Fine motor rehabilitation regulation method and system based on multi-modal signal fusion

Abstract

The present application belongs to the technical field of rehabilitation engineering and neural engineering, and relates to a fine motor rehabilitation evaluation and regulation method and system based on multi-modal signal fusion, each evaluation comprising: collecting pressure distribution signals, surface electromyography signals and electroencephalography signals of hand fine motor; respectively pre-processing, feature extraction and normalization processing the three signals to obtain pressure feature values, surface electromyography feature values and electroencephalography feature values; calculating a conversion coefficient of the current rehabilitation stage based on the number of days of impaired fine motor function of the patient, and calculating the weights of the pressure feature values, surface electromyography feature values and electroencephalography feature values based on the conversion coefficient; weighted summing the pressure feature values, surface electromyography feature values and electroencephalography feature values to obtain an evaluation score; and regulating rehabilitation training parameters based on the evaluation score and the number of days of impaired fine motor function of the patient. The present application can break through the limitations of traditional single-modal physiological signal evaluation, and can improve the comprehensiveness, accuracy and personalized adaptation ability of rehabilitation evaluation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of rehabilitation engineering and neural engineering, and particularly relates to a fine motor rehabilitation regulation method and system based on multi-modal signal fusion. BACKGROUND

[0002] Stroke is a neurological impairment disease caused by sudden rupture or blockage of blood vessels in the brain, resulting in insufficient blood supply to brain tissue. Among them, limb spasticity is a common complication after stroke, with a high incidence of 65%. Persistent spasm can induce pain in the affected limb, and gradually develop into muscle spastic atrophy, which seriously affects the quality of life of patients. Current stroke rehabilitation treatment follows a closed-loop process of "function assessment, plan development, plan implementation and effect re-evaluation", in which accurate motor function assessment and closed-loop neural regulation are the core to ensure the effectiveness of rehabilitation. However, the current clinical rehabilitation assessment of hand fine motor function mainly relies on scales, lacking objective and quantitative evaluation indicators. At the same time, traditional rehabilitation training mainly uses fixed parameter electrical stimulation or physical training methods, which are difficult to dynamically adjust according to the rehabilitation progress of patients, resulting in limited rehabilitation effect.

[0003] The current method for evaluating the hand function of stroke patients mainly includes clinical scale evaluation based on Fugl-Meyer upper limb assessment scale (FMA-UE), nine-hole peg test 9HPT and modified Ashworth spasm assessment scale, machine vision-based limb and motor function evaluation, and intelligent evaluation based on physiological signals. However, these evaluation methods still have some technical limitations that need to be solved when applied to the rehabilitation evaluation of hand fine motor action:

[0004] (1) Scale evaluation: first, scale evaluation is time-consuming, has poor consistency of evaluation results, and is highly dependent on the clinical experience of rehabilitation therapists; second, scale evaluation is not sensitive enough to the changes in hand fine motor function in the short-term rehabilitation stage, making it difficult to accurately reflect the progressive rehabilitation progress of patients; moreover, scale evaluation has subjective factors, affecting the accuracy and reliability of the evaluation, and cannot meet the clinical needs of quantitative evaluation of fine motor function;

[0005] (2) Machine vision-based limb and motor function evaluation: patients with hand dysfunction often have high muscle tone, making the amplitude of hand fine motor action very small or even difficult to complete fine motor action, so there are great limitations in capturing fine motor action with a camera; at the same time, there are often occlusions during the execution of complex hand fine motor action, affecting the complete capture of the motion trajectory and reducing the accuracy of the evaluation; in addition, the overall structure of the evaluation system is complex, the equipment cost is high, and the application and promotion are limited;

[0006] (3) Intelligent assessment based on physiological signals: Due to the limited muscle activity of the affected limb of the patient, the physiological signal acquisition process is easily disturbed by external interference, resulting in a decrease in signal quality and affecting the stability of the assessment. At present, most assessment systems only rely on a single sensor or data acquisition system, which is difficult to fully reflect the hand function status of the patient. In addition, due to significant physiological differences between individuals, existing machine learning algorithms are difficult to effectively adapt to the characteristics of different patients, and the generalization ability is insufficient, which seriously affects the accuracy and universality of the assessment. SUMMARY

[0007] The present application aims to solve the problems of long time-consuming scale assessment, poor consistency of assessment results, complex overall structure of the assessment system, high equipment cost, and poor accuracy and universality of the assessment in the existing assessment method. A multi-modal signal fusion fine motor rehabilitation regulation method and system are proposed, which breaks through the limitations of single-modal physiological signal assessment, constructs a neural-muscle function assessment system based on multi-modal signal fusion, proposes a feature fusion algorithm based on dynamic weight distribution in different rehabilitation stages, and adjusts the weight coefficient dynamically to adapt to the needs of different patients and different rehabilitation stages, improving the comprehensiveness, accuracy and personalized adaptation ability of rehabilitation assessment.

[0008] To achieve the above purpose, the technical scheme adopted by the present application is as follows:

[0009] In a first aspect, the present application provides a multi-modal signal fusion fine motor rehabilitation regulation method, each assessment comprising the following steps:

[0010] S1. Collecting multi-modal signals, the multi-modal signals comprising: pressure distribution signals of hand fine motor, surface electromyography signals and electroencephalogram signals;

[0011] S2. Preprocessing, feature extraction and normalization processing are performed on the pressure distribution signals of hand fine motor, surface electromyography signals and electroencephalogram signals respectively, to obtain pressure feature values, surface electromyography feature values and electroencephalogram feature values;

[0012] S3. Calculating the conversion coefficient of the current rehabilitation stage based on the number of days of fine motor function impairment of the patient, and calculating the weight of the pressure feature values, surface electromyography feature values and electroencephalogram feature values based on the conversion coefficient;

[0013] S4. Weighted sum of the pressure feature values, surface electromyography feature values and electroencephalogram feature values to obtain an assessment score;

[0014] S5. Regulating the rehabilitation training parameters based on the assessment score and the number of days of fine motor function impairment of the patient.

[0015] As one possible implementation, the stress characteristics include the maximum stress value, the minimum stress value, and the trend quantification value; the surface electromyography (EMG) characteristics are the root mean square (RMS) values ​​of the EMG signals; and the EEG characteristics include the RMS values ​​of the EEG signals and event-related desynchronization.

[0016] The assessment score is calculated using the following method:

[0017]

[0018] in, Indicates the assessment score. This indicates the number of days the patient's fine motor function was impaired. Indicates the number of days damaged. The weights of time pressure eigenvalues Indicates the number of days damaged. The weights of surface electromyographic features, Indicates the number of days damaged. Weights of EEG feature values Indicates the maximum pressure. Indicates the minimum pressure. Indicates the trend quantification value. , , These represent the weights of the maximum pressure value, the minimum pressure value, and the trend quantification value, respectively. ; This represents the root mean square value of the surface electromyography signal. This represents the root mean square value of the electroencephalogram (EEG) signal. Indicates that the event is related to synchronization. , and represent the root mean square value of the EEG signal and the weight of event-related desynchronization, respectively. .

