CN108836328A - A kind of equipment for being monitored for electromyography signal and generating corresponding electrical stimulation signal - Google Patents

Abstract

A device for monitoring electromyographic signals and generating corresponding electrical stimulation signals, including: monitoring and stimulating equipment, card controllers, and data processing and display devices, wherein the monitoring and stimulating equipment adopts a band-aid design and has muscle An electrical signal acquisition module and one or more electrical stimulation modules, the one or more electrical stimulation modules are used to generate one or more electrical discharge stimuli; the data processing and display device can analyze various characteristics of the myoelectric signal Extraction is performed to obtain multiple myoelectric signal features. The present invention obtains multiple features of the electromyographic signal and fuses the multiple features of the electromyographic signal, so that it can perform a more comprehensive analysis of various muscle states; in addition, the stimulation waveform in the present invention consists of square wave and sine It is composed of two types of waves, and has low frequency and intermediate frequency, so that it can produce a variety of electrical stimulation schemes; and by using biphasic electrical stimulation, it is safer and more versatile.

Description

A device for monitoring myoelectric signals and generating corresponding electrical stimulation signals

technical field

The invention relates to the field of medical equipment, in particular to a device for monitoring myoelectric signals and generating corresponding electrical stimulation signals, which is used for analysis of myoelectric state and generation of electrical stimulation signals.

Background technique

Muscle is the main source of the human body's exercise power, and it is widely distributed in the human body, accounting for about 40% of the total body weight. When in the EMG state, the overall mental state and motion perception function of the human body will gradually decline, manifested as distraction, slowed operation speed, reduced coordination and flexibility of movements, and will eventually weaken the working ability of the entire human body and increase errors. And the incidence of accidents, resulting in a reduction in the efficiency of completing tasks. When the muscles are severely fatigued, it will even affect the normal work and life order of the human body. However, as the pace of life becomes faster, many white-collar workers, students, and IT practitioners are prone to neck and shoulder pain due to long-term desk work. Some passenger or freight drivers often cause shoulder and waist muscle discomfort due to long-term driving. Athletes and sports enthusiasts It can also cause myoelectric state and damage in various parts of the body due to excessive exercise and exercise. According to statistics, in the process of high-load work and unscientific exercise methods, the myoelectric state and pain symptoms caused by unfavorable factors such as fixed sedentary, poor posture, and excessive activities are increasing year by year.

At present, there are more mature myoelectric state monitoring technologies in the market, such as surface electromyography (SEMG) technology, which can objectively reflect the level of muscle activity and functional state, and has the characteristics of accurate operation and repeatable measurement and monitoring, and realizes non-invasive , Real-time, quantitative assessment and measurement of myoelectric state. However, most products using this technology have single functions, fixed electrical stimulation waveforms, and fixed electrical stimulation schemes, which make them unable to meet the requirements of different users and have poor versatility. The specific disadvantages are as follows:

(1) Mainly based on the independent channel mode, with a single function and a small scope of use, it is often limited to specific myoelectric state monitoring, or is limited to the discharge of a single electrical stimulation signal, with low integration and weak expansion.

(2) Part of the myoelectric state monitoring and electrical stimulation mechanisms are too large, expensive, and complex in structure, requiring professionals to operate, limited in use places, difficult to wear, and unsuitable for large-scale practical applications.

(3) Not based on the user's own state and the state of local myoelectricity, the user cannot adjust the electrical stimulation signal by himself through an adaptive algorithm according to his actual situation, resulting in poor or harmful effects.

Contents of the invention

In view of the above problems, the present invention discloses a device for monitoring electromyographic signals and generating corresponding electrical stimulation signals, including: monitoring and stimulation equipment, a card controller, and a data processing and display device; wherein, the monitoring and stimulation The equipment adopts a band-aid design, including a myoelectric signal acquisition module and an electrical stimulation module, and the electrical stimulation module is used to generate one or more discharge stimuli; The features are extracted to obtain a plurality of myoelectric signal features; the monitoring and stimulation equipment and the data processing and display device communicate with the card controller through wireless transmission respectively.

Preferably, the multiple discharge stimuli include: low-frequency discharge stimulation and medium-frequency discharge stimulation.

Preferably, the various features include: time domain features, frequency domain features, dual frequency domain features, and time-frequency domain features.

Preferably, the monitoring and stimulation equipment includes: surface myoelectric electrodes, surface array electrical stimulation electrodes, myoelectric signal feedback display module, first wireless communication module and battery, wherein the surface myoelectric electrodes and the surface array electrical electrodes The stimulating electrodes adopt an array design, which are respectively used for the collection of myoelectric signals and the discharge of electrical stimulation signals.

More preferably, the myoelectric state feedback display module includes: a buzzer element and an LED display light element, and the myoelectric state feedback display module is used to feedback and display the state information contained in the myoelectric signal through the combination of sound and light .

