Rehabilitation immediate adaptation mode construction method based on SVI-sEMG dual-mode correction

The SVI-sEMG dual-mode correction rehabilitation immediate adaptation mode, combined with surface electromyography and stroke volume index, solves the problem of high cardiovascular risk in ICU patients' rehabilitation exercises, realizes safe and effective rehabilitation assessment and dynamic adjustment, and improves rehabilitation efficiency.

CN118942626APending Publication Date: 2024-11-12THE SECOND HOSPITAL OF SHANDONG UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410981189.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-22
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

Existing technologies lack systematic tolerance assessment in rehabilitation exercises for ICU patients, leading to cardiovascular autonomic imbalance and abnormal myocardial blood supply. Furthermore, relying on heart rate and blood pressure assessments has large errors and cannot effectively monitor cardiac load, increasing the risks of cardiopulmonary exercise testing.

Method used

A rehabilitation immediate adaptation mode based on SVI-sEMG dual-mode correction is adopted. Combining surface electromyography (sEMG) and stroke volume index (SVI), the intensity assessment model is established by denoising and standardizing the signal through short-time Fourier transform, wavelet transform, bandpass filter and wavelet transform. Kalman filtering and fuzzy rules are used to adjust the intensity of rehabilitation exercises and dynamically adjust the safety threshold to achieve immediate adaptation and safety assessment.

Benefits of technology

Effective monitoring of ICU patients' exercise status and cardiac load ensures the safety and effectiveness of rehabilitation exercises, reduces cardiovascular risk, and improves rehabilitation efficiency and patients' physiological safety.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118942626A_ABST
    Figure CN118942626A_ABST
Patent Text Reader

Abstract

The invention belongs to the technical field of medical rehabilitation nursing, and relates to a rehabilitation immediate adaptation mode construction method based on SVI-sEMG dual-mode correction, which comprises the steps of patient resting state data acquisition, resting data preprocessing, patient motion state data acquisition, motion data preprocessing, motion intensity prediction, intensity adjustment and motion intensity prediction. And generating an intensity adjusting exercise intensity scheme. According to the method, multiple monitoring means of the SVI, the sEMG and the respiratory function are combined, reasonable threshold value parameters are established, the SVI and sEMG baseline values of the patient are collected, the exercise intensity and duration are adjusted in time according to the SVI, sEMG and respiratory function index changes during the upper limb rehabilitation period of the patient, the current state of the patient is adapted, and the rehabilitation effect of the patient is improved. A new method is provided for medical staff to formulate a reasonable maximum rehabilitation prescription, physiological and pathological safety of a patient is guaranteed, and rehabilitation efficiency of the patient is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of medical rehabilitation nursing, and in particular relates to a method for constructing an instant adaptation mode of rehabilitation based on SVI-sEMG dual-mode correction. Background Art

[0002] ICU-acquired weakness (ICU-AW) is a clinical syndrome characterized by widespread limb weakness that develops during a patient's stay in an intensive care unit (ICU) and cannot be explained by other causes other than the critical illness. [1] The incidence rate can be as high as 50%-100% [2] Once patients develop ICU-acquired weakness, the duration of mechanical ventilation and ICU hospitalization will be prolonged, the quality of life will be reduced, the mortality rate will increase, and the economic burden of treatment will increase. However, this abnormality of muscle strength and function is usually reversible in the early stages. Around 2000, Western European scholars proposed the theory of early rehabilitation of ICU patients. [3] , aiming to slow down skeletal muscle atrophy in critically ill patients and prevent the occurrence of ICU-AW. In 2011, Professor LevineS [4] Through AKT-FOXO signaling, it was determined that the rate of disuse atrophy of the diaphragm in ICU patients undergoing ventilator-assisted ventilation was 8 times that of skeletal muscle. [5] It is pointed out that rehabilitation measures such as using breathing trainers, adjusting patient position, and pursed lip breathing can improve lung compliance, increase functional residual capacity, and prevent alveolar collapse. At the same time, corresponding upper limb exercises, especially activities that involve the pectoralis major muscles, can increase respiratory muscle strength, promote lung re-expansion, and help improve ventilation.

[0003] ICU patients require more life support equipment and more indwelling catheters due to the severity and complexity of their illnesses. They are also usually required to be immobilized or rest in bed due to their condition. The purpose is to avoid the risks of patient activities to a certain extent. Currently, early rehabilitation activities are mainly based on lower limb rehabilitation, mainly for patients with neurological injuries. There are two main types of rehabilitation for ICU patients: passive rehabilitation exercises and active rehabilitation exercises. Currently, early rehabilitation for ICU patients is mostly passive rehabilitation exercises led by rehabilitation therapists. [6] There is still a lack of systematic tolerance assessment, and the intensity and time of training are mainly increased or decreased based on the changes in the patient's heart rate and blood pressure. However, ICU patients generally have abnormal heart and respiratory function. Professor Milani JGPO [7]A study found that for critically ill patients undergoing rehabilitation exercises, especially those treated with B-receptor antagonists, there is a large error in using the maximum heart rate model to evaluate the exercise effect. When ICU patients are doing rehabilitation exercises, if there is an error in the maximum heart rate assessment or the threshold is set improperly, cardiovascular autonomic nerve imbalance may occur, causing abnormal myocardial blood supply, and may even reduce the patient's cerebral blood flow and oxygen saturation. HellerydE, ​​LinPY and many other scholars [8-10] The research results of the National Cancer Institute all agree with this view, and indicate that this situation can cause myocardial damage in patients, even sudden death, and aggravate the patient's condition.

[11] Research has pointed out that cardiopulmonary exercise testing (CPET) is the most effective cardiopulmonary safety device for evaluating patients with abnormal heart and respiratory function during exercise, but it is not suitable for ICU patients, especially ICU patients supported by mechanical ventilation.