[0019] As one possible implementation, S3 includes the following sub-steps:

[0020] S30. Determine the number of days the patient's fine motor function was impaired. ;

[0021] S31. Based on the number of damaged days Calculate the conversion factor for the current rehabilitation stage. The conversion factor includes... coefficients and sum coefficient;

[0022] S32. Based on the number of damaged days Determine the initial weights for pressure characteristic values, surface electromyography characteristic values, and electroencephalography characteristic values, based on coefficients and sum The coefficients are used to calculate the weights of pressure characteristic values, surface electromyography characteristic values, and electroencephalography characteristic values.

[0023] As a possible implementation, the following method is used to calculate the coefficient:

[0024]

[0025] The following method is used to calculate the coefficient:

[0026]

[0027] wherein, represents the number of days of impaired fine motor function of the patient, represents that the current rehabilitation stage is acute stage, represents that the current rehabilitation stage is recovery stage, represents that the current rehabilitation stage is chronic stage.

[0028] As a possible implementation, the following method is used to determine the initial weight:

[0029] If it is the first assessment, according to the number of days of impairment, the initial weight ratio of the stress feature value, the surface electromyography feature value and the electroencephalogram feature value is set to: acute stage: the initial weight ratio is 1:1:4 to 1:1:6;

[0030] recovery stage: the initial weight ratio of the stress feature value, the surface electromyography feature value and the electroencephalogram feature value is set to 1:2:2 to 1:2:3;

[0031] chronic stage: the initial weight ratio of the stress feature value, the surface electromyography feature value and the electroencephalogram feature value is set to 3:1:1 to 4:1:1;

[0032] If it is not the first assessment, the weighted weight of the stress feature value, the surface electromyography feature value and the electroencephalogram feature value calculated in the last assessment is used as the initial weight of the current assessment.

[0033] As a possible implementation, the following method is used to calculate the weight of the electroencephalogram feature value:

[0034]

[0035] The following method is used to calculate the weight of the surface electromyography feature value:

[0036]

[0037] The following method is used to calculate the weight of the stress feature value:

[0038]

[0039] and satisfies: ​​

[0040]

[0041] wherein, represents the weight of the brain electrical characteristic value, represents the weight of the surface electromyogram characteristic value, represents the weight of the stress characteristic value, respectively represent the initial weight of the brain electrical characteristic value in the acute phase, the recovery phase and the chronic phase, respectively represent the initial weight of the surface electromyogram characteristic value in the acute phase, the recovery phase and the chronic phase, respectively represent the initial weight of the stress characteristic value in the acute phase, the recovery phase and the chronic phase.

[0042] As a possible implementation manner, S5 is specifically:

[0043] S50. The evaluation score is divided into a low level score, a medium level score and a high level score; wherein the low level score is 0-40 points of the evaluation score, the medium level score is 41-70 points of the evaluation score, and the high level score is 71-100 points of the evaluation score;

[0044] S51. The rehabilitation training parameters are regulated based on the evaluation score and the number of days of impaired fine motor function of the patient, specifically:

[0045] when the current rehabilitation stage of the patient is the acute phase and the evaluation score is the low level score, the current amplitude is set to 12-14 mA, the frequency is set to 15-20 Hz, the pulse width is set to 180-220 μs, and the training frequency is set to 10-12 times / group;

[0046] when the current rehabilitation stage of the patient is the acute phase and the evaluation score is the medium level score, the current amplitude is set to 14-16 mA, the frequency is set to 20 Hz, the pulse width is set to 200 μs, and the training frequency is set to 12-14 times / group;

[0047] when the current rehabilitation stage of the patient is the acute phase and the evaluation score is the high level score, the current amplitude is set to 15-17 mA, the frequency is set to 20 Hz, the pulse width is set to 200 μs, and the training frequency is set to 12-15 times / group;

[0048] when the current rehabilitation stage of the patient is the recovery phase and the evaluation score is the low level score, the current amplitude is set to 15-17 mA, the frequency is set to 25 Hz, the pulse width is set to 220 μs, and the training frequency is set to 15-18 times / group;

[0049] ​​​​​​When the current rehabilitation stage of the patient is recovery and the evaluation score is a medium level score, the current amplitude is set to 18-20 mA, the frequency is set to 30 Hz, the pulse width is set to 230 μs, and the training times are set to 18-20 times / group;

[0050] When the current rehabilitation stage of the patient is recovery and the evaluation score is a high level score, the current amplitude is set to 20-22 mA, the frequency is set to 30-35 Hz, the pulse width is set to 240-250 μs, and the training times are set to 20-22 times / group;

[0051] When the current rehabilitation stage of the patient is chronic and the evaluation score is a low level score, the current amplitude is set to 18-20 mA, the frequency is set to 30 Hz, the pulse width is set to 240 μs, and the training times are set to 18-20 times / group;

[0052] When the current rehabilitation stage of the patient is chronic and the evaluation score is a medium level score, the current amplitude is set to 22-24 mA, the frequency is set to 35 Hz, the pulse width is set to 250 μs, and the training times are set to 22 times / group;

[0053] When the current rehabilitation stage of the patient is chronic and the evaluation score is a high level score, the current amplitude is set to 24-26 mA, the frequency is set to 40 Hz, the pulse width is set to 260-280 μs, and the training times are set to 25 times / group.

[0054] In a second aspect, the present application provides a fine motor rehabilitation control system based on multi-modal signal fusion, comprising:

[0055] A multi-modal signal acquisition module is configured to acquire multi-modal signals, wherein the multi-modal signals include pressure distribution signals of hand fine motor, surface electromyography signals and electroencephalography signals;

[0056] A data processing module is configured to pre-process, extract features and normalize the acquired pressure distribution signals, surface electromyography signals and electroencephalography signals, and obtain pressure features, surface electromyography features and electroencephalography features, respectively;

[0057] A rehabilitation evaluation module is configured to automatically evaluate rehabilitation progress based on a feature-level fusion algorithm;

[0058] A feedback and adaptive control module is configured to dynamically control the weights of the pressure distribution signals, surface electromyography signals and electroencephalography signals based on the rehabilitation progress.

[0059] As a possible implementation manner, the multi-modal signal acquisition module includes a pressure distribution signal acquisition unit, a surface electromyography signal acquisition unit and an electroencephalography signal acquisition unit;

[0060] The pressure distribution signal acquisition unit comprises an array type flexible pressure sensor subunit, a pressure distribution signal processing subunit and an upper computer, the array type flexible pressure sensor subunit comprises a hand holding sensing array and a sleeve sensing array, the hand holding sensing array comprises high-density independent sensitive points, the center distance between adjacent sensitive points is 0.5-2mm, and the scanning frequency is 50-500Hz; the sleeve sensing array comprises high-density independent sensitive points, the center distance between adjacent sensitive points is 0.5-1mm, and the scanning frequency is 50-500Hz; the array type flexible pressure sensor subunit is connected with the pressure distribution signal processing subunit through a flexible circuit board, the pressure distribution signal processing subunit adopts a sliding differential algorithm to eliminate the static error introduced by the deformation of the sensor base, and transmits the pressure data filtered to the upper computer; and the upper computer is used for calculating the pressure values of the sensitive points, mapping the pressure values of the sensitive points to a standard anatomical coordinate system, and generating a pressure thermal map in real time to display the pressure distribution of each region of the hand.

