid, F i =1; otherwise, if it exceeds this range, it is determined as invalid data, F i =0, the system triggers the prompt process and checks the invalid data.

[0058] The dynamic multi-scale library stores a plurality of assessment scales for different rehabilitation fields, and the assessment scales include but are not limited to functional assessment scales, symptom assessment scales and quality of life scales; the acquisition of the multi-scale generally needs to screen the scales verified by the reliability and validity through the authoritative medical literature retrieval (such as PubMed and China National Knowledge Network), or the industry guidelines recommendation (such as the standard of the International Society of Physical Medicine and Rehabilitation Medicine), and form a customized scale set according to the assessment target combination, such as the Fugl-Meyer Assessment Scale covering motor function or the Montreal Cognitive Assessment Scale MoCA covering cognitive function for the functional assessment scale, the Visual Analogue Scale VAS for pain or the Brief Fatigue Inventory BFI for the symptom assessment scale, and the SF-36 Health Survey for the quality of life scale, which are all prior art and will not be described in detail here.

[0059] The dynamic scale adaptation module 2 dynamically selects the assessment scale combination most suitable for the current rehabilitation stage and condition of the patient from the dynamic multi-scale library built in the system based on the information acquired by the patient information acquisition module 1 by using a preset adaptation algorithm;

[0060] The adaptation algorithm in the dynamic scale adaptation module 2 adopts a strategy of combining decision trees and Bayesian networks to construct an adaptation algorithm model, and the decision trees take the disease type, rehabilitation stage, age and gender of the patient as decision nodes to perform scale screening;

[0061] The Bayesian network is used to calculate the scale applicability probability, and event A represents selecting a scale, and event B represents a set of patient characteristics. According to the formula: Wherein, P(A|B) is the probability of selecting scale A under the condition that the patient has characteristic B, P(B|A) is the probability of the patient having characteristic B under the condition of selecting scale A, P(A) is the prior probability of selecting scale A, and P(B) is the probability of the patient having characteristic B. The probability values are calculated through historical data, and the branch weight of the decision tree is dynamically adjusted. The Bayesian network is used to process the uncertainty in the data. According to the historical assessment data and the patient population characteristics, the applicability probability of different scales under certain conditions is calculated, and the branch weight of the decision tree is dynamically adjusted, so that the scale selection is more suitable for the actual situation of the patient.

[0062] The automatic data acquisition and calculation module 3 acquires physiological and motion data by connecting with the intelligent medical device, acquires dynamic assessment scale answer data, and adopts Kalman filtering formula for prediction.

[0063] By connecting with intelligent medical devices (such as wearable sensors, built-in sensors of rehabilitation training devices, etc.), real-time physiological data and action data of patients during the evaluation process are collected, such as heart rate, limb movement trajectory, strength output, etc. At the same time, the answer data of the patients on the electronic evaluation scale are automatically collected. Then, using the preset calculation model and algorithm, the collected data are automatically calculated and analyzed to obtain the specific values and evaluation results of each evaluation index. For example, according to the collected limb movement trajectory data, the joint range of motion, movement speed and other indicators are calculated, and the movement function of the patient is comprehensively evaluated in combination with the scale answer data;

[0064] For intelligent medical device data, connection is established through Bluetooth, Wi-Fi and other wireless communication technologies, and real-time data transmission is realized by following the Internet of Things communication protocol (such as MQTT);

[0065] In the automatic data acquisition and calculation module 3, Kalman filtering formula is used to remove the noise of the data collected in the intelligent medical device, and the specific steps are as follows:

[0066] Through the filtered data combined with the kinematics principle, the collected physiological signals and action data are preprocessed and feature extracted,

[0067] Prediction step:

[0068] Update step:

[0069]

[0070] P k|k =(I-K k H k )P k|k-1 ;