[0004] With the deepening of rehabilitation medicine research, HashemMD et al. [12-13] Scholars have found that active rehabilitation exercises can enhance patients' confidence in fighting the disease, reduce the incidence of delirium and bring many other psychological constructive advantages. However, the exercise process mainly relies on the patient's autonomous expression to judge the fatigue of a single training session, so the rehabilitation effect is not significant. Surface electromyography (sEMG) is the combined effect of superficial muscle EMG and electrical activity on the nerve trunk on the surface of the skin. Its signal is a one-dimensional voltage time series signal obtained by guiding, amplifying, displaying and recording the bioelectric changes of the neuromuscular system during voluntary and involuntary activities through surface electrodes. Its amplitude is about 0-5000uV and the frequency is 30-350Hz. The changes in the amplitude and frequency characteristics of the signal depend on the combined effects of physiological factors (motor unit activity, muscle fiber recruitment and decreased muscle fiber excitation conduction velocity under different muscle activity levels and functional states) and measurement factors (detection electrode position, signal stringing, skin temperature, muscle length and muscle contraction mode). Professor Xiao Jinzhuang, Professor Larsen S and other scholars [14-16] The research results show that sEMG can directly reflect the changes in muscle activity level and muscle fatigue through the electrode signals of each target muscle, thereby guiding the effect of active rehabilitation exercise for ICU patients and avoiding excessive muscle fatigue.

[17] It is pointed out that excessive strength training will cause a large amount of lactic acid to be produced in the human body, which may lead to lactic acidosis in ICU patients and even affect the cardiovascular system: excessive muscle fatigue will lead to the release of inflammatory mediator interleukin-6 (IL-6). ICU patients generally have low resistance and are at risk of infection. The release of IL-6 will lead to an aggravation of the inflammatory response. Therefore, evaluating the patient's exercise status through sEMG can effectively evaluate the patient's tolerance to single training, improve rehabilitation effects, and reduce the incidence of complications.

[0005] Rehman A

[18] Studies have shown that chest expansion exercises have a more significant passive stretching effect on respiratory muscles than lower limb exercises. Active upper limb rehabilitation exercises for ICU patients can effectively exercise target muscle groups, promote the recovery of cardiopulmonary function, and significantly reduce the incidence of ICU-AW. [7] Professor Kho ME

[19] Both conducted research on the maximum exercise prescription for patients with heart failure and ICU patients. The results showed that relying solely on heart rate and blood pressure cannot meet the safety assessment of patients' rehabilitation exercises. Heredia-Elvar JR, Barbado D and other scholars [20-21] It is proposed that a variety of monitoring methods can be used to develop safer thresholds, and personalized exercise intensity can be developed in an intelligent way. By establishing reasonable value parameters, the training intensity and duration can be quickly adjusted according to the patient's specific indicators, helping medical staff to develop reasonable maximum rehabilitation prescriptions, while ensuring the patient's physiological and pathological safety and improving their rehabilitation efficiency.

[0006] In summary, a method for constructing an immediate adaptation model of rehabilitation based on SVI-sEMG dual-mode correction was proposed. Summary of the invention

[0007] The purpose of the present invention is to provide a method for constructing an instant adaptation mode of rehabilitation based on SVI-sEMG dual-mode correction, which has active upper limb rehabilitation exercise monitoring and evaluation based on surface electromyography and an instant adaptation mode based on dual-mode correction, and studies the role of the stroke volume index SVI in non-invasive cardiac output in monitoring the cardiac load in exercise state, so as to provide a new method for monitoring the cardiac load during the rehabilitation process of ICU patients.

[0008] In order to achieve the above object, the technical solution adopted by the present invention is that the present invention provides a method for constructing an immediate adaptation mode of rehabilitation based on SVI-sEMG dual-mode correction, comprising the following steps:

[0009] 1) Start;

[0010] 2) Data collection of patients in resting state;

[0011] 3) Resting data preprocessing;

[0012] 4) Patient movement status data collection;

[0013] 5) Motion data preprocessing;

[0014] 6) Exercise intensity prediction;

[0015] 7) Adjust the intensity of exercise prediction;

[0016] 8) Generate an intensity-adjusted exercise plan;

[0017] 9) End.

[0018] Specifically, the preprocessing methods in step 3) and step 5) are extraction, noise filtering and standardization.

[0019] Specifically, the extraction method is to use short-time Fourier transform or wavelet transform to analyze the time-frequency characteristics of sEMG signals and extract the features of SVI signals; the denoising method is to use a bandpass filter to eliminate low-frequency baseline drift and high-frequency noise in sEMG signals and use a wavelet transform denoising method to remove artifacts and noise of SVI signals; the standardization method is to process sEMG and SVI signals using a standardized formula to eliminate differences between individuals, and the standardized formula is:

[0020] Where μ and σ are the mean and standard deviation of the signal, respectively.

[0021] Specifically, the exercise intensity prediction in step 6) includes establishing an intensity assessment model, data fusion, and feedback regulation.

[0022] Specifically, the establishment of the intensity assessment model includes modal effective value and multimodal intensity assessment, wherein the multimodal intensity assessment adopts a stacking method combined with linear regression and random forest to estimate the intensity of rehabilitation exercises.

[0023] Specifically, the data fusion uses Kalman filtering through real-time state estimation and recursive updating, and verifies and makes a prediction plan for intensity adjustment of rehabilitation exercises based on preliminary prediction results of multimodal intensity assessment.

[0024] Specifically, the feedback regulation adjusts the intensity of rehabilitation exercise through fuzzy rules.

[0025] Specifically, it includes dynamic threshold judgment, dynamically adjusting the safety threshold based on historical data and current trends.

