Rehabilitation assessment method and system based on patient state perception

Abstract

The invention relates to the technical field of physiological status monitoring, in particular to a rehabilitation assessment method and system based on patient status awareness, and the method comprises the following steps: collecting myoelectricity, heart rate and skin conductance signal data of a patient, normalizing time sequence data, setting a time window, calculating a mean value and a mean square error, and analyzing a signal covariance variation trend. And generating a cross-domain signal stability list. According to the method, the physiological signals of the patient are collected and analyzed, the method has important significance on improvement of the rehabilitation effect, the physiological signals such as myoelectricity, heart rate and skin conductance are collected and analyzed in real time, tiny changes of the state of the patient can be carefully captured, more accurate state evaluation and rehabilitation progress monitoring are achieved, and the rehabilitation effect is improved. The accuracy of data processing and the efficiency of signal decoding are ensured by analyzing the signal stability and optimizing and adjusting the time window, and the rehabilitation scheme is further adjusted according to the specific response of the patient, so that the rehabilitation process is more personalized and targeted.

Description

Technical Field

[0001] The present invention relates to the technical field of physiological state monitoring, and in particular to a rehabilitation assessment method and system based on patient state perception. Background Art

[0002] The field of physiological status monitoring technology includes various methods and systems for real-time or non-real-time monitoring of various physiological signals of the human body, including heart rate, blood pressure, body temperature, respiratory rate and various physiological parameters. The core content is mainly to capture physiological signals through various sensors and devices, and convert physiological signals into useful data for health monitoring and decision support in the process of disease recovery. The systematic introduction of the technical field involves sensor technology, signal acquisition methods, data conversion technology, and data analysis and application. The common goal is to improve the accuracy and practicality of physiological monitoring so that it can be effectively applied in medical and health management.

[0003] Among them, the patient state-aware rehabilitation assessment method refers to the use of specific technical means to assess the patient's physiological and psychological state during the rehabilitation process in order to provide support for rehabilitation. This method includes the use of physiological state monitoring equipment to collect the patient's physiological data, such as motor ability, muscle activity, and related physiological parameters. The technical matters targeted by the patent subject include collecting data and how to effectively assess the patient's rehabilitation progress based on the data. Specific means include using wearable devices or fixed devices to monitor the patient's activity status and physiological reactions, and assessing the patient's rehabilitation needs and progress based on the data.

[0004] Existing technologies lack sufficient flexibility and personalization in physiological status monitoring and rehabilitation assessment, which is mainly reflected in the relatively fixed processing of patients' physiological signals and the lack of dynamic adjustment mechanisms for individual differences. This fixed processing method leads to misunderstandings or delayed responses to patients' status, especially in the early stages of rehabilitation and when rehabilitation status changes rapidly. Existing systems fail to fully utilize all available physiological data to optimize the rehabilitation process, such as failing to adjust rehabilitation plans in real time to adapt to patients' immediate needs. This deficiency leads to low rehabilitation efficiency, prolongs the rehabilitation cycle, and even has an adverse effect on patients' rehabilitation motivation and rehabilitation effects. Summary of the invention

[0005] In order to solve the problem that the existing technology lacks sufficient flexibility and personalization in physiological status monitoring and rehabilitation assessment, which is mainly reflected in the relatively fixed processing of patient physiological signals and the lack of dynamic adjustment mechanisms for individual differences. This fixed processing method leads to misunderstanding of the patient's status or delayed response, especially in the early stages of rehabilitation and when the rehabilitation status changes rapidly. The existing system fails to make full use of all available physiological data to optimize the rehabilitation process, such as failing to adjust the rehabilitation plan in real time to adapt to the patient's immediate needs. This deficiency leads to technical problems such as low rehabilitation efficiency, prolonged rehabilitation cycle, and even adverse effects on the patient's rehabilitation motivation and rehabilitation effect. The embodiment of the present invention provides a rehabilitation assessment method and system based on patient status perception. The technical solution is as follows:

[0006] In one aspect, a rehabilitation assessment method based on patient state perception is provided, the method comprising:

[0007] S1: Collect the patient's electromyography, heart rate and skin conductance signal data, normalize the time series data, set the time window, calculate the mean and mean square error, analyze the trend of signal covariance, and generate a cross-domain signal stability list;

[0008] S2: Through the cross-domain signal stability list, a time window with stability priority is selected as a decoding period, the fluctuation interval of the patient's status is detected, the data segment whose fluctuation amplitude exceeds the stability variation range is marked, adjacent stable signal points are called for interpolation correction, the window length is adjusted, and an optimized decoding time window is generated;

[0009] S3: Based on the optimized decoding time window, the angle change rate of the joint motion trajectory and the gradient of the neural signal emission cycle are calculated, the time series gradient change rate is extracted, the change trend of the electromyographic signal and the motion trajectory is compared, the patient's rehabilitation status is classified according to the growth and attenuation characteristics of the physiological signal, and the gradient change trend of the physiological signal is generated;

[0010] S4: using the physiological signal gradient change trend to identify changes in the patient's physiological signal, calculating the rehabilitation speed index in combination with the patient's rehabilitation status, analyzing the signal fluctuation range with reference to the speed change in the rehabilitation recovery stage, and generating a rehabilitation progress recognition result.

[0011] As a further solution of the present invention, the cross-domain signal stability list includes stability threshold, fluctuation trend, and covariance coefficient; the optimized decoding time window includes window start and end times, window dynamic interval, and interpolation correction point; the physiological signal gradient change trend includes gradient change rate, trend synchronization, and signal amplitude range; the rehabilitation progress identification result includes rehabilitation stage index, fluctuation amplitude level, and time window correction amount.