[0061] The surface electromyography signal acquisition unit adopts an 8-channel electromyography acquisition bracelet to acquire the surface electromyography signal.

[0062] The electroencephalogram signal acquisition unit adopts a saline electrode to acquire the electroencephalogram signal.

[0063] As a possible implementation manner, the hand holding sensing array comprises 4000 independent sensitive points, and the arrangement mode of the 4000 independent sensitive points is 80 rows and 50 columns; the sleeve sensing array comprises 1000 independent sensitive points, and the arrangement mode of the 1000 independent sensitive points is 40 rows and 25 columns.

[0064] Compared with the prior art, the present application has the following beneficial effects:

[0065] 1. The multi-modal signal fusion fine motor rehabilitation regulation method breaks through the limitation of traditional single-modal physiological signal evaluation, constructs a neural-muscular function evaluation system based on multi-modal signal fusion, proposes a feature fusion algorithm based on dynamic weight distribution in different rehabilitation stages, dynamically adjusts the weight coefficient to adapt to the needs of different patients and different rehabilitation stages, and improves the comprehensiveness, accuracy and personalized adaptation ability of rehabilitation evaluation.

[0066] 2. The multi-modal signal fusion fine motor rehabilitation regulation method and system adopt a high-density distributed flexible pressure sensing matrix, can acquire and quantify the pressure distribution and dynamic change of patients in hand fine movements such as gripping and pointing in real time and accurately, overcome the limitation of traditional single-point measurement, and improve the evaluation accuracy.

[0067] 3. The multi-modal signal fusion fine motor rehabilitation control method according to the present application constructs a 3x3 joint control matrix of evaluation scores and rehabilitation stages to ensure that the electrical stimulation parameters are dynamically matched with the patient's movement state. This breaks through the limitations of traditional fixed parameter electrical stimulation schemes, significantly improves the accuracy, personalization and training effect of rehabilitation training, enhances the autonomous activation of the patient's neuromuscular system, promotes neural plasticity, and provides support for long-term rehabilitation of patients.

[0068] 4. The multi-modal signal fusion fine motor rehabilitation control method according to the present application determines whether it is the first evaluation according to the actual situation of the patient, dynamically allocates initial weights based on the rehabilitation stage of the patient during the first evaluation, and different rehabilitation stages are dominated by different modal signals; when not the first evaluation, the weight of the last evaluation is inherited, and the continuity of the evaluation index between multiple training is ensured by inheritance, and the dynamic evolution trend is completely retained, which can ensure the accuracy and continuity of the evaluation. BRIEF DESCRIPTION OF DRAWINGS

[0069] The accompanying drawings, which are included to provide a further understanding of the application and constitute a part of this application, illustrate certain illustrative embodiments of the application and together with the description serve to explain the application. In the drawings:

[0070] Figure 1 A multi-modal signal fusion fine motor rehabilitation control method flowchart is provided for the embodiments of the present application;

[0071] Figure 2 A multi-modal signal fusion fine motor rehabilitation control system structure schematic diagram is provided for the embodiments of the present application;

[0072] Figure 3 A pressure distribution signal acquisition unit structure schematic diagram is provided for the embodiments of the present application.

[0073] REFERENCE NUMERALS

[0074] 1 - multi-modal signal acquisition module, 10 - pressure distribution signal acquisition unit, 100 - array type flexible pressure sensor subunit, 1000 - hand sensing array, 1001 - sleeve sensing array, 101 - pressure distribution signal processing subunit, 102 - upper computer, 11 - surface electromyography signal acquisition unit, 12 - electroencephalogram signal acquisition unit, 2 - data processing module, 3 - rehabilitation evaluation module, 4 - feedback and adaptive control module. DETAILED DESCRIPTION

[0075] In order to clearly describe the technical solutions of the embodiments of the present application, in the embodiments of the present application, the terms of "first", "second", etc. are used to distinguish the same or similar items with basically the same function and role. For example, the first threshold and the second threshold are only used to distinguish different thresholds, and the order is not limited. Those skilled in the art can understand that the terms of "first", "second", etc. do not limit the number and execution order, and the terms of "first", "second", etc. also do not necessarily mean different.

[0076] It should be noted that in the present application, the words "exemplary" or "for example" are used to mean serving as an example, instance, or illustration. Any embodiment or design presented as "exemplary" or "for example" in the present application should not be interpreted as being more preferred or advantageous than other embodiments or designs. Rather, the use of the words "exemplary" or "for example" is intended to present concepts in a particular manner.

[0077] In the present application, "at least one" means one or more, and "multiple" means two or more. The association relationship of the associated objects is described, which means that there can be three relationships, for example, A and / or B, which can represent the following cases: A exists alone, A and B exist together, and B exists alone, where A and B can be singular or plural. The character " / " generally represents an "or" relationship between the associated objects. The following at least one (or similar expressions) means any combination of these items, including any combination of single (or multiple) items. For example, at least one of a, b or c can represent: a, b, c, a and b, a and c, b and c, or a, b and c, where a, b, c can be single or multiple.

[0078] The embodiments of the present application aim to provide a fine motor rehabilitation control method based on multi-modal signal fusion. By means of the pressure distribution signal of hand fine motor, the surface electromyography signal and the electroencephalogram signal, the key physiological parameters are obtained in real time, the evaluation score is calculated combined with the duration of impaired fine motor function of the patient, and then the rehabilitation training parameters are dynamically adjusted, thereby constructing a whole-process personalized rehabilitation scheme covering "signal acquisition-multi-modal signal fusion-quantitative output-closed-loop control", and improving the accuracy, effectiveness, personalization and adaptability of the rehabilitation training.

[0079] In a first aspect, the embodiments of the present application provide a fine motor rehabilitation control method based on multi-modal signal fusion, referring to Figure 1 Each evaluation includes the following steps:

[0080] S1. Collecting multi-modal signals, the multi-modal signals including: pressure distribution signals of hand fine motor, surface electromyography signals and electroencephalogram signals;

[0081] As an example, a torsion bar for rehabilitation is taken as a carrier, and an 80-row 50-column array flexible pressure sensor is integrated on the torsion bar, the sensor thickness can be selected between 0.1 mm and 0.3 mm, 4000 independent pressure sensing points are formed, the center distance between adjacent sensing points is 2 mm, the scanning frequency is 50 Hz, the pressure signal is converted into an electric signal through the pressure sensor, and high-sensitivity pressure detection is realized. Such design can adapt to the hand size of different patients, accurately measure the pressure distribution of each finger part of the stroke patient during the gripping process in different rehabilitation stages, and dynamically record the change of pressure with time, so as to ensure the accuracy and universality of the rehabilitation training. Based on the array flexible pressure sensor and the medical elastic fabric, a 40-row 25-column array pressure sensor (1000 sensing points, each sensing point has a spacing of 1 mm) is designed to form a thumb sleeve type pressure sensing unit, real-time acquisition of dynamic pressure changes of the patient during gripping, finger pointing and other processes is realized, and the dynamic mechanical characteristics of the fine motor of the hand are evaluated through time series analysis. Based on the multi-channel portable electromyography bracelet, the surface electromyography (sEMG) signals of the key muscle groups related to rehabilitation are collected, which are used to evaluate the activity state and movement execution ability of the muscle. The saline electrode cap is used to collect the electroencephalogram (EEG) signals of the brain area related to movement, to decode the movement intention and evaluate the control ability of the brain to the hand movement.