[0026] Specifically, the implementation steps include:

[0027] 1. Start;

[0028] 2. Motion parameter detection;

[0029] 3. Prediction of its own strength;

[0030] 4. Revision of self-strength prediction;

[0031] 5. Adjust the intensity plan forecast;

[0032] 6. Correct and adjust the intensity plan;

[0033] 7. Adjust exercise prescription;

[0034] 8. Safety monitoring;

[0035] 9. End.

[0036] Specifically, the dynamic threshold of the sEMG signal sEMG,dynamic (t), defined as:

[0037] Threshold sEMG,dynamic (t) = μ sEMG (t)+k·σ sEMG (t)

[0038] Dynamic threshold of SVI signal SVI,dynamic (t), defined as:

[0039] Threshold SVI,dynamic (t) = μ SVI (t)+k·σ SVI (t)

[0040] Where k is the threshold adjustment parameter;

[0041] For the moving average μ of the sEMG signal sEMG (t), defined as:

[0042]

[0043] For the moving average μ of the SVI signal SVI (t), defined as:

[0044]

[0045] The standard deviation of the sEMG signal σ sEMG (t), defined as:

[0046]

[0047] For the standard deviation of the SVI signal σ SVI (t), defined as:

[0048]

[0049] Among them, X sEMG (i) and X SVI (i) are the sEMG signal value and SVI signal value at the i-th moment, respectively, and w is the calculation window.

[0050] Compared with the prior art, the advantages and positive effects of the present invention are:

[0051] 1. The present invention combines the stroke volume index SVI, surface electromyography sEMG and respiratory function monitoring methods to establish reasonable threshold parameters. By collecting the patient's SVI and sEMG baseline values, the patient can adjust the exercise intensity and duration in real time according to the changes in SVI, sEMG and respiratory function indicators during the patient's autonomous upper limb rehabilitation to adapt to the patient's current state, providing a new method for medical staff to formulate a reasonable maximum rehabilitation prescription, while ensuring the patient's physiological and pathological safety and improving their rehabilitation efficiency;

[0052] 2. The present invention has active upper limb rehabilitation exercise monitoring and evaluation based on surface electromyography and an instant adaptation mode based on dual-mode correction. It studies the role of the stroke volume index SVI in non-invasive cardiac output in monitoring the cardiac load during exercise, and provides a new method for monitoring the cardiac load during the rehabilitation process of ICU patients. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0054] Figure 1 This is a flow chart of an embodiment of the present invention;

[0055] Figure 2 This is a flow chart of an embodiment of the present invention;

[0056] Figure 3 The figure is a flow chart of upper limb rehabilitation training according to an embodiment of the present invention. DETAILED DESCRIPTION

[0057] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described below in conjunction with the accompanying drawings and embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0058] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways than those described herein. Therefore, the present invention is not limited to the specific embodiments of the following disclosure.

[0059] Embodiment 1, as Figure 1 As shown, a method for constructing an immediate adaptation mode of rehabilitation based on SVI-sEMG dual-mode correction includes the following steps:

[0060] 1) Start;

[0061] 2) Data collection of patients in resting state;

[0062] 3) Resting data preprocessing;

[0063] 4) Patient movement status data collection;

[0064] 5) Motion data preprocessing;

[0065] 6) Exercise intensity prediction;

[0066] 7) Adjust the intensity of exercise prediction;

[0067] 8) Generate an intensity-adjusted exercise plan;

[0068] 9) End.

[0069] The following is a detailed description of the specific design of the above key steps:

[0070] Specifically, the preprocessing methods in step 3) and step 5) are extraction, noise filtering and standardization.

[0071] Specifically, the extraction method is to use short-time Fourier transform or wavelet transform to analyze the time-frequency characteristics of sEMG signals and extract the features of SVI signals; the noise filtering method is to use a bandpass filter to eliminate low-frequency baseline drift and high-frequency noise in sEMG signals and use a wavelet transform denoising method to remove artifacts and noise of SVI signals;

[0072] Instantaneous frequency and time-frequency analysis: Short-time Fourier transform (STFT) or convolutional wavelet transform (CWT) is used to analyze the time-frequency characteristics of sEMG signals.

[0073] The short-time Fourier transform (STFT) is defined as:

[0074]

[0075] Among them, x(t) is the original signal; w(τ―t) is the window function; f represents the frequency; and X(t,f) is the representation in the time-frequency domain.

[0076] The wavelet transform (CWT) is defined as:

[0077]

[0078] in is the mother wavelet function at scale s.

[0079] SVI signal feature extraction

[0080] (1) Stroke volume SV: reflects the amount of blood pumped out by the heart during each contraction.

[0081]

[0082] Among them, CO is cardiac output and HR is heart rate.

[0083] (2) Cardiac output (CO): reflects the amount of blood pumped by the heart per unit time.

[0084] CO=SV×HR

[0085] During the treatment and nursing process, critically ill patients are often accompanied by ECG monitoring, ventilator-assisted ventilation, and even CRRT kidney filtration, IABP, and ECOM heart support. These methods will generate a large number of interference signals, which will have a great interference on sEMG signals.

[0086] Bandpass filter: used to eliminate low-frequency baseline drift and high-frequency noise in sEMG signals. The formula is:

[0087] y(t)=h(t)*x(t)

[0088] Where y(t) is the filtered signal, x(t) is the original sEMG signal, and h(t) is the impulse response of the filter.

[0089] Denoising algorithm: Use wavelet transform or other signal denoising methods to remove artifacts and noise.

[0090] The formula is:

[0091]

[0092] in, is the denoised sEMG signal, W i and φ i (t) are wavelet coefficients and basis functions respectively.

[0093] The standardization method is to process the sEMG and SVI signals using a standardization formula to eliminate differences between individuals. The standardization formula is:

[0094] Where μ and μ are the mean and standard deviation of the signal, respectively.

[0095] Specifically, the exercise intensity prediction in step 6) includes establishing an intensity assessment model, data fusion, and feedback regulation.