[0012] As a further solution of the present invention, the steps of obtaining the cross-domain signal stability list are specifically as follows:

[0013] S101: Collect the patient's electromyography, heart rate and skin conductance signal data, normalize the signal data, set a fixed time window, divide the time interval, and generate a signal normalization time series;

[0014] S102: calling the signal normalization time series, performing mean and mean square error calculations on the windows of the electromyographic signal, the heart rate signal, and the skin conductance signal, respectively, identifying fluctuations of the patient's physiological signals within the differentiated time window, analyzing fluctuation patterns between differentiated signals, and generating a signal fluctuation interval;

[0015] S103: Call the signal fluctuation interval, calculate the covariance of the signal mean and the mean square error, identify the trend of signal correlation change, analyze the stability between signals, and generate a cross-domain signal stability list.

[0016] As a further solution of the present invention, the step of obtaining the optimized decoding time window is specifically as follows:

[0017] S201: calling the cross-domain signal stability list, and screening the stability-prioritized time windows according to the fluctuation of the signal in the window to obtain a stability-prioritized window set;

[0018] S202: Based on the stability priority window set, detecting the fluctuation interval of the patient's state, identifying the fluctuation amplitude, and comparing the stability change range, marking the data segments beyond the range, screening adjacent stable signal points, performing interpolation correction, and performing continuity check on the corrected data to obtain a corrected signal window;

[0019] S203: Using the modified signal window, adjusting the window length according to the stability change, comparing the fluctuation of the patient's physiological signal before and after the adjustment, optimizing the decoding cycle, and generating an optimized decoding time window.

[0020] As a further solution of the present invention, the step of acquiring the gradient change trend of the physiological signal is specifically:

[0021] S301: calling the optimized decoding time window, calculating the change rate of the joint motion trajectory angle, analyzing the rate change in the time series, and obtaining the joint motion rate data;

[0022] The formula for calculating the change rate of the joint motion trajectory angle is:

[0023]

[0024] Among them, v o Represents the rate of change of the angle of the motion trajectory of joint o, θ o,j represents the angle value of joint o at the jth moment in the time series, θ o,j-1represents the angle value of joint o at the j-1th moment in the time series, F represents the total number of samples in the time window, and Δt represents the time interval between two adjacent moments;

[0025] S302: Based on the joint motion rate data, calculate the gradient of the neural signal emission cycle, analyze the gradient change in the time series, extract the gradient change rate, identify the change trend of the electromyographic signal and the motion trajectory, extract the differentiated trend characteristics, and obtain the time series gradient change rate;

[0026] S303: calling the time series gradient change rate, classifying the patient's rehabilitation status according to the growth and attenuation characteristics of the patient's physiological signal, identifying the gradient change pattern of the physiological signal, and generating the gradient change trend of the physiological signal.

[0027] As a further solution of the present invention, the steps of obtaining the rehabilitation progress recognition result are specifically as follows:

[0028] S401: using the physiological signal gradient change trend, comparing the signal change amounts in adjacent time windows, identifying the direction and amplitude of the signal change, summarizing the signal change pattern, calling the heart rate variability and blood oxygen saturation gradient to compare the numerical differences in adjacent time windows, screening the key change intervals, and generating the patient's physiological signal change rate;

[0029] S402: calling the change rate of the patient's physiological signal, combining the patient's rehabilitation status data, extracting the mean value of the physiological signal in the differentiated rehabilitation stage, identifying the speed change range of multiple stages, calculating the rehabilitation speed index, comparing the speed difference of the differentiated rehabilitation stage, summarizing the signal fluctuation range, and generating the rehabilitation speed change trend;

[0030] S403: calling the rehabilitation speed change trend, analyzing the signal fluctuation range in the differentiation stage, adjusting the time window length, optimizing the signal analysis interval, summarizing the rehabilitation progress, and generating a rehabilitation progress recognition result.

[0031] As a further solution of the present invention, the formula for calculating the recovery speed index is:

[0032]

[0033] Among them, S r Represents the recovery speed index, V max Represents the maximum rate of change of the patient's physiological signal, V min Represents the minimum rate of change of the patient's physiological signal, V avg Represents the average rate of change of the patient's physiological signal, |V max -V min |Represents the fluctuation range of the patient's physiological signal change rate.

[0034] As a further solution of the present invention, the method further comprises step S5:

[0035] S5: calling the rehabilitation progress identification result, adjusting the calculation weight with reference to the patient's physiological adaptability and exercise tolerance, matching the patient's rehabilitation status classification data, setting rehabilitation goals, generating a rehabilitation status evaluation value, and adjusting the rehabilitation training frequency and training load according to the rehabilitation status evaluation value;

[0036] The rehabilitation status assessment value includes rehabilitation goals, training frequency, and load adjustment parameters.

[0037] As a further solution of the present invention, the steps for obtaining the rehabilitation status evaluation value are specifically as follows:

[0038] S501: calling the rehabilitation progress recognition result, referring to the patient's physiological adaptability and exercise tolerance, calculating the influence of multiple indicators, setting calculation weights, and obtaining a rehabilitation parameter weight setting result;

[0039] S502: Based on the rehabilitation parameter weight setting result, matching the patient rehabilitation status classification data, screening the associated rehabilitation categories, combining the rehabilitation goals of the differentiated categories, identifying the rehabilitation progress of the differentiated categories, and obtaining the rehabilitation status evaluation value;

[0040] S503: According to the rehabilitation status evaluation value, the patient's rehabilitation training items are sorted out, the training frequency and training load are adjusted, and the rehabilitation training rhythm is optimized.