[0082] The present application accurately collects the pressure distribution signals of the fine motor of the hand, and simultaneously fuses the surface electromyography signals and the electroencephalogram signals, so as to extract the coordinated control characteristics of the neural-muscular system, realize the synchronous collection, feature extraction and dynamic evaluation of the multi-modal movement information, and help to improve the effect of the fine motor rehabilitation training of the hand.

[0083] S2. The pressure distribution signals, the surface electromyography signals and the electroencephalogram signals of the fine motor of the hand are respectively preprocessed, feature extracted and normalized, to obtain pressure feature values, surface electromyography feature values and electroencephalogram feature values;

[0084] For example, in order to ensure the accuracy of rehabilitation evaluation and regulation, the collected pressure distribution signals, surface electromyography signals and electroencephalogram signals need to be subjected to baseline calibration, filtering and other operations, so as to improve the signal quality.

[0085] For example, the pressure feature values include a maximum pressure value, a minimum pressure value and a trend quantization value; the surface electromyography feature values are the root mean square values of the surface electromyography signals; and the electroencephalogram feature values include the root mean square values and event-related desynchronization of the electroencephalogram signals.

[0086] In specific implementation, the pressure distribution signal is collected in a cycle according to a frequency of 50 Hz, a sliding window of 100 ms is adopted, median filtering of a window length of 5 points is performed, noise in the signal is removed, and the original characteristics of the signal are maintained as much as possible during filtering.

[0087] (1)

[0088] wherein, represents the number of data points of the window, represents the original pressure signal value collected at the i th time point, represents the noise-removed pressure signal value after median filtering processing of the i th time point.

[0089] The maximum value, minimum value and trend quantization value of the pressure are extracted as the pressure characteristic values, as shown in the following formula (2) and formula (3):

[0090] , (2)

[0091] (3)

[0092] wherein, is the time point of the sliding window.

[0093] A fourth-order Butterworth filter is adopted to perform band-pass filtering of 20-500 Hz, and the root mean square value (RMS) of the surface electromyogram signal is calculated as the surface electromyogram characteristic value according to the following formula (4):

[0094] (4)

[0095] wherein, is the surface electromyogram sampling point, is the window size.

[0096] When collecting the electroencephalogram signal, band-pass filtering of 0.1 Hz-35 Hz and 50 Hz notch filtering are performed, the 8-13 Hz band and the 13-30 Hz band signals related to movement are extracted, and the root mean square value (RMS) and event-related desynchronization (ERD) of the electroencephalogram signal are extracted as the electroencephalogram characteristic values, as shown in the following formula (5) and (6): (5)

[0097] wherein,

[0098] ​​​​​​​For the sampling points of the electroencephalogram signal, For the window size.

[0099] (6)

[0100] Wherein, is the average power of 1s before the action (rest state), is the average power of 2s during the action (task state).

[0101] In order to ensure that different modal signals have the same scale, facilitate subsequent multi-modal signal fusion, according to the range of the pressure sensor and the reference value of the healthy population, the maximum threshold is set, and the pressure distribution signal original value is mapped to the [0, 1] interval; for the surface electromyogram signal, according to the maximum expected electromyogram activity intensity, the threshold is set, and the normalization is carried out, so as to avoid individual difference or noise interference; for the electroencephalogram signal, the baseline data of the resting state is used for normalization, and the relative change amount is used, so as to eliminate the individual baseline difference.

[0102] S3. Calculate the conversion coefficient of the current rehabilitation stage based on the number of days of impaired fine motor function of the patient, and calculate the weights of the pressure feature value, the surface electromyogram feature value and the electroencephalogram feature value based on the conversion coefficient;

[0103] As a possible implementation manner, S3 includes the following sub-steps:

[0104] S30. Determine the number of days of impaired fine motor function of the patient ;

[0105] S31. Calculate the conversion coefficient of the current rehabilitation stage based on the number of impaired days , the conversion coefficient includes coefficient and coefficient;

[0106] In specific implementation, for each time window , a multi-dimensional feature vector is constructed, as shown in formula (7):

[0107] (7)

[0108] Based on the feature vector, a feature-level fusion algorithm based on dynamic weight distribution is constructed. The algorithm is based on the neural rehabilitation time window theory, combines the complementarity between multi-modal signals (pressure distribution signal, surface electromyography signal, electroencephalogram signal), and designs a dynamic weight distribution mechanism to adjust the weight of each signal in different recovery stages to optimize the accuracy of data fusion in the patient's rehabilitation process. The neural rehabilitation time window theory divides the rehabilitation stage into five stages: acute stage (0-14 days), subacute stage (15-60 days), recovery stage (60 days-180 days), chronic stage (180 days-365 days) and long-term chronic stage (more than 1 year). Based on the above neural rehabilitation time window theory, the application constructs The coefficient is used to represent the conversion degree of the patient from the acute stage to the recovery stage, and the application constructs The coefficient is used to represent the conversion degree of the patient from the chronic stage to the recovery stage.

[0109] As a possible implementation manner, the following method is used to calculate The coefficient is:

[0110] (8)

[0111] The following method is used to calculate The coefficient is:

[0112] (9)

[0113] Wherein, represents the number of days of impaired fine motor function of the patient, represents that the current rehabilitation stage is the acute stage, represents that the current rehabilitation stage is the recovery stage, represents that the current rehabilitation stage is the chronic stage.

[0114] S32. Based on the number of impaired days Determine the initial weight of the pressure feature value, the surface electromyography feature value and the electroencephalogram feature value, based on The coefficient and The coefficient is used to calculate the weight of the pressure feature value, the surface electromyography feature value and the electroencephalogram feature value.

[0115] As a possible implementation manner, the following method is used to determine the initial weight:

[0116] If it is the first evaluation, according to the number of impaired days, the initial weight ratio of the pressure feature value, the surface electromyography feature value and the electroencephalogram feature value is set to: acute stage: initial weight ratio is 1:1:4 to 1:1:6; wherein, the weight of the electroencephalogram signal is the largest, and highlights the important role of the electroencephalogram signal in decoding the motor intention in the early rehabilitation.

[0117] Recovery stage: the initial weight ratio of the pressure feature value, the surface electromyogram feature value and the electroencephalogram feature value is set to 1:2:2 to 1:2:3; the electroencephalogram signal still dominates in this stage, and the muscle electromyogram feature weight is significantly improved, reflecting the trend of the synergistic enhancement of central motor intention and muscle execution.

[0118] Chronic stage: the initial weight ratio of the pressure feature value, the surface electromyogram feature value and the electroencephalogram feature value is set to 3:1:1 to 4:1:1; in this stage, the pressure signal weight gradually dominates to highlight the core position of the end execution ability and fine motion recovery.