[0096] Specifically, the establishment of the intensity assessment model includes modal effective value and multimodal intensity assessment, wherein the multimodal intensity assessment adopts a stacking method combined with linear regression and random forest to estimate the intensity of rehabilitation exercises.

[0097] Modal effective value AEMV: AEMV can be used to monitor muscle load in real time, mainly to identify signs of muscle fatigue. By calculating the MEV of the sEMG signal, the actual force or activity intensity of the muscle during rehabilitation training can be obtained.

[0098] The formula is:

[0099]

[0100] in,

[0101] Cumulative sum of squares of signals from i from 1 to N; x i is the value of the sEMG signal at the i-th time step. The sEMG signal contains the muscle electrical activity data captured by the electrodes from the skin surface; N is the total number of sampling points of the signal; dt is the sampling time interval, that is, the time interval in signal capture (sampling period).

[0102] Multimodal intensity assessment: Preliminary prediction of the current rehabilitation intensity of critically ill patients through sEMG data and stroke volume index (SVI). After preprocessing the sEMG data through denoising and standardization, the stacking method (Stacking) is used in combination with linear regression and random forest to estimate the intensity of rehabilitation exercises, which is used to deal with the nonlinear relationship that may occur in the rehabilitation process of critically ill patients and improve the robustness of the overall model to noise and outliers in the data.

[0103] The formula for the linear regression model is:

[0104] y=β0+β1X sEMG +β2X SVI +∈

[0105] in,

[0106] y is the target variable, i.e., the predicted value of rehabilitation training effect or physical fitness level; β0 is the intercept term, which indicates the expected value of the target variable when all explanatory variables are zero; β1 is the regression coefficient of sEMG signal, which indicates the influence of sEMG signal on the target variable; β2 is the regression coefficient of stroke volume index (SVI), which indicates the influence of SVI on the target variable; X sEMGis the sEMG signal data, which is one of the explanatory variables; X SVI is the stroke volume index SVI data, which is one of the explanatory variables; ∈ is the error term, which represents other unexplained random errors.

[0107] The formula for the random forest model is:

[0108]

[0109] in,

[0110] y RF is the final prediction result of the random forest model; N is the number of decision trees in the random forest;

[0111] For the i-th decision tree, input X sEMG and X SVI prediction results.

[0112] By combining the prediction results of multiple decision trees, random forests can improve the prediction accuracy and stability of the model and effectively deal with high-dimensional data and noise interference.

[0113] Specifically, the data fusion uses Kalman filtering through real-time state estimation and recursive updating, and verifies and makes a prediction plan for intensity adjustment of rehabilitation exercises based on preliminary prediction results of multimodal intensity assessment.

[0114] The Kalman filter verifies and makes a prediction plan for the intensity adjustment of rehabilitation exercises based on the preliminary prediction results of multimodal intensity assessment through real-time state estimation and recursive updating.

[0115] 1) State prediction:

[0116] The formula is:

[0117]

[0118] in,

[0119] is the prior estimate of the state at time k (using the information at time k-1), which is the prediction of the state before a new observation is made at time kk; F is the state transfer matrix (or state transfer model), which is used to transfer the state from time k-1 to time k, reflecting the dynamic changes of the system within one time step.

[0120] is the a posteriori estimate of the state at time k-1 (combined with the observation data at time k-1); B is the control input matrix, mapping the control input to the state space; u k is the control input, the external control variable that affects the system state at time k.

[0121] 2) Covariance prediction:

[0122] The formula is:

[0123] P k∣k―1 =FP k―1∣k―1 F T +Q

[0124] in,

[0125] P k∣k―1 is the prediction error covariance matrix at time k, which is based on the error covariance matrix P updated at time k-1 k―1∣k―1 And the system noise covariance matrix Q predicted; P k―1∣k―1 is the error covariance matrix after the update at time k-1.

[0126] F is the state transfer matrix; Q is the process noise covariance matrix, which describes the statistical characteristics of the system process noise.

[0127] 3) Update:

[0128] Kalman gain formula:

[0129] K k =P k∣k―1 H T (HP k∣k―1 H T +R) ―1

[0130] in,

[0131] K k is the Kalman gain matrix at time k, which describes the weight balance between the predicted value and the measured value;

[0132] P k∣k―1 is the prediction error covariance matrix at time k;

[0133] H is the observation matrix, which describes how to map from system states to measurements;

[0134] R is the measurement noise covariance matrix, which describes the statistical characteristics of the measurement noise;

[0135] (HP k∣k―1 H T +R) ―1 It is the inverse matrix of the weighting matrix after the prediction covariance is mapped by the measurement.

[0136] Status update formula:

[0137]

[0138] in,

[0139] is the updated state estimate at time k, which is combined with the predicted state and measurement data z k Corrected is the predicted state at time kk;

[0140] K k is the Kalman gain matrix at time k; z k is the actual measured value at time k;

[0141] is the predicted measurement value, that is, the predicted state is mapped to the value of the measurement space through the observation matrix HH;

[0142] To measure residuals or innovation, it is the difference between the actual measured value and the predicted measured value.

[0143] Covariance update equation formula:

[0144] P k∣k =(I-K k H)P k∣k―1

[0145] in,

[0146] P k∣k is the updated state covariance matrix at time k; I is the identity matrix with the same dimension as the state vector; K k is the Kalman gain matrix, which indicates the relative weight of observation information to state estimation; H is the observation matrix, which maps the state space to the observation space; P k∣k―1 is the predicted state covariance matrix at time kk.

[0147] Overall analysis of the Kalman filter formula:

[0148] The prediction results of linear regression and random forest models are introduced into the observation update process of Kalman filter to correct the target variables.