[0041] On the other hand, the rehabilitation assessment system based on patient state perception is used to perform the above-mentioned rehabilitation assessment method based on patient state perception, and the system includes:

[0042] The signal stability analysis module obtains the patient's electromyography, heart rate and skin conductance signal data, calls the time series data, calculates the signal mean and mean square error based on a fixed time window, screens the time window where the signal fluctuation amplitude exceeds the fixed interval, calculates the proportion of signal points that fluctuate beyond the interval, calls adjacent stable signal points for interpolation correction, and generates a cross-domain signal stability list;

[0043] The decoding cycle optimization module selects the time window with priority of signal stability based on the cross-domain signal stability list, calls the signal fluctuation amplitude data in the time window, calculates the interval ratio of the fluctuation amplitude exceeding the stable variation range, adjusts the time window length according to the ratio value, and generates an optimized decoding time window;

[0044] The motion state calculation module calls the optimized decoding time window, obtains the patient's joint motion trajectory, calculates the trajectory angle change rate, extracts the neural signal emission cycle, calculates the time series gradient change rate, compares the gradient change rate with the time series correlation of the electromyographic signal, and generates the physiological signal gradient change trend;

[0045] The rehabilitation progress assessment module uses the physiological signal gradient change trend to obtain the patient's electromyographic signal growth and attenuation characteristics, identifies the physiological signal fluctuation range of the differentiated rehabilitation stage, calculates the patient's state change speed, and generates a rehabilitation progress recognition result;

[0046] The rehabilitation status matching module uses the rehabilitation progress identification result, refers to the patient's physiological adaptability and exercise tolerance, matches the rehabilitation status classification data, adjusts the rehabilitation training frequency, and generates a rehabilitation status evaluation value.

[0047] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0048] Collecting and analyzing the patient's physiological signals is of great significance for improving the rehabilitation effect. By comprehensively processing time series data, adjusting the decoding cycle, and adjusting the frequency and load of rehabilitation training in real time, the dynamic adaptability and accuracy of rehabilitation monitoring can be significantly improved. Real-time collection and analysis of physiological signals such as electromyography, heart rate, and skin conductance enable the subtle changes in the patient's state to be captured in detail, achieving more accurate state assessment and rehabilitation progress monitoring. By analyzing the signal stability and optimizing the time window, the accuracy of data processing and the efficiency of signal decoding are ensured, and the rehabilitation plan is further adjusted according to the patient's specific response, making the rehabilitation process more personalized and goal-oriented. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a schematic diagram of the workflow of the present invention;

[0050] Figure 2 It is a system flow chart of the present invention. DETAILED DESCRIPTION

[0051] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0052] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "example" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of the word "example" is intended to present the concept in a specific way. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or it can be either of the two.

[0053] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0054] See also Figure 1 The embodiment of the present invention provides a rehabilitation assessment method based on patient state perception, and the processing flow of the method may include the following steps:

[0055] S1: Collect the patient's electromyography, heart rate and skin conductance signal data, normalize the time series data, set the time window, calculate the mean and mean square error, analyze the trend of signal covariance, and generate a cross-domain signal stability list;

[0056] S2: Through the cross-domain signal stability list, select the time window with stability priority as the decoding period, detect the fluctuation range of the patient's status, mark the data segment whose fluctuation amplitude exceeds the stability change range, call the adjacent stable signal points for interpolation correction, adjust the window length according to the stability change, and generate the optimized decoding time window;

[0057] S3: Based on the optimized decoding time window, the angle change rate of the joint motion trajectory and the gradient of the neural signal emission cycle are calculated, the time series gradient change rate is extracted, the change trend of the electromyographic signal and the motion trajectory is compared, the patient's rehabilitation status is classified according to the growth and attenuation characteristics of the physiological signal, and the gradient change trend of the physiological signal is generated;

[0058] S4: Use the physiological signal gradient change trend to identify the changes in the patient's physiological signals, calculate the rehabilitation speed index in combination with the patient's rehabilitation status, analyze the signal fluctuation range with reference to the speed changes in the differentiated rehabilitation recovery stage, correct the time window length, and generate the rehabilitation progress recognition result;

[0059] S5: call the rehabilitation progress recognition result, adjust the calculation weight according to the patient's physiological adaptability and exercise tolerance, match the patient's rehabilitation status classification data, set the rehabilitation goal, generate the rehabilitation status evaluation value, and adjust the rehabilitation training frequency and training load according to the rehabilitation status evaluation value;

[0060] The cross-domain signal stability list includes stability threshold, fluctuation trend, and covariance coefficient. The optimized decoding time window includes window start and end times, window dynamic interval, and interpolation correction point. The physiological signal gradient change trend includes gradient change rate, trend synchronization, and signal amplitude range. The rehabilitation progress identification results include rehabilitation stage index, fluctuation amplitude level, and time window correction amount. The rehabilitation status assessment value includes rehabilitation goals, training frequency, and load adjustment parameters.

[0061] The specific steps for obtaining the cross-domain signal stability list are as follows:

[0062] S101: Collect the patient's electromyography, heart rate and skin conductance signal data, normalize the signal data, set a fixed time window, divide the time interval, and generate a signal normalization time series;

[0063] The collection of patients' electromyography, heart rate and skin conductance signal data needs to be completed through wearable devices or medical instruments. Surface electromyography sensors (sEMG) are used to obtain muscle activity data, photoelectric volume pulse wave sensors (PPG) are used to obtain heart rate information, and skin conductance sensors (EDA / GSR) are used to detect skin electrical response. The device transmits the data to the data acquisition device via Bluetooth or wired mode. The received data contains timestamps, original electrical signal values ​​and device calibration parameters. To ensure the consistency of data processing, all signals need to be normalized. The normalization method can be Z-score normalization, that is, for each signal value X i , calculate the normalized value X′ i The formula is as follows:

[0064]

[0065] Among them, μ is the mean of the signal samples, and σ is the standard deviation;

[0066] The original data of the electromyographic signal is set in the range of 0-5000μV. The mean value of a set of collected data is 1500μV and the standard deviation is 400μV. When the original value of a signal point is 2000μV, its normalized value is:

[0067]

[0068] After all signal data are normalized, it is necessary to set a fixed time window for data segmentation. The size of the time window is set according to application requirements. Emotion recognition tasks can use a 1-5 second window, while medical tasks use a 10-30 second window. If the time window is set to 5 seconds, then 60 seconds of signal data can be divided into 12 windows in total. The signal data in each window is classified as a time series. The signal sequence in the window can use a sliding window strategy, sliding for 1 second each time (overlapping sliding) to form multiple normalized time series. Each time series is used for subsequent signal fluctuation analysis and pattern recognition to generate a signal normalized time series.