[0119] As an example, the initial weight distribution of the pressure feature value, the surface electromyogram feature value and the electroencephalogram feature value is shown in Table 1:

[0120] Table 1 Initial weight distribution of the pressure feature value, the surface electromyogram feature value and the electroencephalogram feature value

[0121]

[0122] If it is not the first evaluation, the weighted weights of the pressure feature value, the surface electromyogram feature value and the electroencephalogram feature value calculated in the last evaluation are used as the initial weights of the current evaluation, and the inherited setting can ensure the continuity of the evaluation indexes between multiple training and completely retain the dynamic evolution trend.

[0123] As a possible implementation manner, the weight of the electroencephalogram feature value is calculated by the following method:

[0124] (10)

[0125] The weight of the surface electromyogram feature value is calculated by the following method:

[0126] (11)

[0127] The weight of the pressure feature value is calculated by the following method:

[0128] (12)

[0129] and satisfy:

[0130] (13)

[0131] wherein, represents the weight of the electroencephalogram feature value, represents the weight of the surface electromyogram feature value, represents the weight of the pressure feature value, , , respectively represent the initial weights of the electroencephalogram feature value in the acute stage, the recovery stage and the chronic stage, 、 、 respectively represent the initial weights of the surface electromyogram characteristic values in the acute phase, recovery phase and chronic phase, 、 、 respectively represent the initial weights of the stress characteristic values in the acute phase, recovery phase and chronic phase.

[0132] S4. Weighted sum of the stress characteristic values, the surface electromyogram characteristic values and the electroencephalogram characteristic values to obtain an evaluation score;

[0133] As a possible implementation manner, the evaluation score is obtained by using the following method:

[0134] (14)

[0135] wherein, represents the evaluation score, represents the number of days of impaired fine motor function of the patient, represents the weight of the stress characteristic values when the number of impaired days is represents the weight of the surface electromyogram characteristic values when the number of impaired days is represents the weight of the electroencephalogram characteristic values when the number of impaired days is represents the maximum value of stress, represents the minimum value of stress, represents the trend quantification value, 、 、 respectively represent the weights of the maximum value of stress, the minimum value of stress and the trend quantification value, and ; represents the root mean square value of the surface electromyogram signal, represents the root mean square value of the electroencephalogram signal, represents the event-related desynchronization, 、 respectively represent the weights of the root mean square value of the electroencephalogram signal and the event-related desynchronization, and .

[0136] S5. Regulating the rehabilitation training parameters based on the evaluation score and the number of days of impaired fine motor function of the patient.

[0137] As a possible implementation manner, S5 is specifically:

[0138] S50. Dividing the evaluation score into a low-level score, a medium-level score and a high-level score; wherein the low-level score is 0-40 points of the evaluation score, the medium-level score is 41-70 points of the evaluation score, and the high-level score is 71-100 points of the evaluation score.​​​

[0139] The application divides the evaluation score into three levels: low level score (0-40), medium level score (41-70) and high level score (71-100); at the same time, the rehabilitation stage corresponding to the number of days of impaired fine motor function is divided into: acute stage (≤14 days), recovery stage (15-180 days) and chronic stage (>180 days), forming a 3x3 joint control matrix.

[0140] S51. Regulate the rehabilitation training parameters based on the evaluation score and the number of days of impaired fine motor function of the patient, specifically:

[0141] When the current rehabilitation stage of the patient is the acute stage and the evaluation score is the low level score, set the current amplitude 12-14 mA, the frequency 15-20 Hz, the pulse width 180-220 μs, and the training frequency 10-12 times / group, and the target is to activate the motor intention and central nervous response;

[0142] When the current rehabilitation stage of the patient is the acute stage and the evaluation score is the medium level score, set the current amplitude 14-16 mA, the frequency 20 Hz, the pulse width 200 μs, and the training frequency 12-14 times / group, and the target is to establish the preliminary neural-muscular pathway;

[0143] When the current rehabilitation stage of the patient is the acute stage and the evaluation score is the high level score, set the current amplitude 15-17 mA, the frequency 20 Hz, the pulse width 200 μs, and the training frequency 12-15 times / group, and the target is to stabilize the central control performance;

[0144] When the current rehabilitation stage of the patient is the recovery stage and the evaluation score is the low level score, set the current amplitude 15-17 mA, the frequency 25 Hz, the pulse width 220 μs, and the training frequency 15-18 times / group, and the target is to induce muscle contraction and establish coordinated control;

[0145] When the current rehabilitation stage of the patient is the recovery stage and the evaluation score is the medium level score, set the current amplitude 18-20 mA, the frequency 30 Hz, the pulse width 230 μs, and the training frequency 18-20 times / group, and the target is to strengthen the muscle strength output and endurance training;

[0146] When the current rehabilitation stage of the patient is the recovery stage and the evaluation score is the high level score, set the current amplitude 20-22 mA, the frequency 30-35 Hz, the pulse width 240-250 μs, and the training frequency 20-22 times / group, and the target is to improve the fine operation ability and neuromuscular coordination;

[0147] When the current rehabilitation stage of the patient is the chronic stage and the evaluation score is the low level score, set the current amplitude 18-20 mA, the frequency 30 Hz, the pulse width 240 μs, and the training frequency 18-20 times / group, and the target is to strengthen the residual motor ability;

[0148] When the current rehabilitation stage of the patient is chronic and the evaluation score is a medium level score, set the current amplitude 22-24 mA, the frequency 35 Hz, the pulse width 250 μs, the training times 22 times / group, and enhance the continuous force and fine control;

[0149] When the current rehabilitation stage of the patient is chronic and the evaluation score is a high level score, set the current amplitude 24-26 mA, the frequency 40 Hz, the pulse width 260-280 μs, the training times 25 times / group, and achieve the comprehensive reconstruction of fine motion function.

[0150] Next, the method is used to treat a patient Wang, male, 58 years old, 1 year after stroke (brain stroke) The current rehabilitation stage of the patient is chronic, and the thumb-index finger opposition motion of the affected limb has poor coordination and insufficient strength. The rehabilitation goal is to improve the fine motor ability of the hand, especially the rehabilitation of the opposition function. At present, based on the evaluation system, a week of rehabilitation training has been completed, and the rehabilitation training parameters are adjusted.

[0151] The patient wears a thumb sleeve type flexible pressure sensing unit (40 rows 25 columns, a total of 1000 sensitive points), and completes the thumb and index finger opposition rehabilitation training under the drive of functional electrical stimulation. The pressure between the thumb and index finger during opposition is collected by the thumb sleeve type flexible pressure sensing unit. An 8-channel portable electromyography bracelet is worn on the extensor and flexor muscle groups of the forearm to collect electromyography signals. A saline electrode is used to collect motor area electroencephalogram signals. The patient completes motor imagery by watching virtual animations of opposition, and decodes motor intention to trigger a functional electrical stimulator.

[0152] Place the stimulating electrode on the key muscle groups (long flexor of the thumb, deep flexor of the fingers, etc.) that complete the opposition of the thumb and index finger. Set the stimulation parameters according to the previous rehabilitation training as the default values (such as current intensity 17 mA, pulse width 200 us).