[0149] Status prediction Indicators indicating the patient's current recovery status, namely: muscle activity (one of the internal indicators of sEMG);

[0150] The state transfer matrix F is used to describe the transfer characteristics of the system from the previous time step to the current time step, namely: muscle fatigue (one of the internal indicators of sEMG);

[0151] Control input u k Represents the external control variables in rehabilitation movements, namely: equipment-assisted force;

[0152] The control input matrix B represents the influence of the control input on the system state;

[0153] Error covariance P is the covariance of the estimation error, which is used to measure the uncertainty of the prediction;

[0154] Measurement value z k Real-time data from sEMG and SVI sensors;

[0155] The observation matrix H maps the true measurement values ​​to the state space;

[0156] Kalman gain K k Used to weight predictions and actual measurements to update state estimates;

[0157] The noise covariance matrix Q is the covariance matrix of process noise, and R is the covariance matrix of measurement noise. These two matrices describe the noise characteristics of the system and measurement.

[0158] Specifically, the feedback regulation adjusts the intensity of rehabilitation exercise through fuzzy rules.

[0159] Fuzzy logic control is a control method based on fuzzy rules, which is suitable for handling the dynamic adjustment of rehabilitation exercise intensity. The following is a written explanation of the main concepts and formulas of fuzzy logic control:

[0160] Fuzzy rule formula:

[0161] IFerror is A i AND error rate is B j ,THE N output=C ij

[0162] explain:

[0163] Error (error): is the difference between the current output of the system and the target value (set value).

[0164] Error Rate (error rate): The speed at which the error changes, that is, the rate at which the error changes over time.

[0165] A i (Fuzzy Set of Error): Fuzzy set of error, used to quantify the error into fuzzy variables, including Negative Large, Negative Small, Zero, Positive Small, and Positive Large.

[0166] B j(Fuzzy set of error change rate): Fuzzy set of error change rate, including "negative large", "negative small", "zero", "positive small" and "positive large".

[0167] C ij (Output control quantity): The control output given according to the fuzzy rules is used to adjust the system. The control output is also expressed in fuzzy variables and is the error set A i and error change rate set B j function.

[0168] Specific explanation:

[0169] Error represents the difference between the current output of the system and the target value (set value). For example, in rehabilitation training, the target value may be the patient's ideal muscle activity, while the current output is the actual measured muscle activity.

[0170] The Error Rate indicates the speed of change of the error, that is, the rate of change of the error over time. If the patient's muscle activity approaches the target value quickly, the error change rate is large.

[0171] The fuzzy set A of the error i Used to quantify the error into fuzzy variables. Fuzzy sets include:

[0172] Negative Large, Negative Small, Zero, Positive Small, Positive Large,

[0173] Fuzzy set B of error change rate j Used to quantify the error change rate into a fuzzy variable. Fuzzy sets also include Negative Large, Negative Small, Zero, Positive Small, and Positive Large.

[0174] Output control quantity C ij The control output given by the fuzzy rules is used to adjust the system. The control output is also expressed as a fuzzy variable. For example, the output control quantity can be the adjustment of the auxiliary force of the equipment, quantified as a fuzzy variable, such as "increase by a large amount", "increase by a small amount", "unchanged", "decrease by a small amount", "decrease by a large amount".

[0175] An example of fuzzy rules, assuming the fuzzy rule base is as follows:

[0176] If the error is "negative and large" and the error change rate is "negative and small", the output control amount is "reduced and large";

[0177] If the error is "zero" and the error change rate is "zero", the output control quantity is "remains unchanged";

[0178] If the error is "positive small" and the error change rate is "positive large", the output control amount is "increase small".

[0179] By using fuzzy logic control, the error and error change rate can be converted into fuzzy variables, and the fuzzy rule base can be applied to determine the appropriate output control amount. This method can dynamically adjust the auxiliary strength of the rehabilitation equipment according to the real-time measurement data (data from sEMG and SVI sensors) during rehabilitation exercise, thereby more accurately controlling the patient's exercise intensity and promoting rehabilitation effects.

[0180] Embodiment 2, as Figure 2 As shown, a method for constructing an immediate adaptation mode of rehabilitation based on SVI-sEMG dual-mode correction includes the following steps:

[0181] 1. Start;

[0182] 2. Motion parameter detection;

[0183] 3. Prediction of its own strength;

[0184] 4. Revision of self-strength prediction;

[0185] 5. Adjust the intensity plan forecast;

[0186] 6. Correct and adjust the intensity plan;

[0187] 7. Adjust exercise prescription;

[0188] 8. Safety monitoring;

[0189] 9. End.

[0190] Some steps are implemented according to the scheme described in Reference Example 1.

[0191] Specifically, the safety monitoring in step 8 includes dynamic threshold determination, and the safety threshold is dynamically adjusted based on historical data and current trends.

[0192] Dynamic threshold of sEMG signal sEMG,dynamic (t), defined as:

[0193] Threshold sEMG,dynamic (t) = μ sEMG (t)+k·σ sEMG (t)

[0194] Dynamic threshold of SVI signal SVI,dynamic (t), defined as:

[0195] Threshold SVI,dynamic (t) = μ SVI (t)+k·σ SVI (t)

[0196] Where k is the threshold adjustment parameter;

[0197] For the moving average μ of the sEMG signal sEMG (t), defined as:

[0198]

[0199] For the moving average μ of the SVI signal SVI (t), defined as:

[0200]

[0201] The standard deviation of the sEMG signal σ sEMG (t), defined as:

[0202]

[0203] For the standard deviation of the SVI signal σ SVI (t), defined as:

[0204]

[0205] Among them, X sEMG (i) and X SVI (i) are the sEMG signal value and SVI signal value at the i-th moment, respectively, and w is the calculation window.