Preferably, the card-type controller includes: a myoelectric signal preprocessing module, a second communication module and a control module, wherein the myoelectric signal preprocessing module and the second communication module are respectively connected to the control module.

More preferably, the electromyographic signal preprocessing module includes: an acquisition circuit, an amplifier circuit, a filter circuit, and an A/D conversion circuit, which are respectively used for receiving, amplifying, and denoising the electromyographic signal, wherein the A/D conversion The circuit is used for analog-to-digital or digital-to-analog conversion of myoelectric signals or digital-to-analog or digital-to-analog conversion of said electrical stimulation signals.

Preferably, the data processing and display device includes:

Myoelectric signal off-line data training module, for extracting the eigenvalue in the described myoelectric signal, normalizing the value of the described myoelectric signal and selecting the classification function of the described myoelectric signal;

An electromyographic signal processing module, configured to extract the eigenvalues in the electromyographic signals, and classify the state information in the electromyographic signals according to the classification function;

An electrical stimulation module, configured to generate electrical stimulation signals based on waveform, frequency, pulse width and amplitude;

The human-computer interaction module is used to generate myoelectric signals and display information of electrical stimulation signal waveforms or to set different electrical stimulation schemes;

The database module is used to collect and store the myoelectric signals within a period of time, the state information in the myoelectric signals, the selection of the classification function and the classification result according to the classification function or the electrical stimulation scheme.

More preferably, the classification function includes: a support vector machine classification function.

More preferably, the waveform-based electrical stimulation signals generated in the electrical stimulation module include: sine waves and square waves.

The advantage of the present invention is that it has the integrated function of myoelectric signal monitoring and electrical stimulation, and can perform multiple feature extraction on the collected myoelectric signal, thereby obtaining multiple myoelectric signal features, from the time domain, frequency domain, and dual frequency domain. Combine multiple myoelectric signal features from multiple angles in the time-frequency domain; different states of muscles (muscle fatigue state, muscle strength increase state, muscle strength decline state, muscle strength recovery state) and large muscle movements (such as running ) and fine movements (such as typing on the keyboard) can be analyzed more accurately; the self-adaptive learning method is used to learn offline the past myoelectric characteristics and muscle fatigue status of the individual, so that the device can be targeted, and the individual Muscle state differences between individuals and individuals in different states are reloaded, combined with the online classification results of the current personal EMG characteristics, to form the final result of dynamic muscle fatigue classification, thereby generating corresponding electrical stimulation signals. The stimulation waveform of the electrical stimulation signal is composed of square wave and sine wave, which can be divided into low frequency and intermediate frequency, and adopts biphasic electrical stimulation, which makes it have a variety of electrical stimulation schemes and is safer and more versatile. powerful.

In addition, the device of the present invention is composed of three major components, including: monitoring and electrical stimulation equipment, a card controller, and a data processing and display device. The wireless communication module connects the monitoring and electrical stimulation equipment and the data processing and display device, which reduces the volume of the monitoring and electrical stimulation equipment on the one hand, making it more comfortable and safe to wear, and on the other hand, due to the data processing and display The device is a device such as a smart phone or a Pad installed with an APP corresponding to the device of the present invention, so that the cost of the device of the present invention is lower.

Description of drawings

Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the specific embodiments. The drawings are only for the purpose of illustrating specific embodiments and are not to be considered as limiting the invention. Also throughout the drawings, the same reference numerals are used to designate the same components. In the attached picture:

Fig. 1 is the device composition diagram of the present invention.

Fig. 2 is a flow chart of the control method of the present invention.

Fig. 3 is a working flow chart of the present invention.

Fig. 4 is a flowchart of the electrical stimulation of the present invention.

Fig. 5 is a flow chart of myoelectric signal acquisition and feature extraction in the present invention.

Detailed ways

Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided for more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

The invention discloses a device for monitoring myoelectric signals and generating corresponding electrical stimulation signals, comprising: monitoring and electrical stimulation equipment, a card controller, and a data processing and display device. The monitoring and electrical stimulation equipment described in the present invention adopts a band-aid design, which can collect myoelectric signals for analysis of myoelectric state, and generate corresponding electrical stimulation signals according to said myoelectric state. In order to reduce the The size of the electrical stimulation equipment makes it more comfortable to wear. In the present invention, the myoelectric signal processing module and control module are concentrated on the card controller other than the monitoring and electrical stimulation equipment, and communicate with the electrical stimulation equipment through the wireless communication module. The card controller communicates with the above-mentioned card controller, and then obtains the electrical stimulation signal, in this way, the volume of the monitoring and electrical stimulation equipment is smaller, and it is more comfortable to wear and use; the card-type controller preprocesses the electromyography signal, And the preprocessed myoelectric signal is sent to the data processing and display device for further processing and analysis; the data processing and display device can use mobile devices such as mobile phones and Pads, by installing the APP corresponding to the device of the present invention, and then The electromyographic signal is further processed, and the further processing is such as performing feature extraction on the electromyographic signal through various feature extraction methods, thereby obtaining a plurality of electromyographic signal features; ) classification method classifies the electromyographic state according to the electromyographic signal characteristics within a period of time; finally, the data processing and display device sends corresponding control commands according to the electromyographic state classification result and the various electromyographic signal characteristics, thereby Various electrical stimulation signals are generated by monitoring and electrical stimulation equipment. The above-mentioned multiple myoelectric signal features are peak values of multiple myoelectric signals, and said myoelectric feature is that by said myoelectric signal feature (peak value), the value of multiple myoelectric signal peak values is graded according to a certain range after the result. The equipment and method of the present invention will be described in more detail below in conjunction with the accompanying drawings;