[0153] During the rehabilitation training process, the patient completes 10 times of thumb and index finger opposition training in each block under the drive of functional electrical stimulation. Synchronous acquisition of pressure signals, surface electromyography signals, and electroencephalogram signals is performed, and normalization processing is performed.

[0154] The maximum pressure recorded when completing the opposition motion is 1.3 N, and the minimum value is 0.35 N, which is quantified The current amplitude is 1.1 N / s, and the normalized values are 0.65, 0.35 and 0.55 in turn. The root mean square value of the collected electromyogram of the patient is 0.9 mV, and the normalized value is 0.3. The electroencephalogram of the patient is collected synchronously, and the root mean square value of the collected electroencephalogram of the patient is 6.5 uV, and the event-related desynchronization of the alpha band is 35%, and the normalized values are 0.65 and 0.35 in turn. The weights of the electroencephalogram feature value, the surface electromyogram feature value and the pressure feature value are calculated according to the current recovery stage of the patient and formulas (8), (9), (10), (11), (12), and the weights are 0.433, 0.2835 and 0.2835 in turn. The final evaluation score is calculated according to formula (14) and is 46.

[0155] Based on the regulation method proposed in the embodiment, the score is a medium level score, the patient is in the chronic stage, and according to the score and the current rehabilitation stage, the current amplitude is set to 22-24 mA, the frequency is 35 Hz, the pulse width is 250 us, and the training times are 22 times / group, so as to enhance the continuous force and fine control. During the experiment, the current amplitude of Wang is adjusted from 19 mA to 23 mA, the pulse width is increased from 240 us to 250 us, the training intensity is increased, 3 groups of thumb and index finger fine motor training are added to each block, visual stimulation is introduced combined with brain-computer interface technology, and the accuracy of motion intention decoding is enhanced. After two weeks of new rehabilitation training, the evaluation system score of the patient Wang is increased from 46 to 52, which indicates that the thumb and index finger opposition ability of the patient is significantly improved, and the personalized rehabilitation training scheme is effective.

[0156] In a second aspect, the embodiment of the present application provides a fine motor rehabilitation regulation system based on multi-modal signal fusion, which is shown in Figure 2 , and includes:

[0157] The multi-modal signal acquisition module 1 is used for acquiring multi-modal signals, and the multi-modal signals include: pressure distribution signals of hand fine motor, surface electromyogram signals and electroencephalogram signals.

[0158] Referring to Figure 2 , as a possible implementation manner, the multi-modal signal acquisition module 1 includes a pressure distribution signal acquisition unit 10, a surface electromyogram signal acquisition unit 11 and an electroencephalogram signal acquisition unit 12.

[0159] Referring to Figure 3, the pressure distribution signal acquisition unit 10 comprises an array type flexible pressure sensor subunit 100, a pressure distribution signal processing subunit 101 and a host computer 102, the array type flexible pressure sensor subunit 100 comprises a hand holding sensing array 1000 and a sleeve sensing array 1001, the hand holding sensing array 1000 comprises high-density independent sensitive points, the center distance between adjacent sensitive points is 0.5-2 mm, and the scanning frequency is 50-500 Hz; the sleeve sensing array comprises high-density independent sensitive points, the center distance between adjacent sensitive points is 0.5-1 mm, and the scanning frequency is 50-500 Hz; the array type flexible pressure sensor subunit 100 is connected with the pressure distribution signal processing subunit 101 through a flexible circuit board, the pressure distribution signal processing subunit 101 adopts a sliding differential algorithm to eliminate the static error introduced by the deformation of the sensor substrate, and transmits the pressure data to the host computer 102 after filtering processing; the host computer 102 is used for calculating the pressure values of the sensitive points, mapping the pressure values of the sensitive points to a standard anatomical coordinate system, and generating a pressure thermal map in real time to display the pressure distribution of each region of the hand.

[0160] As a possible implementation manner, the hand holding sensing array comprises 4000 independent sensitive points, and the arrangement mode of the 4000 independent sensitive points is 80 rows 50 columns; the sleeve sensing array comprises 1000 independent sensitive points, and the arrangement mode of the 1000 independent sensitive points is 40 rows 25 columns.

[0161] Referring to Figure 3 As an example, the hand holding sensing array 1000 takes a torsion rod for rehabilitation as a carrier, and an array flexible pressure sensor of 80 rows 50 columns is attached to the surface, the thickness of the sensor can be selected between 0.1 mm and 0.3 mm, 4000 independent sensitive points capable of sensing pressure are formed, the center distance between adjacent sensitive points is 2 mm, the scanning frequency is 50 Hz, the pressure signal is converted into an electric signal through the pressure sensor, and high-sensitivity pressure detection is realized. This design can adapt to the hand size of different patients, accurately measure the pressure distribution of each finger part of a stroke patient during the gripping process in different rehabilitation stages, and dynamically record the change of pressure with time, so as to ensure the accuracy and universality of rehabilitation training. The sleeve sensing array 1001 combines an array flexible pressure sensor and a medical elastic fabric to form a sleeve conforming to human engineering, and the area covers the distal phalanx region. The sleeve sensing array 1001 integrates 40 rows 25 columns of micro flexible pressure sensing units, the distance between adjacent sensitive points is 1 mm, there are 1000 independent sensitive points, the pressure change of the fingertip during fine hand movements such as finger pointing can be accurately monitored, the mechanical characteristics are extracted, and the rehabilitation progress is quantified.

[0162] As an example, the array flexible pressure sensor subunit 100 is connected to the pressure distribution signal processing subunit 101 through a flexible circuit board, realizes array scanning, and adopts a sliding differential algorithm to eliminate static errors introduced by deformation of the sensor substrate. After filtering, the pressure data is transmitted to the host computer 102 through wired or wireless means. The host computer 102 is responsible for real-time data processing, including calculating the pressure values of each sensitive point, extracting key mechanical characteristics such as maximum value, minimum value and trend quantization. At the same time, based on the hand and finger biomechanical model, the pressure data is mapped to the standard anatomical coordinate system to realize accurate positioning. Moreover, the host computer can generate a pressure heat map in real time, which can intuitively display the pressure distribution of each region, and allow users to view the specific pressure values of each sensitive point, providing an efficient and intuitive rehabilitation evaluation tool.

[0163] The surface electromyography signal acquisition unit 11 adopts an 8-channel electromyography acquisition bracelet to acquire surface electromyography signals.

[0164] Even in the case of severely impaired motor function, patients may still have low-amplitude spontaneous electromyographic activity and produce corresponding electromyographic responses during passive movement induced by functional electrical stimulation (FES). The present embodiment uses a high-sensitivity portable electromyography bracelet to acquire surface electromyography signals. An 8-channel electromyography acquisition bracelet is used, based on metal electrodes, integrating an active signal acquisition unit with high input impedance, which can effectively reduce external noise interference and improve the signal-to-noise ratio and stability of surface electromyography signals. The use of a Bluetooth module enables wireless transmission of surface electromyography signals, ensuring comfort and convenience during evaluation. According to specific fine motor rehabilitation tasks and different rehabilitation stages, the electromyography bracelet is worn on the key motor groups of the corresponding movements, enabling surface electromyography signal acquisition. By accurately capturing muscle activity patterns, the bracelet can support dynamic assessment of hand fine motor rehabilitation in stroke patients and provide reliable data support for developing personalized rehabilitation intervention strategies.