[0071] Wherein, is the state prediction value based on k-1 time at k time, F k is the state transition matrix, is the state estimation value at k-1 time, B k is the control input matrix, u k is the control input, P k|k-1 is the prediction error covariance, Q k is the process noise covariance, K k is the Kalman gain, H k is the observation matrix, z k is the observation value, R k is the observation noise covariance, is the updated state estimation value at k time, P k|k is the updated error covariance;

[0072] The joint range of motion is calculated by the formula: R = max(θ1, θ2,..., θ n )-min(θ1, θ2,..., θ n );

[0073] The evaluation result is obtained by the scale answer data, and the formula is:

[0074] wherein θ1, θ2,..., θ n are the angle data of the joint collected at different times, R is the joint range of motion, w i is the score weight of a question in the scale, s i is the score of the patient's answer, S is the comprehensive score of motor function, and a is the motor data weight coefficient, and m is the number of scale questions;

[0075] Taking motor function evaluation as an example, a convolutional neural network (CNN) is used to extract features from the limb motion trajectory data, identify key indicators such as joint angle change and motion speed, etc.

[0076] The data analysis and report generation module 4 performs in-depth analysis on the evaluation results obtained by the automated data acquisition and calculation module 3, obtains the change trend of the patient's rehabilitation status, and claims a rehabilitation evaluation report;

[0077] The evaluation results obtained by the automated data acquisition and calculation module 3 are analyzed in depth, compared with the built-in rehabilitation evaluation standard and historical evaluation data in the system, and the change trend of the patient's rehabilitation status is analyzed. According to the analysis result, a detailed rehabilitation evaluation report is automatically generated, and the report content includes the current rehabilitation status evaluation of the patient, the rehabilitation progress summary, the existing problems and the personalized rehabilitation suggestions;

[0078] The output end of the automated data acquisition and calculation module 3 is electrically connected with the input end of the data analysis and report generation module 4, and the data analysis and report generation module 4 includes a data analysis unit 401 and a report generation unit 402. The data analysis unit 401 is used to suggest an analysis model, and the analysis model is used to analyze the evaluation results obtained by the automated data acquisition and calculation module 3 in depth and compare them with the built-in rehabilitation evaluation standard and historical evaluation data in the system, and analyze the change trend of the patient's rehabilitation status. The analysis model adopts an ARIMA model, and specifically:

[0079] Let the time series y t , and the ARIMA(p, d, q) model expression is: wherein, is the autoregressive part; is the difference operator, and is used to make the time series stationary, and Θ(B) = 1 + θ1B +... + θ qB q For the moving average portion; ∈ t The sequence is a white noise sequence, where p is the autoregressive order, d is the difference order, and q is the moving average order.

[0080] Model parameters are estimated using the least squares method. θ i This allows for the prediction of future rehabilitation indicator values. During comparative analysis, the current assessment indicator x is calculated. cur Compared with historical indicators x his The rate of change r is specifically: The rate of change r reflects the dynamic changes in rehabilitation effect. A positive rate of change r indicates that the rehabilitation effect has improved, and the larger the value, the more significant the improvement. A negative rate of change r indicates that the rehabilitation effect has regressed.

[0081] The report generation unit 402 generates an evaluation report based on the analysis of the data analysis unit 401 using the Freemarker template engine and the ECharts chart library;

[0082] The rehabilitation management module 5, based on the rehabilitation suggestions provided by the assessment result analysis and report generation module 4 and combined with the system's built-in rehabilitation plan knowledge base, formulates a rehabilitation plan for the patient. At the same time, the rehabilitation management module 5 is also used to track and manage the implementation of the rehabilitation plan and adjust the rehabilitation plan according to the patient's rehabilitation progress.