[0069] S102: calling the signal normalization time series, performing mean and mean square error calculations on the windows of the electromyographic signal, the heart rate signal, and the skin conductance signal, respectively, identifying the fluctuation of the patient's physiological signal within the differentiated time window, analyzing the fluctuation pattern between the differentiated signals, and generating a signal fluctuation interval;

[0070] Obtain normalized data of electromyographic signal (EMG), heart rate signal (HR) and skin conductance signal (EDA), set the time window length W to perform mean calculation and mean square error (MSE) calculation, and calculate the mean in each window And the mean square error:

[0071]

[0072] Among them, S i is the signal value in the window, and n is the number of data points in the window;

[0073] Set the EMG data in a time window [0.2, 0.25, 0.3, 0.35, 0.4];

[0074] Then the mean value is calculated:

[0075]

[0076] The MSE is calculated as:

[0077]

[0078] Traverse all windows to calculate the complete mean and mean square error sequence, analyze the fluctuation of the signal in different time windows, extract the local maximum and minimum values ​​of the signal, and calculate the fluctuation amplitude:

[0079] A=S max -S min ;

[0080] The EMG signal fluctuation threshold of 0.05 was set as the significant fluctuation standard to identify the fluctuation patterns of different signal types and generate the signal fluctuation interval.

[0081] S103: calling the signal fluctuation interval, calculating the covariance of the signal mean and the mean square error, identifying the trend of signal correlation change, analyzing the stability between signals, and generating a cross-domain signal stability list;

[0082] Calculate the covariance of the signal mean and mean square error to obtain the mean sequence of the electromyographic signal, heart rate signal and skin conductance signal and mean square error sequence MSE 1 , MSE 2 , MSE 3 ;

[0083] Compute the covariance between signals:

[0084]

[0085] Among them, X and Y represent the mean or mean square error of two different signals;

[0086] Set the mean values ​​of EMG and HR in a time window to [0.2, 0.3, 0.35, 0.4] and [80, 85, 90, 95] respectively, and calculate the mean Then the covariance is:

[0087]

[0088] Traverse all signal combinations to calculate the covariance matrix, analyze the trend of signal correlation changes, identify signal stability, and generate a cross-domain signal stability list

[0089] The specific steps for obtaining the optimized decoding time window are:

[0090] S201: calling the cross-domain signal stability list, screening the time window with stability priority according to the fluctuation of the signal in the window, and obtaining the stability priority window set;

[0091] The cross-domain signal stability list stores the signal fluctuations of multiple time windows. The signal of each window shows different change characteristics on the time axis, including stable intervals and fluctuation intervals. After calling the list, the signal data of each window is read one by one, and the fluctuation amplitude, mean and variance of the signal are calculated for each window. As the core indicator of stability evaluation, a stability priority screening standard is set to compare the fluctuation amplitudes of all windows, exclude windows with excessive fluctuations, and only retain windows with smaller signal changes. In this process, real-time data will be referenced to calculate the overall signal fluctuations, and a reasonable stability threshold will be set accordingly. When monitoring the patient's heart rate signal, if the heart rate in a window fluctuates slightly, such as between 60-65, and the heart rate in another window fluctuates greatly between 60-90, the latter is affected by external interference or sudden changes in the patient's status. Therefore, the former is more in line with the stability priority screening standard. During the screening process, the signal continuity in the window must also be checked to avoid windows that change frequently in a short period of time from being mistakenly included in the stability priority window set.

[0092] S202: Based on the stability priority window set, detect the fluctuation range of the patient's state, identify the fluctuation amplitude, and compare the stability change range, mark the data segment that exceeds the range, screen adjacent stable signal points, perform interpolation correction, and perform continuity check on the corrected data to obtain a corrected signal window;

[0093] The fluctuation range of the patient's state is identified, and the fluctuation amplitude of the signal in each time window is analyzed. If the fluctuation amplitude of a certain window exceeds the set stability change range, the data segment is marked as abnormal data. When continuously monitoring blood oxygen saturation, if the blood oxygen saturation in a certain time window suddenly drops from 97% to 85%, and then quickly recovers to 96%, the window is affected by signal interference or transient state changes of the patient and needs further processing. After detecting such abnormal data segments, the adjacent stable signal points before and after the abnormal segment will be found, and data correction will be performed based on the stable points. If the blood oxygen values ​​of the two adjacent stable points are 96% and 97% respectively, the interpolation method can be used to fill the middle abnormal interval to restore the continuity of the signal. After the correction is completed, the continuity of the corrected signal is checked to ensure that the corrected signal will not have sudden changes or unreasonable jumps. If the corrected blood oxygen data still has an abnormal value of 85%, it indicates that there is a problem with the interpolation method and it is necessary to reselect a suitable correction method. After a series of processing, the corrected signal window is obtained.

[0094] S203: using a modified signal window, adjusting the window length according to the stability change, comparing the fluctuation of the patient's physiological signal before and after the adjustment, optimizing the decoding cycle, and generating an optimized decoding time window;

[0095] On the basis of correcting the signal window, the window length is adjusted according to the change of signal stability. By calculating the signal fluctuation in the window, it is determined whether the window needs to be expanded or shortened. If the internal data fluctuation of the corrected signal window is still large, the window needs to be shortened to reduce the interference of fluctuations. If the signal is stable, the window can be appropriately extended to improve the signal coverage. In the process, the signal fluctuations before and after the adjustment will be compared. If the heart rate fluctuations in the original window are large, and the fluctuation range is reduced after adjusting the window length, it means that the adjustment is effective. The decoding cycle will be optimized according to the fluctuation situation to make data processing more accurate and generate an optimized decoding time window.