[0165] The electroencephalogram signal acquisition unit 12 uses a saline electrode to acquire electroencephalogram signals.

[0166] As an example, a saline electrode is used to acquire electroencephalogram signals, and KCl or NaCl solution is used as the conductive medium, effectively reducing the contact impedance between the electrode and the skin, overcoming the cumbersome process of applying conductive paste to traditional wet electrodes and the need for patients to wash, and achieving convenient acquisition of high-quality electroencephalogram signals. During the acquisition process, a 64-channel standard electroencephalogram cap is used, according to the international 10-20 system standard electrode placement method, to ensure that the electrodes remain moist but do not cause short circuits between the electrodes. When the patient performs hand fine motor tasks (such as grasping and finger-to-finger movements), the electroencephalogram signals related to motor intention are acquired and analyzed in real time, and the motor intention characteristics are extracted, providing reliable neural feedback data for brain-controlled functional electrical stimulation.

[0167] a data processing module 2 configured to preprocess, extract features from, and normalize the collected pressure distribution signals, surface electromyography signals, and electroencephalography signals to obtain pressure features, surface electromyography features, and electroencephalography features, respectively;

[0168] a rehabilitation assessment module 3 configured to automatically assess rehabilitation progress based on the features-level fusion algorithm;

[0169] a feedback and adaptive regulation module 4 configured to dynamically regulate the weights of the pressure distribution signals, surface electromyography signals, and electroencephalography signals based on the rehabilitation progress.

[0170] Although the present application has been described in connection with various embodiments thereof, those skilled in the art will understand that various modifications in form and detail can be made therein without departing from the spirit and scope of the application. In the description of the specification the word "comprising" does not exclude other elements or steps and the indefinite article "a" or "an" does not exclude a plurality. A single processor or other unit can fulfil the functions of several items recited in the specification. The mere fact that certain measures are recited in mutually different embodiments does not indicate that a combination of these measures cannot be used to an advantage.

[0171] Although the present application has been described in connection with particular features and embodiments thereof, it will be evident that various modifications and combinations can be made thereto, and that the generic principles defined herein can be applied to other embodiments and applications without departing from the spirit and scope of the application. Accordingly, the description and drawings are to be regarded as illustrative in nature and not as restrictive. Obviously, many modifications and variations of the present application are possible in light of the above teachings. It is therefore to be understood that within the scope of the application, modifications can be made by those skilled in the art in their implementation of the application. Accordingly, the disclosure and practice can only be limited by the claims that follow.

Claims

1. A method for fine motor rehabilitation regulation through multimodal signal fusion, characterized in that, Each evaluation includes the following steps: S1. Acquire multimodal signals, including: pressure distribution signals of fine hand movements, surface electromyography signals, and electroencephalography signals; S2. The pressure distribution signal, surface electromyography signal and electroencephalogram (EEG) signal of the fine hand movements are preprocessed, feature extracted and normalized respectively to obtain pressure feature values, surface electromyography feature values ​​and EEG feature values; S3. Calculate the conversion coefficient of the current rehabilitation stage based on the number of days the patient's fine motor function is impaired, and calculate the weights of pressure characteristic value, surface electromyography characteristic value and electroencephalography characteristic value based on the conversion coefficient; S3 includes the following sub-steps: S30. Determine the number of days the patient's fine motor function was impaired. ; S31. Based on the number of damaged days Calculate the conversion factor for the current rehabilitation stage, the conversion factor including coefficients and sum coefficient; Calculate using the following method coefficient: Calculate using the following method coefficient: in, This indicates the number of days the patient's fine motor function was impaired. This indicates that the current recovery phase is the acute phase. This indicates that the current recovery phase is the recovery period. This indicates that the current recovery phase is the chronic phase; S32. Based on the number of damaged days Determine the initial weights for pressure characteristic values, surface electromyography characteristic values, and electroencephalography characteristic values, and based on... coefficients and sum The coefficients are used to calculate the weights of pressure characteristic values, surface electromyography characteristic values, and electroencephalography characteristic values. S4. The pressure characteristic value, surface electromyography characteristic value, and electroencephalography characteristic value are weighted and summed to obtain an evaluation score; the pressure characteristic value includes the maximum pressure value, the minimum pressure value, and the trend quantification value; the surface electromyography characteristic value is the root mean square value of the surface electromyography signal; the electroencephalography characteristic value includes the root mean square value of the electroencephalography signal and the event-related desynchronization; The assessment score is calculated using the following method: in, Indicates the assessment score. This indicates the number of days the patient's fine motor function was impaired. Indicates the number of days damaged. The weights of time pressure eigenvalues Indicates the number of days damaged. The weights of surface electromyographic features, Indicates the number of days damaged. Weights of EEG feature values Indicates the maximum pressure. Indicates the minimum pressure. Indicates the trend quantification value. , , These represent the weights of the maximum pressure value, the minimum pressure value, and the trend quantification value, respectively. ; This represents the root mean square value of the surface electromyography signal. This represents the root mean square value of the electroencephalogram (EEG) signal. Indicates that the event is related to synchronization. , and represent the root mean square value of the EEG signal and the weight of event-related desynchronization, respectively. ; S5. Adjust rehabilitation training parameters based on assessment scores and the number of days the patient's fine motor function is impaired. The rehabilitation training parameters are the current amplitude, frequency, pulse width, and number of training sessions under different level scores.

2. The method for fine motor rehabilitation regulation based on multimodal signal fusion according to claim 1, characterized in that, The initial weights are determined using the following method: If this is the first assessment, the initial weighting ratio of pressure characteristic value, surface electromyography characteristic value and electroencephalography characteristic value is set to: acute phase: 1:1:4 to 1:1:6, based on the number of days of injury. Recovery period: The initial weighting ratio of pressure characteristic value, surface electromyography characteristic value and electroencephalography characteristic value is set to 1:2:2 to 1:2:3; Chronic phase: The initial weighting ratio of pressure characteristic value, surface electromyography characteristic value and electroencephalography characteristic value is set to 3:1:1 to 4:1:1; If this is not the first assessment, the weighted average of the pressure characteristic value, surface electromyography characteristic value, and electroencephalography characteristic value calculated in the previous assessment will be used as the initial weight for this assessment.

3. The method for fine motor rehabilitation regulation based on multimodal signal fusion according to claim 1, characterized in that, The weights of EEG feature values ​​are calculated using the following method: The weights of surface electromyography (EMG) features are calculated using the following method: The weights of the pressure eigenvalues ​​are calculated using the following method: And it satisfies: in, , , These represent the initial weights of the EEG characteristic values ​​during the acute, recovery, and chronic phases, respectively. , , These represent the initial weights of the surface electromyography (SEM) characteristics during the acute, recovery, and chronic phases, respectively. , , These represent the initial weights of the stress characteristic values ​​during the acute, recovery, and chronic phases, respectively.