[0083] The input terminal of the rehabilitation management module 5 is electrically connected to the output terminal of the data analysis and report generation module 4. The rehabilitation management module 5 includes a treatment plan formulation unit 501 and a treatment plan adjustment unit 502. The treatment plan formulation unit 501 is used to construct a treatment plan formulation model. By constructing the treatment plan formulation model and combining it with the system's built-in rehabilitation treatment plan knowledge base, a rehabilitation plan is formulated for the patient. The rehabilitation treatment plan knowledge base stores historical cases. The treatment plan formulation model executes the following strategies:

[0084] Case-based reasoning (CBR) calculates the current patient feature vector X = [x1, x2, ..., x...]. n ] and historical case feature vector Y j =[y j1 ,y j2 ,...,y jn The similarity sim(X,Y) j This allows for matching with historical cases using the Euclidean distance formula.

[0085] Select historical cases with high similarity as references and adjust the solution in conjunction with the rule engine;

[0086] Let a rule in the rule set be: If a patient's muscle strength is below a threshold T, then the training intensity I is denoted as I. min ,Right now:

[0087] wherein n is the number of dimensions of the features, I min is the minimum training intensity, I default is the default training intensity.

[0088] The scheme adjustment 502 is used to adjust the rehabilitation scheme according to the rehabilitation progress of the patient. The scheme adjustment 502 adopts the Q-learning algorithm in reinforcement learning, and specifically is:

[0089] Let the state space be S and the action space be A. Performing an action a in state s moves to state s' and obtains a reward r'. The Q value update formula is:

[0090]

[0091] wherein a is the learning rate, and a e [0, 1], g is the discount factor, and g e [0, 1], the state space S includes multi-dimensional index features such as patient physiological indicators, rehabilitation process stages, and complication risks, and each s e indicates the current rehabilitation state of the patient, the action space A covers rehabilitation intervention measures such as training plans of different intensities, physiotherapy mode combinations, and drug dose adjustments;

[0092] By continuously iterating and updating the Q value, the rehabilitation scheme adjustment strategy is optimized, and finally a rehabilitation scheme adjustment strategy that maximizes the long-term cumulative reward is formed, so as to realize treatment path dynamic planning.

[0093] The database storage center 6 is used to store system related data, adopts master-slave replication and database and table splitting technology, encrypts data using AES algorithm, and optimizes data query using index coverage optimization and Redis cache technology.

[0094] The system related data in the database storage center 6 includes patient basic information, rehabilitation diagnosis information, dynamic evaluation scale data, evaluation result data, rehabilitation scheme data, and all data related to rehabilitation medical evaluation management.

Claims

1. A rehabilitation medical assessment and management system based on dynamic multi-scale adaptation and automated calculation, characterized in that: The assessment and management system includes: a patient information collection module (1), a dynamic scale adaptation module (2), an automated data collection and calculation module (3), a data analysis and report generation module (4), a rehabilitation management module (5), and a database storage center (6). The patient information collection module (1) is used to collect patient information and store the collected information in the database storage center (6); The dynamic scale adaptation module (2) uses the information obtained by the patient information collection module (1) and a preset adaptation algorithm to dynamically select the most suitable combination of assessment scales for the current rehabilitation stage and condition of the patient from the built-in dynamic multi-scale library. The automated data acquisition and calculation module (3) collects physiological and motor data by connecting with the intelligent medical device, collects dynamic assessment scale response data, and uses the Kalman filter formula for prediction. The data analysis and report generation module (4) performs in-depth analysis on the evaluation results obtained by the automated data acquisition and calculation module (3), obtains the changing trend of the patient's recovery status, and generates a rehabilitation evaluation report. The rehabilitation management module (5) formulates a rehabilitation plan for the patient based on the rehabilitation suggestions provided by the assessment result analysis and report generation module (4) and the built-in rehabilitation plan knowledge base. At the same time, the rehabilitation management module (5) is also used to track and manage the implementation of the rehabilitation plan and adjust the rehabilitation plan according to the patient's rehabilitation progress. The database storage center (6) is used to store system-related data. It adopts master-slave replication and database sharding technology, uses AES algorithm to encrypt data, and uses index coverage optimization and Redis caching technology to optimize data query.

