CN-121999969-A - Facial paralysis physiotherapy system based on brain electricity feedback

CN121999969ACN 121999969 ACN121999969 ACN 121999969ACN-121999969-A

Abstract

The invention discloses a facial paralysis physiotherapy system based on electroencephalogram feedback, which relates to the technical field of medical rehabilitation equipment and comprises an electroencephalogram data acquisition and processing module, a characteristic extraction module, a model training module and a physiotherapy adjustment module, wherein the electroencephalogram data acquisition and processing module is used for acquiring an original electroencephalogram signal and then converting the original electroencephalogram signal into a digital signal which can be used for wireless transmission, the characteristic extraction module is used for carrying out characteristic extraction on the electroencephalogram digital signal to obtain a corresponding electroencephalogram signal characteristic vector, the model training module is used for optimizing a support vector machine to train to obtain an electroencephalogram signal classification model by using a multi-universe optimization algorithm based on historical data of relevance of the electroencephalogram signal and tolerance, and the physiotherapy adjustment module is used for obtaining feedback information about time of a tester for generating tolerance in an acupuncture massage process through the electroencephalogram signal classification model according to the feedback information and adjusting current physiotherapy parameters according to the feedback information. The invention solves the problem of poor physiotherapy effect of the facial paralysis physiotherapy system in the prior art.

Inventors

TAO YAQIN
WANG SHANCHENG
XIONG ZICHEN

Assignees

南昌耀光科技有限公司

Dates

Publication Date: 20260508
Application Date: 20251203

Claims (10)