4. The method for fine motor rehabilitation regulation based on multimodal signal fusion according to claim 1, characterized in that, Specifically, S5 is: S50. The assessment scores are divided into low-level scores, medium-level scores, and high-level scores; where low-level scores are 0-40 points, medium-level scores are 41-70 points, and high-level scores are 71-100 points. S51. Adjust rehabilitation training parameters based on assessment scores and the number of days the patient's fine motor function was impaired, specifically as follows: When the patient is currently in the acute phase of rehabilitation and the assessment score is low, set the current amplitude to 12-14 mA, the frequency to 15-20 Hz, the pulse width to 180-220 μs, and the number of training sessions to 10-12 times per set. When the patient is currently in the acute phase of rehabilitation and the assessment score is at a moderate level, the current amplitude is set to 14-16 mA, the frequency to 20 Hz, the pulse width to 200 μs, and the number of training sessions to 12-14 times per set. When the patient is currently in the acute phase of rehabilitation and the assessment score is at a high level, set the current amplitude to 15-17mA, the frequency to 20Hz, the pulse width to 200μs, and the number of training sessions to 12-15 times per set. When the patient is currently in the recovery phase and the assessment score is low, set the current amplitude to 15-17mA, the frequency to 25Hz, the pulse width to 220μs, and the number of training sessions to 15-18 times per set. When the patient is currently in the recovery phase and the assessment score is at a moderate level, set the current amplitude to 18-20mA, the frequency to 30Hz, the pulse width to 230μs, and the number of training sessions to 18-20 times per set. When the patient is currently in the recovery phase and the assessment score is at a high level, set the current amplitude to 20-22mA, the frequency to 30-35Hz, the pulse width to 240-250μs, and the number of training sessions to 20-22 times per set. When the patient's current rehabilitation stage is chronic and the assessment score is low, set the current amplitude to 18-20mA, the frequency to 30Hz, the pulse width to 240μs, and the number of training sessions to 18-20 times per set. When the patient's current rehabilitation stage is chronic and the assessment score is at a moderate level, the current amplitude is set to 22-24mA, the frequency to 35Hz, the pulse width to 250μs, and the number of training sessions to 22 times per set. When the patient is currently in the chronic phase of rehabilitation and has a high assessment score, set the current amplitude to 24-26 mA, the frequency to 40 Hz, the pulse width to 260-280 μs, and the number of training sessions to 25 per set.

5. A fine motor rehabilitation control system based on multimodal signal fusion, characterized in that, include: A multimodal signal acquisition module is used to acquire multimodal signals, including: pressure distribution signals of fine hand movements, surface electromyography signals, and electroencephalography signals. The data processing module is used to preprocess, extract features, and normalize the collected pressure distribution signals, surface electromyography signals, and electroencephalogram (EEG) signals to obtain pressure feature values, surface electromyography feature values, and EEG feature values, respectively. The rehabilitation assessment module is used to calculate the conversion coefficient of the current rehabilitation stage based on the number of days the patient's fine motor function is impaired, and to calculate the weights of pressure characteristic values, surface electromyography characteristic values, and electroencephalography characteristic values ​​based on the conversion coefficient; it includes the following steps: A0. Determine the number of days the patient's fine motor function was impaired. ; A1. Based on the number of damaged days Calculate the conversion factor for the current rehabilitation stage, the conversion factor including coefficients and sum coefficient; Calculate using the following method coefficient: Calculate using the following method coefficient: in, This indicates the number of days the patient's fine motor function was impaired. This indicates that the current recovery phase is the acute phase. This indicates that the current recovery phase is the recovery period. This indicates that the current recovery phase is the chronic phase; A2. Based on the number of damaged days Determine the initial weights for pressure characteristic values, surface electromyography characteristic values, and electroencephalography characteristic values, and based on... coefficients and sum The coefficients are used to calculate the weights of pressure characteristic values, surface electromyography characteristic values, and electroencephalography characteristic values. The evaluation score is obtained by weighted summation of the pressure characteristic values, surface electromyography (EMG) characteristic values, and electroencephalography (EEG) characteristic values. The pressure characteristic values ​​include the maximum pressure value, the minimum pressure value, and the trend quantification value. The surface EMG characteristic value is the root mean square (RMS) value of the EMG signal. The EEG characteristic values ​​include the RMS value of the EEG signal and event-related desynchronization. The evaluation score is calculated using the following method: in, Indicates the assessment score. This indicates the number of days the patient's fine motor function was impaired. Indicates the number of days damaged. The weights of time pressure eigenvalues Indicates the number of days damaged. The weights of surface electromyographic features, Indicates the number of days damaged. Weights of EEG feature values Indicates the maximum pressure. Indicates the minimum pressure. Indicates the trend quantification value. , , These represent the weights of the maximum pressure value, the minimum pressure value, and the trend quantification value, respectively. ; This represents the root mean square value of the surface electromyography signal. This represents the root mean square value of the electroencephalogram (EEG) signal. Indicates that the event is related to synchronization. , and represent the root mean square value of the EEG signal and the weight of event-related desynchronization, respectively. ; The feedback and adaptive control module adjusts rehabilitation training parameters based on assessment scores and the number of days the patient's fine motor function is impaired. The rehabilitation training parameters are the current amplitude, frequency, pulse width, and number of training sessions under different level scores.

6. The fine motor rehabilitation control system based on multimodal signal fusion according to claim 5, characterized in that, The multimodal signal acquisition module includes a pressure distribution signal acquisition unit, a surface electromyography signal acquisition unit, and an electroencephalogram (EEG) signal acquisition unit. The pressure distribution signal acquisition unit includes an array-type flexible pressure sensor subunit, a pressure distribution signal processing subunit, and a host computer. The array-type flexible pressure sensor subunit includes a hand grip sensor array and a sleeve sensor array. The hand grip sensor array includes high-density independent sensitive points with a center-to-center spacing of 0.5–2 mm between adjacent sensitive points and a scanning frequency of 50–500 Hz. The sleeve sensor array includes high-density independent sensitive points with a center-to-center spacing of 0.5–1 mm between adjacent sensitive points and a scanning frequency of 50–500 Hz. The array-type flexible pressure sensor subunit and the pressure distribution signal processing subunit are connected via a flexible circuit board. The pressure distribution signal processing subunit uses a sliding differential algorithm to eliminate static errors introduced by sensor substrate deformation, filters the pressure data, and then transmits it to the host computer. The host computer calculates the pressure value of each sensitive point, maps the pressure value of each sensitive point to a standard anatomical coordinate system, and generates a pressure heatmap in real time to display the pressure distribution in different areas of the hand. The surface electromyography (EMG) signal acquisition unit uses an 8-channel EMG acquisition wristband to acquire surface EMG signals. The electroencephalogram (EEG) signal acquisition unit uses saline electrodes to acquire EEG signals.

7. The fine motor rehabilitation control system based on multimodal signal fusion according to claim 6, characterized in that, The hand grip sensor array includes 4,000 independent sensitive points, arranged in an 80-row × 50-column configuration; the sleeve sensor array includes 1,000 independent sensitive points, arranged in a 40-row × 25-column configuration.

Images