2. A rehabilitation medical assessment and management system based on dynamic multi-scale adaptation and automated calculation, characterized in that: The output of the patient information collection module (1) is electrically connected to the input of the dynamic scale adaptation module (2). When the patient information collection module (1) is connected to the hospital's electronic medical record system, it follows the HL7 standard protocol or FHIR specification. The structured data is automatically extracted through the API interface. The patient information includes the patient's basic information, rehabilitation diagnosis information and current rehabilitation stage information. The information is obtained by connecting to the hospital's electronic medical record system, patient self-entry, and manual input by medical staff. The validity of the age is judged, a flag is set to judge the entry of required words, and the information is stored in the database storage center (6).

3. The rehabilitation medical assessment and management system based on dynamic multi-scale adaptation and automated calculation according to claim 2, characterized in that: The patient information collection module (1) uses the following strategy to determine the validity of the data when judging the validity of the age: Where 1 indicates that the age data is valid, and 0 indicates that it is invalid, the flag bit F is set. i If a field is required and has not been entered, then F i =0, the system triggers a prompt process; if data has already been entered, then F i =1; When the system receives age data, it compares the input value with the valid age range. If the age falls within the range of 0 to 120 years, the data is marked as valid. i =1; conversely, if it exceeds this range, it is considered invalid data, F i =0, the system triggers a prompt process and checks invalid data.

4. The rehabilitation medical assessment and management system based on dynamic multi-scale adaptation and automated calculation according to claim 3, characterized in that: The dynamic multi-scale library stores multiple assessment scales for different rehabilitation areas, including but not limited to functional assessment scales, symptom assessment scales, and quality of life scales. The adaptation algorithm in the dynamic scale adaptation module (2) adopts a strategy of combining decision tree and Bayesian network to build the adaptation algorithm model, and the decision tree uses the patient's disease type, rehabilitation stage, age and gender as decision nodes to screen the scale. A Bayesian network is used to calculate the applicability probability of a scale. Let event A represent selecting a certain scale, and event B represent the patient's feature set. According to the formula: Wherein, P(A|B) is the probability of selecting scale A given that the patient has feature B, P(B|A) is the probability that the patient has feature B given that scale A is selected, P(A) is the prior probability of scale A being selected, and P(B) is the probability that the patient has feature B. The probability values ​​are calculated using historical data, and the branch weights of the decision tree are dynamically adjusted.

5. The rehabilitation medical assessment and management system based on dynamic multi-scale adaptation and automated calculation according to claim 4, characterized in that: The automated data acquisition and calculation module (3) uses the Kalman filter formula to remove noise from the data acquired by the intelligent medical device. The specific steps are as follows: By combining filtered data with kinematic principles, preprocessing and feature extraction are performed on the collected physiological signals and motion data. Prediction steps: Update steps: P k|k =(I-K k H k )P k|k-1 ; in, F is the state prediction value at time k based on time k-1. k Here is the state transition matrix. B is the state estimate at time k-1. k To control the input matrix, u k To control the input, P k|k-1 For the prediction error covariance, Q k For process noise covariance, K k For Kalman gain, H k Let z be the observation matrix. k For the observed value, R k To observe the noise covariance, Let P be the state estimate after time k. k|k This is the updated error covariance; The formula for calculating the range of motion of a joint is: R = max(θ1, θ2, ..., θ n )-min(θ1,θ2,...,θ n ); The assessment results are derived from the data obtained from the scale responses, and the formula is as follows: Where θ1,θ2,...,θ n To collect angle data of the joint at different times, R represents the range of motion of the joint, and w i s represents the weight of a score on a certain item in the scale. i The patient was given a score, where S is the overall score of motor function, α is the weighting coefficient of the motor data, and m is the number of items on the scale.