[0206] The present invention involves the design of a basic intelligent autonomous rehabilitation training device that can intelligently adjust the intensity and duration of activities according to the instant adaptation mode. After completing the construction of the active rehabilitation instant adaptation mode based on SVI-sEMG dual-mode correction and forming a safety threshold monitoring and adjustment framework, the design can complete the research and development of intelligent critically ill conscious patients' autonomous function training rehabilitation equipment through the introduction of newly formed monitoring, algorithm patents and software processing soft science. The current basic equipment applies to patents: Autonomous function training rehabilitation equipment for bedridden patients [P]. Shandong Province: CN219376006U, 2023-07-21.

[0207] Example 3, as shown in Example 1 or 2, a method for constructing a rehabilitation instant adaptation model based on SVI-sEMG dual-mode correction is designed according to the technical content. The specific design content is as follows:

[0208] 1) Selection of experimental subjects

[0209] This experiment adopted the purposive sampling method and selected mechanically ventilated patients admitted to the intensive care unit (ICU) of a tertiary-level general hospital in Shandong Province as the research subjects. Baseline data such as patient age, gender, SVI and sEMG of patients before the experiment were collected.

[0210] The inclusion criteria for the research subjects were: ① age > 18 years old; ② patients with mechanical ventilation in the ICU; ③ agreed to participate in this experiment, and the patient himself or his immediate family members signed the relevant informed consent form.

[0211] Exclusion criteria: ① Patients with myocardial infarction, heart failure and heart failure; ② Patients with joint lesions; ③ Patients with myasthenia gravis; ④ Patients with pulmonary parenchymal lesions such as partial or total lung resection. The hospital ethics committee approved this experiment before implementation.

[0212] 2) Experimental methods

[0213] Phase 1: Pre-experimental stage

[0214] In order to preliminarily form an immediate adaptation model for upper limb rehabilitation with SVI-sEMG dual-mode correction, we used the convenience sampling method in the early stage to recruit 15-30 research subjects from a tertiary-level A general hospital in Jinan City, Shandong Province from January to June 2024.

[0215] Collect baseline data of the pre-investigation subjects: ① Demographic data: gender, age, education level, occupation, marital status, place of residence, type of medical insurance, personal monthly income, etc.; ② Other relevant data: SVI before and after the experiment and sEMG related indicators before and after the experiment.

[0216] Inclusion criteria for the research subjects: ① Age > 18 years old; ② Normal cognitive and communication abilities, able to clearly express their own feelings; ③ Voluntarily participate in this experiment and sign the informed consent.

[0217] Exclusion criteria: ① patients with cognitive and communication disorders; ② patients with other serious physical diseases, such as severe heart, liver, brain and kidney dysfunction; ③ patients with muscle weakness; ④ patients who are unable to cooperate with the completion of the trial or drop out midway.

[0218] Phase II: Validation of the immediate adaptation model for upper limb rehabilitation using SVI-sEMG dual-mode correction

[0219] This experiment adopted a randomized controlled trial research design. Patients were numbered according to the order of admission and the subjects were divided into a control group and an experimental group using a random number table. The specific intervention plan is as follows:

[0220] ① Control group: Mechanically ventilated patients who received routine upper limb rehabilitation exercises, that is, rehabilitation technicians assisted critically ill patients in completing upper limb rehabilitation training, twice a day, 30 minutes each time.

[0221] ② Experimental group: Mechanically ventilated patients who received upper limb rehabilitation training using the SVI-sEMG dual-mode correction upper limb rehabilitation instant adaptation mode formed in the early stage, that is, professional medical staff used the independently developed patented autonomous functional exercise rehabilitation device CN219376006U for bedridden patients to help patients complete rehabilitation training twice a day, but the intensity and time of each training session were adjusted according to the model threshold formed in the early stage. At this time, as far as the rehabilitation training method for mechanically ventilated patients is concerned, medical staff can set relevant parameters on the functional exercise rehabilitation device according to the threshold model to ensure that patients can perform upper limb rehabilitation training within a safe range of activity.

[0222] sENG electromyographic signals exist on the surface of the human body and inside the muscles. To obtain electromyographic signals, electrodes are needed to export the signals. In this experiment, MyoSstem1400 is used as the sEMG signal acquisition device. The device has the function of wireless sampling of electromyographic signals and the sampling frequency is optional, which meets the hardware requirements of this experiment. Before connecting the sEMG, use 75% alcohol to clean the skin surface of the chest area to reduce skin resistance and reduce interference. Generally, the impedance value is controlled at 1-20KΩ, and it is confirmed that the electrode contact is good and not easy to fall off. Silver Ag / AgCI wire electrodes are used as signal acquisition electrodes. The position of the surface electrode is placed: connect the sEMG electrode wire to the pectoralis major on both sides of the patient, with a distance of 2c between the two electrodes. When the patient's upper limb muscles are relaxed and contracted, the pectoralis major electrodes on both sides are used to collect the original data such as electromyographic information and inertial information of the pectoralis major channels on both sides in real time:

[0223] ① Pre-baseline test: In this experiment, the sEMG baseline data of the subjects before upper limb rehabilitation training were obtained, and all motor unit action potential (MUAP) signals of the patients' pectoralis major muscles during isometric contraction at 1% of the maximum volume contraction force were recorded;

[0224] ②5-10 minutes after the start of rehabilitation training, the data of sEMG-related observation indicators were collected again. The signal sampling frequency range was set to 50-200Hz during the entire experiment. In order to reduce the influence of individual differences, the sEMG data was normalized after feature extraction. The calculation formula is as follows: Pnormalization = P / Pmax-Pstatic.