[0096] The specific steps for obtaining the gradient change trend of physiological signals are as follows:

[0097] S301: Calling the optimized decoding time window, calculating the change rate of the joint motion trajectory angle, analyzing the rate change in the time series, and obtaining the joint motion rate data;

[0098] The formula for calculating the rate of change of the joint motion trajectory angle is:

[0099]

[0100] Among them, v o Represents the rate of change of the angle of the motion trajectory of joint o, θ o,j represents the angle value of joint o at the jth moment in the time series, θ o,j-1represents the angle value of joint o at the j-1th moment in the time series, F represents the total number of samples in the time window, and Δt represents the time interval between two adjacent moments;

[0101] Parameter meaning and calculation process:

[0102] Joint angle measurement (θ o,j ,θ o,j-1 ) Acquisition method: Use an inertial measurement unit (IMU) sensor to monitor the angle change of joint o in real time. The IMU sensor detects the angular velocity and integrates it to calculate the angle value at each time point;

[0103] Numerical example: Assume that the angle measurements at time points j-1 and j are: θ o,j-1 =30°,θ o,j =35°;

[0104] Time interval (Δt) acquisition method: It is determined by the sampling frequency of the IMU sensor and represents the time interval between adjacent sampling points;

[0105] Numerical example: If the IMU sampling frequency is 100 Hz, the time interval is:

[0106]

[0107] The number of samples (F) is obtained by the set time window length and IMU sampling frequency, indicating the total number of sampling points in the time window;

[0108] Numerical example: If the time window is set to 1 second and the IMU sampling frequency is 100 Hz, then:

[0109] F = 1 × 100 = 100;

[0110] Angle change (|θ o,j -θ o,j-1 |) Calculation method: Calculate the absolute value of the angle change at adjacent time points;

[0111] Numerical example:

[0112] |θ o,j -θ o,j-1 |=|35-30|=5;

[0113] Total angle change Calculation method: sum the angle changes of all adjacent time points in the time window;

[0114] Numerical example: Assume that the angle change at each adjacent time point in the time window is 5, then the sum is:

[0115]

[0116] The rate of change of the angle of the joint motion trajectory (v o ) Calculation method: Divide the total angle change by the product of the time interval and the number of samples;

[0117] Numerical example:

[0118]

[0119] Meaning of the results: The average angle change rate of joint o within the selected time window was 495° / s, which can be used to analyze joint movement trends and stability, such as determining the smoothness of movement or detecting abnormal movement patterns.

[0120] S302: Based on the joint motion rate data, the gradient of the neural signal emission cycle is calculated, the gradient change in the time series is analyzed, the gradient change rate is extracted, the change trend of the electromyographic signal and the motion trajectory is identified, the differential trend characteristics are extracted, and the time series gradient change rate is obtained;

[0121] Calculate the gradient of the neural signal emission cycle, obtain neural signal data, perform data preprocessing, such as denoising, smoothing and normalization, and perform finite difference calculation of the time series. The gradient calculation formula is:

[0122]

[0123] Among them, S(t) is the normalized neural signal value, Δt is the time interval;

[0124] Set the time interval to 0.1s, and the signal value changes from 0.8 to 1.2 at a certain moment, then calculate:

[0125]

[0126] Traverse the entire time series to obtain gradient data and calculate the gradient change rate, that is, the first-order difference of the gradient series:

[0127] ΔG = G(t+1) - G(t);

[0128] If the gradient value is 4 at a certain moment and 4 at the next moment, then:

[0129] ΔG = 1;

[0130] Indicates that the gradient increases. Analyze the correlation between the gradient change rate and the motion rate data, and calculate the correlation coefficient. If the correlation coefficient is high, it indicates that there is a synchronous change trend between the EMG signal and the motion trajectory. Further extract the differentiated trend characteristics and obtain the time series gradient change rate.

[0131] S303: calling the time series gradient change rate, classifying the patient's rehabilitation status according to the growth and attenuation characteristics of the patient's physiological signal, identifying the gradient change pattern of the physiological signal, and generating the gradient change trend of the physiological signal;

[0132] The time series gradient change rate is called to obtain the patient's physiological signal time series data, including heart rate, electromyographic signal and skin conductance signal, etc. The growth and attenuation characteristics are analyzed according to the signal changes in different time windows, and different time periods are set to compare the rising or falling trends of the signals in each time period. The change pattern is identified by calculating the time series gradient change rate, and the data is classified according to the signal change rate. Faster growing signals are classified as high growth mode, slower growing signals are classified as stable mode, and signals showing a downward trend are classified as attenuation mode. The gradient change characteristics of multiple physiological signals are combined to analyze the performance of signals in different rehabilitation stages, and different rehabilitation states are further subdivided. Signals that remain stable for a long time are classified as recovery period, and signals with large changes are classified as early rehabilitation state. The physiological signal gradient change trend is generated according to the signal trend.

[0133] The specific steps for obtaining the rehabilitation progress identification results are as follows:

[0134] S401: using the change trend of physiological signal gradient, comparing the signal change amount in adjacent time windows, identifying the direction and amplitude of signal change, summarizing the signal change pattern, calling the heart rate variability and blood oxygen saturation gradient to compare the numerical difference of adjacent time windows, screening the key change interval, and generating the change rate of the patient's physiological signal;

[0135] Extract the patient's heart rate, blood oxygen saturation and other physiological signal data, and organize them in chronological order. The signal change trend of each time window needs to be evaluated by calculating the changes in adjacent time points. In actual operation, the signals in adjacent time windows are compared through data processing. The signal changes in adjacent time windows are first calculated to determine the direction of change, such as whether the heart rate rises or falls, whether the blood oxygen saturation increases or decreases, and the amplitude of the signal change is compared. The value is compared with the set threshold, and the time period with significant changes is screened out. Further summarize the signal change pattern. In the process of screening the key change interval, certain judgment criteria need to be adopted. When the heart rate change exceeds a specific threshold and the blood oxygen saturation When the change in the degree also exceeds the set range, this time period can be identified as a key change interval of the physiological signal. This process can use data statistical tools to perform trend analysis on long-term monitoring data, compare the frequency of signal changes in different time periods, the amplitude of fluctuations and the differences between adjacent time points, and identify the intervals with large fluctuations in the patient's physiological state. It is necessary to combine all the screened key change intervals and calculate the rate of change of the physiological signal in the interval. The specific method is to analyze the signal change rate at different stages according to the degree of change and time span of the physiological signal, and store it for further analysis and research. Through this process, the fluctuation of the patient's physiological signal in different time periods can be obtained, forming physiological signal change rate data.