1. Facial paralysis physiotherapy system based on brain electrical feedback, characterized in that it comprises: the electroencephalogram data acquisition and processing module is used for converting the original electroencephalogram signals into digital signals which can be used for wireless transmission after acquiring the original electroencephalogram signals and uploading the electroencephalogram digital signals to the cloud server end; The feature extraction module is used for carrying out feature extraction on the brain electrical digital signals uploaded to the cloud server to obtain corresponding brain electrical signal feature vectors; The model training module is used for obtaining an electroencephalogram signal classification model by optimizing support vector machine training through a multi-universe optimization algorithm based on historical data of correlation of electroencephalogram signals and tolerance; the physiotherapy adjusting module is used for obtaining feedback information about the moment of generating tolerance of a tester in the acupuncture massage process through an electroencephalogram signal classification model according to the electroencephalogram signal feature vector, and adjusting current physiotherapy parameters according to the feedback information.
2. The facial paralysis physiotherapy system based on brain electrical feedback according to claim 1, wherein the step of converting the original brain electrical signal into a digital signal usable for wireless transmission after the original brain electrical signal is collected, and uploading the brain electrical digital signal to a cloud server comprises: The original electroencephalogram signals are acquired by a single-channel circuit based on TGAM to obtain electroencephalogram digital signals, and the acquired electroencephalogram digital signals are transmitted to a cloud server based on an HC-05 Bluetooth module.
3. The facial paralysis physiotherapy system based on brain electrical feedback of claim 2, wherein the step of extracting features of the brain electrical digital signal uploaded to the cloud server to obtain a corresponding brain electrical signal feature vector comprises: carrying out wavelet packet decomposition on the brain electrical digital signals, and then reconstructing the decomposed node signals; Processing the electroencephalogram digital signal after wavelet packet decomposition and reconstruction by using a nonlinear analysis method dispersion entropy to obtain a fuzzy dispersion entropy characteristic; extracting the airspace characteristics of the electroencephalogram digital signals by utilizing a co-space mode to obtain airspace characteristic vectors; Integrating the obtained fuzzy dispersion entropy features with the airspace feature vectors to construct input feature vectors: the input feature vector is standardized to obtain an electroencephalogram signal feature vector: Wherein, the The fuzzy dispersion entropy features and the spatial domain feature vectors are respectively.
4. The facial paralysis physiotherapy system based on brain feedback of claim 3, wherein the step of performing wavelet packet decomposition on the brain electrical digital signal and then reconstructing the decomposed node signal comprises: wavelet packet decomposition is carried out on the electroencephalogram digital signal, binary tree nodes obtained by wavelet packet transformation and decomposition are set as (j, n), j is the number of decomposed layers, n represents the number of nodes of the layer, and the decomposition coefficient of the electroencephalogram digital signal at the layer point is as follows: Reconstructing the decomposed node signals, wherein a wavelet packet reconstruction formula of a kth node of the j-1 layer is as follows: after obtaining each node frequency, the node frequencies of mu and beta rhythms are related, db4 is selected as a wavelet basis function, and signals are reconstructed; Wherein, the Wavelet packet coefficients for the j-th layer, 2 n-th subband, k being the time index of the coefficient, Coefficients of a low-pass analysis filter, belonging to a filter bank of wavelet transforms, for extracting low-frequency characteristics of signals, The wavelet packet coefficient of the nth sub-band of the j-1 layer, l is the time index of the coefficient of the layer, which is used as the input coefficient of the decomposition of the current layer, The representation sums up the integer l, Wavelet packet coefficients for the j-th layer 2n+1 th subband, Is a coefficient of a high-pass divider Jie Lvbo and is used for extracting high-frequency characteristics of signals.
5. The facial paralysis physiotherapy system based on electroencephalogram feedback according to claim 4, wherein said step of processing the electroencephalogram digital signal after wavelet packet decomposition and reconstruction by using the nonlinear analysis method dispersion entropy to obtain a fuzzy dispersion entropy feature comprises: the wavelet packet is used for decomposing and reconstructing the brain electrical digital signal to be an original time sequence, and the original time sequence is mapped to the original time sequence by using a normal distribution function Then, the linear transformation pair is carried out Mapping to ; Solving for And find each scatter pattern probability, and then calculate normalized fuzzy scatter entropy to obtain fuzzy scatter entropy features.
6. The facial paralysis physiotherapy system of claim 5, wherein the expression of the normal distribution function is: Wherein, the As a result of the original time series, For the length of the time series, μ and Representing the expectation and the variance respectively, Is natural logarithm; Wherein, the Is the quantization level.
7. The facial paralysis physiotherapy system based on brain feedback of claim 6, wherein said calculation And solving for each dispersion pattern probability, and then calculating normalized fuzzy dispersion entropy to obtain fuzzy dispersion entropy characteristics: The single scattering mode is ; Each scatter pattern probability is calculated: calculating normalized fuzzy dispersion entropy: Wherein, the Is that The class corresponding to m values of d points apart in the sequence, In order to quantify the level of quantization, As a result of the normalization factor, For a particular scattering mode The probability of occurrence in all embedded vectors, m is the embedding dimension, d is the delay factor, Is the length of the time series.
8. The facial paralysis physiotherapy system based on brain electrical feedback of claim 3, wherein said step of extracting spatial features of brain electrical digital signals by using a co-space mode to obtain spatial feature vectors comprises: Finding an optimal spatial filter through matrix diagonalization by utilizing a co-spatial mode algorithm, wherein the optimal spatial filter is used for maximizing variance difference of two target signals which are tolerant and intolerant, so that the signal distinction is higher; And integrating the screened high-discrimination feature vectors to form a feature matrix so as to obtain a space domain feature vector.
9. The facial paralysis physiotherapy system based on electroencephalogram feedback according to claim 1, wherein the step of obtaining an electroencephalogram classification model by optimizing support vector machine training using a multivariate universe optimization algorithm based on historical data of correlation of electroencephalogram signals and tolerability comprises: Collecting brain electrical signals of a preset number of testers in the acupuncture massage process, and synchronously labeling corresponding tolerance states to form a training data set; denoising the original electroencephalogram signals, extracting features, and obtaining feature vectors which can be input by the model; feeding a support vector machine model framework by using a training data set, searching optimal C and gamma parameters by using a multi-universe optimization algorithm, dividing the training set/testing set to verify the model performance, and performing iterative optimization to obtain a final electroencephalogram classification model; And deploying the trained electroencephalogram signal classification model to a cloud server.
10. The facial paralysis physiotherapy system based on electroencephalogram feedback of claim 9, wherein the step of obtaining the electroencephalogram classification model by optimizing support vector machine training using a multivariate universe optimization algorithm based on the history data of the correlation of electroencephalogram signals and tolerability further comprises: setting a multi-element universe optimization algorithm parameter and a C parameter and gamma parameter range of a support vector machine; Initializing a universe position; Calculating the expansion rate of each universe to execute a roulette selection mechanism to select white holes; updating WEP and TDR values and searching for an optimal individual; Judging whether the optimal universe or the maximum iteration number is reached; If yes, outputting the optimal C, g value to construct an electroencephalogram signal classification model; if not, the step of initializing the universe position is re-executed.