[0225] In addition, in order to ensure the reliability of the evaluation results, the upper limb rehabilitation training equipment has set the gradient value for the strength of the patient's upper limbs. During the signal acquisition process, the patient's upper limb training requires the patient to complete the movement as much as possible. If the patient's strength is too weak to trigger the minimum threshold of the gradient of the upper limb rehabilitation equipment, the equipment will pull the patient's upper limb for rehabilitation training; if the patient's upper limb strength reaches the minimum threshold of the gradient, the equipment will not be triggered to pull the patient's upper limb for rehabilitation training, and the patient will complete the entire training process independently. Figure 3 This is the flow chart of upper limb rehabilitation training for the experimental group.

[0226] 2) Observation indicators

[0227] Before the test, the upper limb muscles of the subjects were in a non-fatigued state, and the sEMG data of the patients were collected; according to the experimental design, the sEMG data of the patients were collected again 5-10 minutes after the start of the test. The sEMG related data included: muscle activity level and muscle fatigue degree, which had obvious signal responses, so the SVI-sEMG signal was used to evaluate the work-rest state.

[0228] The standard parts used in the present invention can all be purchased from the market, and special-shaped parts can be customized according to the description in the specification and the drawings. The specific connection methods of each part adopt conventional means such as mature bolts, rivets, welding, etc. in the prior art. Machinery, parts and equipment all adopt conventional models in the prior art, and the circuit connection adopts the conventional connection method in the prior art. It will not be described in detail here. The content not described in detail in this specification belongs to the prior art known to professional and technical personnel in this field.

[0229] The above description is only a preferred embodiment of the present invention and does not limit the present invention in other forms. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

[0230] Main references:

[0231] [1] Jones Jennifer RA, Karahalios Amalia, Puthucheary Zudin A etal. Responsiveness of Critically Ill Adults With Multimorbidity to Rehabilitation Interventions: A Patient-Level Meta-Analysis Using IndividualPooled Data From Four Randomized Trials.[J]·Crit Care Med, 2023,51:1373-1385.

[0232] [2] Yang Xia, Yu Hongjing, Pan Zelin, Liang Ziyang, Ye Minyi, He Zhili, Mei Chunan, Qin Xiangjun, Ling Donglan, Chen Lifen. Construction and application of ICU-acquired myasthenia risk prediction model. [J]. Nursing Research, 2023, 37(15): 2673-2683.

[0233] [3] Kondo Yutaka, Fuke Ryota, Hifumi Toru et al. Early rehabilitation for the prevention of postintensive care syndrome in critically ill patients: astudy protocol for a systematic review and meta-analysis. [J]. BMJ Open, 2017, 7: e013828.

[0234] [4]Levine Sanford,Biswas Chhanda,Dierov Jamil et al.Increasedproteolysis,myosin depletion,and atrophic AKT-FOXO signaling in humandiaphragm disuse.[J].Am J Respir Crit Care Med,2011,183:483-490.

[0235] [5]Jang Myung Hun,Shin Myung-Jun,Shin Yong Beom,Pulmonary andPhysical Rehabiliation in Critically Il l Patients.[J].Acute Crit Care,2019,34:1-13.

[0236] [6]Doiron Ratherine A,Hoffmann Tammy C,Beller Elaine M,Earlyintervention(mobilization or active exercise)for critically ill adults in theintensive care unit.[J].Cochrane Database Syst Rev,2018,3:CD010754·

[0237] [7]Milani Juliana GouI art Prata Oliveira,Milani Mauricio,CiprianoGraziella Franga Bernardelli et al.Estimation of the maximal heart rate forexercise prescription in heart failure patients:are we there yet?[J]·Eur JPrev Cardiol,2023,30:419-421.

[0238] [8]Helleryd Edvin,Rawshani Araz,Rawshani Aidin et al.Associationbetween exercise load,resting heart rate,and maximum heart rate and risk offuture ST-segment elevation myocardial infarction(STEMI)·[J].Open Heart,2023,10:e002307.

[0239] [9]Lin Ping-Yen,Tsai Cheng-Ting,Hsu Chang Francis et al.The AutonomicImbalance of Myocardial Ischemia during Exercise Stress Testing:Insight fromShort-Term Heart Rate Variability Analysis·[J]·Int J Environ Res PublicHealth,2022,19:15096.

[0240]

[10] Kaufmann Sebastian,Gronwald Thomas,Herold Fabian et al.Heart RateVariability-Derived Thresholds for Exercise Intensity Prescription inEndurance Sports:A Systematic Review of Interrelations and Agreement withDifferent Ventilatory and Blood Lactate Thresholds.[J].Sports Med Open,2023,9:59.

[0241]

[11] Lan Chou-Chin,Yeh Kuan-Hung,Tzeng I-Shiang et al.Evaluation ofthe relationship of tricuspid regurgitation peak gradient / tricuspid annulusplane systolie excursion to exercise capacity,cardiac index,and ventilatoryfunction during exercise in patients with COPD.[J].Heart Lung,2023,62:22-27.

[0242]

[12] Hashem Mohamed D, Parker Ann M, Needham Dale M, Early Mobilization and Rehabilitation of Patients Who Are Critically Ill.[J]·Chest,2016,150:722-731.

[0243]

[13] Tsoukos Athanasios,Brown Lee E,Terzis Gerasimos et al.Changes inEMG and movement velocity during a set to failure against different loads in the bench press exercise.[J].Scand J Med Sci Sports,2021,31:2071-2082·

[0244]

[14] Xiao Jinzhuang, Sun Jinli, Wang Hongrui, et al. Analysis of differences in muscle fatigue during core stability training[J]. Journal of Biomedical Engineering, 2017, 34(2): 227-232.

[0245]

[15] Larsen Stian,Haugen Markus,van den Tillaar Roland,Comparison ofKinematics and Electromyographic Activity in the Last Repetition duringDifferent Repetition Maximums in the Bench Press Exercise·[J]·Int J EnvironRes Public Health,2022,19:14238.