[0136] S402: calling the change rate of the patient's physiological signal, combining the patient's rehabilitation status data, extracting the mean value of the physiological signal in the differentiated rehabilitation stage, identifying the speed change range of multiple stages, calculating the rehabilitation speed index, comparing the speed difference of the differentiated rehabilitation stage, summarizing the signal fluctuation range, and generating the rehabilitation speed change trend;

[0137] The formula for calculating the recovery speed index is:

[0138]

[0139] Among them, S r Represents the recovery speed index, V max Represents the maximum rate of change of the patient's physiological signal, V min Represents the minimum rate of change of the patient's physiological signal, V avg Represents the average rate of change of the patient's physiological signal, |V max -V min |Represents the fluctuation range of the patient's physiological signal change rate;

[0140] Parameter meaning and calculation process:

[0141] Maximum heart rate change rate V max : Find the maximum value of the heart rate change rate within the selected time window;

[0142] Minimum heart rate change rate V min : Find the minimum value of the heart rate change rate within the same time window;

[0143] Average heart rate change rate V avg : Calculate the average value of the heart rate change rate in the time window;

[0144] Through monitoring and calculation, the following data are obtained:

[0145] Maximum heart rate change rate V max : 100 beats / minute, minimum heart rate change rate V min : 60 times / minute, average heart rate change rate V avg : 80 times / minute;

[0146] The calculation process is as follows:

[0147] Fluctuation range:

[0148] |V max -V min |=|100-60|=40;

[0149] Normalized volatility:

[0150]

[0151] Mean square error calculation:

[0152]

[0153] Calculate the recovery speed index S r :

[0154] S r =0.5×20=10;

[0155] The results show that the recovery speed index is 10, which indicates the degree of fluctuation of the patient's heart rate change rate within the selected time window. The larger the index value, the more drastic the heart rate change; the smaller the value, the more stable the heart rate change. r The changing trend of the blood pressure can be used to analyze the patient's recovery process.

[0156] S403: calling the rehabilitation speed change trend, analyzing the signal fluctuation range in the differentiation stage, adjusting the time window length, optimizing the signal analysis interval, summarizing the rehabilitation progress, and generating the rehabilitation progress recognition result;

[0157] Analyze the signal fluctuation range at different stages, adjust the time window length to optimize the signal analysis interval, obtain the signal fluctuation range data at each stage, and calculate the signal standard deviation:

[0158]

[0159] To measure the degree of signal fluctuation, for example, the signal data of a certain stage is [1.0, 1.3, 1.5, 1.2, 1.4], and the mean is 1.28, then the standard deviation is calculated as:

[0160]

[0161] Adjust the time window length W according to the signal fluctuations, and set the adjustment formula:

[0162]

[0163] Among them, W 0 is the initial window length, σ ref is the reference standard deviation;

[0164] If the initial window is 0.5s and the reference standard deviation is 0.2, the adjusted window length is:

[0165]

[0166] After optimizing the time window, the signal is summarized, the rehabilitation progress is summarized based on the calculated rehabilitation speed trend data, and the rehabilitation progress recognition result is output.

[0167] The specific steps for obtaining the rehabilitation status assessment value are as follows:

[0168] S501: calling the rehabilitation progress recognition result, referring to the patient's physiological adaptability and exercise tolerance, calculating the influence of multiple indicators, setting calculation weights, and obtaining the rehabilitation parameter weight setting result;

[0169] Obtain the patient's physiological adaptation and exercise tolerance data, including heart rate recovery index (HRR), maximum oxygen uptake (VO 2 max), blood lactate clearance rate (BLC), maximum endurance time (TTE) and exercise recovery time (RT). The indicators reflect the patient's physical adaptability and recovery ability during the rehabilitation process. For each indicator, normalization processing is used to compare the data on the same scale. At the same time, its impact on the rehabilitation progress is analyzed, the changing trend of each indicator is identified, and the key influencing factors are determined according to the fluctuation. The data of multiple patients are compared, the variation range of each indicator is statistically analyzed, and a weight allocation strategy is established in combination with real-time data. The weight setting is adjusted according to the degree of influence of different indicators. If the patient's HRR changes rapidly, the weight of the influence of this indicator on the rehabilitation progress increases, otherwise it decreases. Combined with expert experience and clinical data, the weight is dynamically adjusted to ensure that the importance of different indicators in different rehabilitation stages is reasonably distributed, and the rehabilitation parameter weight setting results are obtained.

[0170] S502: Based on the rehabilitation parameter weight setting result, match the patient rehabilitation status classification data, screen the associated rehabilitation categories, combine the rehabilitation goals of the differentiated categories, identify the rehabilitation progress of the differentiated categories, and obtain the rehabilitation status evaluation value;

[0171] Match the patient's rehabilitation status classification data, establish a rehabilitation status classification system, divide the rehabilitation process into different stages, the initial recovery stage, the functional improvement stage and the consolidation stage, screen the rehabilitation category that matches the patient's physiological signal data and exercise tolerance index, calculate the deviation between the patient's current status and the standard category, analyze the mean and fluctuation range of key rehabilitation indicators of different categories, if the patient's rehabilitation data is close to the mean of a certain category, it is judged to belong to that category, otherwise the gap between it and different categories is calculated and classified into the closest category, combined with the rehabilitation goals of different rehabilitation categories, analyze the patient's rehabilitation progress, set a patient's maximum oxygen uptake to reach 70% of the target value in the recovery stage, and shorten the exercise recovery time to 80% of the baseline value, it indicates that the rehabilitation progress is good, otherwise if some key indicators still deviate greatly, it means that the rehabilitation process is slow, and calculate the patient's rehabilitation status assessment value based on different indicators.