Description

Facial paralysis physiotherapy system based on brain electricity feedback Technical Field The invention relates to the technical field of medical rehabilitation equipment, in particular to a facial paralysis physiotherapy system based on brain electrical feedback. Background Electric acupuncture and massage are one of the most effective physical therapies for treating facial paralysis clinically at present. The two therapies show remarkable curative effects in a plurality of medical fields such as nervous system diseases, osteoarthropathy, endocrine disorders, respiratory diseases and the like by directly stimulating acupoints and deep tissues. The traditional acupuncture adopts hand needle therapy initially, but has the problems of high operation technical requirements, easy occurrence of needle sickness, broken needle and the like. Along with technological progress, the electric acupuncture technology gradually develops and matures, and the electric acupuncture technology simulates silver acupuncture stimulation through current pulses and is conducted to target acupuncture points along channels and collaterals, so that not only is the treatment effect improved, but also flexible adjustment of the stimulation mode and intensity is realized, and obvious technical advantages are shown. Compared with the traditional mechanical equipment, the modern electric massage device has the characteristics of simple and convenient operation, various modes and the like, and is more suitable for the masses. Physiological studies have shown that the presence of specific responses of human receptors to different stimuli produces an adaptive modulation of the intensity and duration of the stimuli. However, sustained single stimulation results in reduced sensory nerve responsiveness and reduced microcirculation rates, a phenomenon known as "stimulus tolerance". Once tolerance is established, continued use of the same stimulation will significantly reduce the physiotherapy effect. The current multifunctional meridian therapeutic apparatus on the market has a plurality of physiotherapy modes, but still lacks the capability of monitoring tolerance in real time. Thereby causing a problem of poor physiotherapy effect. Disclosure of Invention Therefore, the invention aims to provide a facial paralysis physiotherapy system based on brain electrical feedback, and aims to solve the problem that the facial paralysis physiotherapy system in the prior art is poor in physiotherapy effect. In one aspect, the invention provides a facial paralysis physiotherapy system based on brain electrical feedback, which comprises: the electroencephalogram data acquisition and processing module is used for converting the original electroencephalogram signals into digital signals which can be used for wireless transmission after acquiring the original electroencephalogram signals and uploading the electroencephalogram digital signals to the cloud server end; The feature extraction module is used for carrying out feature extraction on the brain electrical digital signals uploaded to the cloud server to obtain corresponding brain electrical signal feature vectors; The model training module is used for obtaining an electroencephalogram signal classification model by optimizing support vector machine training through a multi-universe optimization algorithm based on historical data of correlation of electroencephalogram signals and tolerance; the physiotherapy adjusting module is used for obtaining feedback information about the moment of generating tolerance of a tester in the acupuncture massage process through an electroencephalogram signal classification model according to the electroencephalogram signal feature vector, and adjusting current physiotherapy parameters according to the feedback information. Further, in the facial paralysis physiotherapy system based on electroencephalogram feedback, the step of converting the original electroencephalogram signal into a digital signal which can be used for wireless transmission after the original electroencephalogram signal is acquired and uploading the electroencephalogram digital signal to the cloud server end comprises the following steps: The original electroencephalogram signals are acquired by a single-channel circuit based on TGAM to obtain electroencephalogram digital signals, and the acquired electroencephalogram digital signals are transmitted to a cloud server based on an HC-05 Bluetooth module. Further, in the facial paralysis physiotherapy system based on electroencephalogram feedback, the step of extracting features of the electroencephalogram digital signals uploaded to the cloud server to obtain corresponding electroencephalogram feature vectors comprises the steps of: carrying out wavelet packet decomposition on the brain electrical digital signals, and then reconstructing the decomposed node signals; Processing the electroencephalogram digital signal after wavelet packet decomposition and re