[0246]

[16] Kang Fu-Jie,Ou Hsiang-Ling,Lin Kun-Ying et al.Serratus Anteriorand Upper Trapezius Electromyographic Analysis of the Push-Up Plus Exercise:ASystematic Review and Meta-Analysis.[J]·J Athl Train,2019,54:1156-1164.

[0247]

[17] Wang Le jun,Wang Yuting,Ma Aidi.et al.A Comparative Study ofEMGIn dices in Muscle Fatigue Evaluation Based on Grey Relational Analysisduring All-Out Cycling Exercise.[J].Biomed Res Int,2018,2018:9341215·

[0248]

[18] Rehman Asma,Ganai Jyoti,Aggarwal Rajeev et al.Effec t of PassiveStretching of Respiratory Muscles on Chest Expansion and 6-Minute WalkDistanee inCOPD Patients.[J]·Int J Environ Res Public Health,2020,17:6480.

[0249]

[19] Kho Michelle E,Reid Julie,Molloy Alexander J et al.Critical CareCling toImprove ower xtremity Strength(CYCLE):protocol for an international,mult icentre randomised clinical trial of early in-bed cycling formechanically ventilatedpatients,[J].BMJ Open,2023,13:e075685.

[0250]

[20] Heredia-Elvar Juan R,Juan-Recio Casto,Prat-Luri Amaya et al.

[0251] Observati.onal Screening Guidelines and Smartphone AccelerometerThresholds toEstablish the Intensity of Some of the Most Popular CoreStability Exercises·[J]·Front Physiol,2021,12:751569.

[0252]

[21] Barbado David,Irles-Vidal Belen,Prat-Luri Amaya et al.Trainingintensityquantification of core stability exercises based on a smartphoneaccelerometer.[J]·PLoS One,2018,13:e0208262。

Claims

1. A method for constructing an immediate adaptation model of rehabilitation based on SVI-sEMG dual-mode correction, characterized in that: The following steps are involved: 1) Start; 2) Data collection of patients in resting state; 3) Resting data preprocessing; 4) Patient movement status data collection; 5) Motion data preprocessing; 6) Exercise intensity prediction; 7) Adjust the intensity of exercise prediction; 8) Generate an intensity-adjusted exercise plan; 9) End.

2. The method for constructing an immediate adaptation mode of rehabilitation based on SVI-sEMG dual-mode correction according to claim 1, characterized in that: The preprocessing methods in step 3) and step 5) are extraction, noise filtering and standardization.

3. The method for constructing an immediate adaptation mode of rehabilitation based on SVI-sEMG dual-mode correction according to claim 2, characterized in that: The extraction method is to use short-time Fourier transform or wavelet transform to analyze the time-frequency characteristics of sEMG signals and extract the features of SVI signals; the denoising method is to use a bandpass filter to eliminate low-frequency baseline drift and high-frequency noise in sEMG signals and use a wavelet transform denoising method to remove artifacts and noise of SVI signals; the standardization method is to process sEMG and SVI signals using a standardized formula to eliminate differences between individuals, and the standardized formula is: Where μ and σ are the mean and standard deviation of the signal, respectively.

4. The method for constructing an immediate adaptation mode of rehabilitation based on SVI-sEMG dual-mode correction according to claim 1, characterized in that: The exercise intensity prediction in step 6) includes establishing an intensity assessment model, data fusion and feedback regulation.

5. The method for constructing an immediate adaptation mode of rehabilitation based on SVI-sEMG dual-mode correction according to claim 4, characterized in that: The strength assessment model established includes modal effective value and multimodal strength assessment, wherein the multimodal strength assessment uses a stacking method combined with linear regression and random forest to estimate the intensity of rehabilitation exercises.

6. The method for constructing an immediate adaptation mode of rehabilitation based on SVI-sEMG dual-mode correction according to claim 4, characterized in that: The data fusion uses Kalman filtering through real-time state estimation and recursive updating to verify and make a prediction plan for intensity adjustment of rehabilitation exercises based on the preliminary prediction results of multimodal intensity assessment.

7. The method for constructing an immediate adaptation mode of rehabilitation based on SVI-sEMG dual-mode correction according to claim 4, characterized in that: The feedback regulation adjusts the rehabilitation exercise intensity through fuzzy rules.

8. The method for constructing an immediate adaptation mode of rehabilitation based on SVI-sEMG dual-mode correction according to claim 1, characterized in that: The intensity adjustment exercise intensity program described in step 8) includes dynamic threshold determination, and dynamically adjusts the safety threshold based on historical data and current trends.

9. The method for constructing an immediate adaptation mode of rehabilitation based on SVI-sEMG dual-mode correction according to claim 8, characterized in that: The intensity adjustment exercise intensity program in step 8) includes the following implementation steps:

1. Start; 2. Motion parameter detection; 3. Prediction of its own strength; 4. Revision of self-strength prediction; 5. Adjust the intensity plan forecast; 6. Correct and adjust the intensity plan; 7. Adjust exercise prescription; 8. Safety monitoring; 9. End.

10. The method for constructing an immediate adaptation mode of rehabilitation based on SVI-sEMG dual-mode correction according to claim 8 or 9, characterized in that: Dynamic threshold of sEMG signal sEMG,dynamic (t), defined as: Threshold sEMG,dynamic (t)=μ sEMG (t)+k·σ sEMG (t) Dynamic threshold of SVI signal SVI,dynamic (t), defined as: Threshold SVI,dynamic (t)=μ SVI (t)+k·σ SVI (t) Where k is the threshold adjustment parameter; For the moving average μ of the sEMG signal sEMG (t), defined as: For the moving average μ of the SVI signal SVI (t), defined as: The standard deviation of the sEMG signal σ sEMG (t), defined as: For the standard deviation of the SVI signal σ SVI (t), defined as: Among them, X sEMG (i) and X SVI (i) are the sEMG signal value and SVI signal value at the i-th moment, respectively, and w is the calculation window.