[0172] S503: sorting out the patient's rehabilitation training items based on the rehabilitation status evaluation value, adjusting the training frequency and training load, and optimizing the rehabilitation training rhythm;

[0173] Organize the patient's rehabilitation training items, analyze the changing trends of the evaluation values, adjust the training frequency and training load. If the evaluation values ​​indicate that the rehabilitation progress is fast, gradually increase the training intensity, extend the training time, and reduce the rest interval. If the evaluation values ​​indicate that the rehabilitation progress is slow or stagnant, reduce the training intensity, increase the rest time, and adjust the training plan in combination with the patient's exercise tolerance. Set a patient's maximum endurance time to increase by 10% and the heart rate recovery index to increase by 5% within a week. Then, the training load can be appropriately increased by 5%-10%. On the contrary, if the endurance time changes by less than 2% and the heart rate recovery index decreases, it is necessary to reduce the training frequency and increase recovery training. It is also necessary to monitor the patient's exercise fatigue. If the heart rate recovers slowly after training and the blood lactate clearance rate decreases, it is necessary to reduce high-intensity training and increase low-intensity recovery training. Optimize the training rhythm according to the rehabilitation status so that the training plan meets the patient's individualized rehabilitation needs.

[0174] See also Figure 2 , a rehabilitation assessment system based on patient status perception, the system includes:

[0175] The signal stability analysis module obtains the patient's electromyography, heart rate and skin conductance signal data, calls the time series data, calculates the signal mean and mean square error based on a fixed time window, screens the time window where the signal fluctuation amplitude exceeds the fixed interval, calculates the proportion of signal points that fluctuate beyond the interval, calls adjacent stable signal points for interpolation correction, and generates a cross-domain signal stability list;

[0176] The decoding cycle optimization module selects the time window with signal stability priority based on the cross-domain signal stability list, calls the signal fluctuation amplitude data in the time window, calculates the interval ratio of the fluctuation amplitude exceeding the stable change range, adjusts the time window length according to the ratio value, and generates an optimized decoding time window;

[0177] The motion state calculation module calls the optimized decoding time window, obtains the patient's joint motion trajectory, calculates the trajectory angle change rate, extracts the neural signal emission cycle, calculates the time series gradient change rate, compares the gradient change rate with the time series correlation of the electromyographic signal, and generates the gradient change trend of the physiological signal;

[0178] The rehabilitation progress assessment module uses the gradient change trend of physiological signals to obtain the growth and attenuation characteristics of the patient's electromyographic signals, identify the fluctuation range of physiological signals in the differentiated rehabilitation stages, calculate the speed of patient status changes, and generate rehabilitation progress recognition results;

[0179] The rehabilitation status matching module uses the rehabilitation progress identification results, refers to the patient's physiological adaptability and exercise tolerance, matches the rehabilitation status classification data, adjusts the rehabilitation training frequency, and generates a rehabilitation status assessment value.

[0180] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A rehabilitation assessment method based on patient status perception, characterized in that: The following steps are involved: S1: Collect the patient's electromyography, heart rate and skin conductance signal data, normalize the time series data, set the time window, calculate the mean and mean square error, analyze the trend of signal covariance, and generate a cross-domain signal stability list; S2: Through the cross-domain signal stability list, a time window with stability priority is selected as a decoding period, the fluctuation interval of the patient's status is detected, the data segment whose fluctuation amplitude exceeds the stability variation range is marked, adjacent stable signal points are called for interpolation correction, the window length is adjusted, and an optimized decoding time window is generated; S3: Based on the optimized decoding time window, the angle change rate of the joint motion trajectory and the gradient of the neural signal emission cycle are calculated, the time series gradient change rate is extracted, the change trend of the electromyographic signal and the motion trajectory is compared, the patient's rehabilitation status is classified according to the growth and attenuation characteristics of the physiological signal, and the gradient change trend of the physiological signal is generated; S4: using the physiological signal gradient change trend to identify changes in the patient's physiological signal, calculating the rehabilitation speed index in combination with the patient's rehabilitation status, analyzing the signal fluctuation range with reference to the speed change in the rehabilitation recovery stage, and generating a rehabilitation progress recognition result.

2. The rehabilitation assessment method based on patient state perception according to claim 1, characterized in that: The cross-domain signal stability list includes stability threshold, fluctuation trend, and covariance coefficient; the optimized decoding time window includes window start and end times, window dynamic interval, and interpolation correction point; the physiological signal gradient change trend includes gradient change rate, trend synchronization, and signal amplitude range; the rehabilitation progress identification result includes rehabilitation stage index, fluctuation amplitude level, and time window correction amount.

3. The rehabilitation assessment method based on patient state perception according to claim 1, characterized in that: The steps for obtaining the cross-domain signal stability list are specifically as follows: S101: Collect the patient's electromyography, heart rate and skin conductance signal data, normalize the signal data, set a fixed time window, divide the time interval, and generate a signal normalization time series; S102: calling the signal normalization time series, performing mean and mean square error calculations on the windows of the electromyographic signal, the heart rate signal, and the skin conductance signal, respectively, identifying fluctuations of the patient's physiological signals within the differentiated time window, analyzing fluctuation patterns between differentiated signals, and generating a signal fluctuation interval; S103: Call the signal fluctuation interval, calculate the covariance of the signal mean and the mean square error, identify the trend of signal correlation change, analyze the stability between signals, and generate a cross-domain signal stability list.

4. The rehabilitation assessment method based on patient state perception according to claim 1, characterized in that: The steps for obtaining the optimized decoding time window are specifically as follows: S201: calling the cross-domain signal stability list, and screening the stability-prioritized time windows according to the fluctuation of the signal in the window to obtain a stability-prioritized window set; S202: Based on the stability priority window set, detecting the fluctuation interval of the patient's state, identifying the fluctuation amplitude, and comparing the stability change range, marking the data segments beyond the range, screening adjacent stable signal points, performing interpolation correction, and performing continuity check on the corrected data to obtain a corrected signal window; S203: Using the modified signal window, adjusting the window length according to the stability change, comparing the fluctuation of the patient's physiological signal before and after the adjustment, optimizing the decoding cycle, and generating an optimized decoding time window.