6. The rehabilitation medical assessment and management system based on dynamic multi-scale adaptation and automated calculation according to claim 5, characterized in that: The output of the automated data acquisition and calculation module (3) is electrically connected to the input of the data analysis and report generation module (4), and the data analysis and report generation module (4) includes a data analysis unit (401) and a report generation unit (402). The data analysis unit (401) is used to suggest an analysis model, and to perform in-depth analysis on the evaluation results obtained by the automated data acquisition and calculation module (3) through the analysis model, and compare them with the rehabilitation evaluation standards and historical evaluation data built into the system to analyze the changing trend of the patient's rehabilitation status. The analysis model adopts the ARIMA model, specifically: Let time series y t The ARIMA(p,d,q) model expression is: in, For the autoregressive part; It is a difference operator used to stationary time series, Θ(B)=1+θ1B+...+θ q B q For the moving average portion; ∈ t The sequence is a white noise sequence, where p is the autoregressive order, d is the difference order, and q is the moving average order. Model parameters are estimated using the least squares method. θ i This allows for the prediction of future rehabilitation indicator values. During comparative analysis, the current assessment indicator x is calculated. cur Compared with historical indicators x his The rate of change r is specifically: The rate of change r reflects the dynamic changes in rehabilitation effect. A positive rate of change r indicates that the rehabilitation effect has improved, and the larger the value, the more significant the improvement. A negative rate of change r indicates that the rehabilitation effect has regressed.

7. The rehabilitation medical assessment and management system based on dynamic multi-scale adaptation and automated calculation according to claim 6, characterized in that: The report generation unit (402) generates an evaluation report based on the analysis of the data analysis unit (401) using the Freemarker template engine and ECharts chart library.

8. The rehabilitation medical assessment and management system based on dynamic multi-scale adaptation and automated calculation according to claim 7, characterized in that: The input end of the rehabilitation management module (5) is electrically connected to the output end of the data analysis and report generation module (4), and the rehabilitation management module (5) includes a plan formulation unit (501) and a plan adjustment unit (502). The plan formulation unit (501) is used to construct a plan formulation model. By constructing the plan formulation model and combining it with the system's built-in rehabilitation plan knowledge base, a rehabilitation plan is formulated for the patient. The rehabilitation plan knowledge base stores historical cases. The plan formulation model executes the following strategies: Case-based reasoning (CBR) calculates the current patient feature vector X = [x1, x2, ..., x...]. n ] and historical case feature vector Y j =[y j1 ,y j2 ,...,y jn The similarity sim(X,Y) j This allows for matching with historical cases using the Euclidean distance formula. Select historical cases with high similarity as references and adjust the solution in conjunction with the rule engine; Let a rule in the rule set be: If a patient's muscle strength is below a threshold T, then the training intensity I is denoted as I. min ,Right now: Where n is the number of dimensions of the feature, I min For minimum training intensity, I default This is the default training intensity.

9. The rehabilitation medical assessment and management system based on dynamic multi-scale adaptation and automated calculation according to claim 8, characterized in that: The proposed adjustment (502) is used to adjust the rehabilitation plan based on the patient's rehabilitation progress. The proposed adjustment (502) employs the Q-learning algorithm in reinforcement learning, specifically as follows: Let the state space be S and the dynamic space be A. In state s, action a transitions to state s′ and yields a reward r′. The Q-value update formula is: Where α is the learning rate, and α∈[0,1], and γ is the discount factor, and γ∈[0,1]. By continuously iterating and updating the Q value, the rehabilitation program adjustment strategy is optimized, ultimately forming a rehabilitation program adjustment strategy that maximizes long-term cumulative rewards, thus achieving dynamic planning of the treatment path.

10. The rehabilitation medical assessment and management system based on dynamic multi-scale adaptation and automated calculation according to claim 9, characterized in that: The system-related data in the database storage center (6) includes patients' basic information, rehabilitation diagnosis information, dynamic assessment scale data, assessment result data, rehabilitation plan data, and all data related to rehabilitation medical assessment and management.

Images