5. The rehabilitation assessment method based on patient state perception according to claim 1, characterized in that: The steps for obtaining the gradient change trend of the physiological signal are specifically as follows: S301: calling the optimized decoding time window, calculating the change rate of the joint motion trajectory angle, analyzing the rate change in the time series, and obtaining the joint motion rate data; The formula for calculating the change rate of the joint motion trajectory angle is: Among them, v o Represents the rate of change of the angle of the motion trajectory of joint o, θ o,j represents the angle value of joint o at the jth moment in the time series, θ o,j-1 represents the angle value of joint o at the j-1th moment in the time series, F represents the total number of samples in the time window, and Δt represents the time interval between two adjacent moments; S302: Based on the joint motion rate data, calculate the gradient of the neural signal emission cycle, analyze the gradient change in the time series, extract the gradient change rate, identify the change trend of the electromyographic signal and the motion trajectory, extract the differentiated trend characteristics, and obtain the time series gradient change rate; S303: calling the time series gradient change rate, classifying the patient's rehabilitation status according to the growth and attenuation characteristics of the patient's physiological signal, identifying the gradient change pattern of the physiological signal, and generating the gradient change trend of the physiological signal.

6. The rehabilitation assessment method based on patient state perception according to claim 1, characterized in that: The steps for obtaining the rehabilitation progress recognition result are specifically as follows: S401: using the physiological signal gradient change trend, comparing the signal change amounts in adjacent time windows, identifying the direction and amplitude of the signal change, summarizing the signal change pattern, calling the heart rate variability and blood oxygen saturation gradient to compare the numerical differences in adjacent time windows, screening the key change intervals, and generating the patient's physiological signal change rate; S402: calling the change rate of the patient's physiological signal, combining the patient's rehabilitation status data, extracting the mean value of the physiological signal in the differentiated rehabilitation stage, identifying the speed change range of multiple stages, calculating the rehabilitation speed index, comparing the speed difference of the differentiated rehabilitation stage, summarizing the signal fluctuation range, and generating the rehabilitation speed change trend; S403: calling the rehabilitation speed change trend, analyzing the signal fluctuation range in the differentiation stage, adjusting the time window length, optimizing the signal analysis interval, summarizing the rehabilitation progress, and generating a rehabilitation progress recognition result.

7. The rehabilitation assessment method based on patient state perception according to claim 6, characterized in that: The formula for calculating the recovery speed index is: Among them, S r Represents the recovery speed index, V max Represents the maximum rate of change of the patient's physiological signal, V min Represents the minimum rate of change of the patient's physiological signal, V avg Represents the average rate of change of the patient's physiological signal, |V max -V min |Represents the fluctuation range of the patient's physiological signal change rate.

8. The rehabilitation assessment method based on patient state perception according to claim 1, characterized in that: The method further comprises step S5: S5: calling the rehabilitation progress identification result, adjusting the calculation weight with reference to the patient's physiological adaptability and exercise tolerance, matching the patient's rehabilitation status classification data, setting rehabilitation goals, generating a rehabilitation status evaluation value, and adjusting the rehabilitation training frequency and training load according to the rehabilitation status evaluation value; The rehabilitation status assessment value includes rehabilitation goals, training frequency, and load adjustment parameters.

9. The rehabilitation assessment method based on patient state perception according to claim 8, characterized in that: The steps for obtaining the rehabilitation status assessment value are specifically as follows: S501: calling the rehabilitation progress recognition result, referring to the patient's physiological adaptability and exercise tolerance, calculating the influence of multiple indicators, setting calculation weights, and obtaining a rehabilitation parameter weight setting result; S502: Based on the rehabilitation parameter weight setting result, matching the patient rehabilitation status classification data, screening the associated rehabilitation categories, combining the rehabilitation goals of the differentiated categories, identifying the rehabilitation progress of the differentiated categories, and obtaining the rehabilitation status evaluation value; S503: According to the rehabilitation status evaluation value, the patient's rehabilitation training items are sorted out, the training frequency and training load are adjusted, and the rehabilitation training rhythm is optimized.

10. A rehabilitation assessment system based on patient status perception, characterized in that: According to any one of claims 1 to 9, the rehabilitation assessment method based on patient state perception comprises: The signal stability analysis module obtains the patient's electromyography, heart rate and skin conductance signal data, calls the time series data, calculates the signal mean and mean square error based on a fixed time window, screens the time window where the signal fluctuation amplitude exceeds the fixed interval, calculates the proportion of signal points that fluctuate beyond the interval, calls adjacent stable signal points for interpolation correction, and generates a cross-domain signal stability list; The decoding cycle optimization module selects the time window with priority of signal stability based on the cross-domain signal stability list, calls the signal fluctuation amplitude data in the time window, calculates the interval ratio of the fluctuation amplitude exceeding the stable variation range, adjusts the time window length according to the ratio value, and generates an optimized decoding time window; The motion state calculation module calls the optimized decoding time window, obtains the patient's joint motion trajectory, calculates the trajectory angle change rate, extracts the neural signal emission cycle, calculates the time series gradient change rate, compares the gradient change rate with the time series correlation of the electromyographic signal, and generates the physiological signal gradient change trend; The rehabilitation progress assessment module uses the physiological signal gradient change trend to obtain the patient's electromyographic signal growth and attenuation characteristics, identifies the physiological signal fluctuation range of the differentiated rehabilitation stage, calculates the patient's state change speed, and generates a rehabilitation progress recognition result; The rehabilitation status matching module uses the rehabilitation progress identification result, refers to the patient's physiological adaptability and exercise tolerance, matches the rehabilitation status classification data, adjusts the rehabilitation training frequency, and generates a rehabilitation status evaluation value.

Images