As shown in FIG. 1 , it is a composition diagram of the equipment of the present invention. The equipment of the present invention includes: monitoring and electrical stimulation equipment, a card controller, and a data processing and display device. Among them, the monitoring and electrical stimulation equipment is used to monitor the myoelectric signal or generate a corresponding electrical stimulation signal according to the myoelectric state in the myoelectric signal, and the card controller has a preprocessing function and acts as a hub , for connecting the monitoring and electrical stimulation equipment and the data processing and display device.

Specifically, the monitoring and electrical stimulation device includes: a surface myoelectric electrode, a surface array electrical stimulation electrode, a myoelectric state feedback display module, a first wireless communication module, and a battery. Wherein, the surface myoelectric electrodes are used to collect myoelectric signals and send them to the card control device through the first wireless communication module, and the surface array electrical stimulation electrodes are used to receive signals from the first wireless communication module. V. The electrical stimulation signal of the card controller, thereby generating electrical stimulation pulses. In addition, the myoelectric state feedback display module includes: a buzzer element and an LED display light element, which are used to feedback and display the current myoelectric state through the combination of sound and light. In the present invention, there are three types of myoelectric state classification , which are respectively displayed by three kinds of LED display lamp components of red, yellow and green, accompanied by buzzer alarm sounds of different frequencies.

The card controller includes: a myoelectric signal preprocessing module, a second communication module and a control module, wherein the myoelectric signal preprocessing module and the second communication module are respectively connected to the control module. More specifically, the EMG signal preprocessing module includes: an acquisition circuit, an amplifier circuit, a filter circuit, and an A/D conversion circuit, which are respectively used to receive and amplify the EMG signal under the control of the control module. , denoising and analog-to-digital conversion operations, and digital-to-analog conversion functions for electrical stimulation signals; the functions of the control module also include: initialization functions, such as device clock initialization, port initialization, timer initialization, and other functions, such as detecting electrodes Resistance value, set the electrode position, set the default electrical stimulation program after starting up, and normalize the collected EMG signal features from 0 to 1.

The data processing and display device is a smart phone, a Pad or a computer having an APP corresponding to the equipment of the present invention, and the APP corresponding to the equipment of the present invention includes: an off-line data training module for myoelectric signals, which is used to analyze the characteristics of the myoelectric signals Extracting the value, normalizing the value of the electromyographic signal and selecting the classification function of the electromyographic signal; the electromyographic signal processing module is used to extract the feature value in the electromyographic signal , and classify the state information in the myoelectric signal according to the classification function; the electrical stimulation module is used to generate the electrical stimulation signal based on waveform, frequency, pulse width and amplitude; the human-computer interaction module is used to generate the myoelectric signal Signal and electrical stimulation signal waveform display information or set different electrical stimulation programs;

The database module is used to collect and store the myoelectric signals within a period of time, the state information in the myoelectric signals, the selection of the classification function and the classification result according to the classification function or the electrical stimulation scheme.

Specifically, on the one hand, in the present invention, various features in the electromyographic signal are extracted through an offline data training module to obtain a plurality of electromyographic signal features, and then the support vector machine classification method is used to classify the electromyographic signal features Carry out classification, obtain electromyographic state classification, wherein, described multiple electromyographic signal characteristic is the peak value of corresponding electromyographic signal in different feature extraction methods, and described electromyographic state is the division result according to the peak range of described electromyographic signal . On the other hand, in the present invention, feature extraction is carried out to described electromyographic signal in real time through electromyographic signal processing module, and obtain real-time electromechanical signal feature, the classification result of electromyographic state in described off-line data training module is described real-time The myoelectric signal is classified to determine the myoelectric state where the real-time myoelectric signal is located, so as to generate different electrical stimulation schemes according to different myoelectric states. The various feature extractions mentioned above include: time-domain feature extraction, frequency-domain feature extraction, dual-frequency-domain feature extraction, and time-frequency domain feature extraction. In the present invention, the electrical stimulation signal is generated by the electrical stimulation module, and various electrical stimulation schemes are obtained by generating electrical stimulation signals whose waveform, frequency, pulse width and amplitude vary with time. In the above, the generation frequency The electrical stimulation signals that vary with time include: low-frequency electrical stimulation and medium-frequency electrical stimulation, and the electrical stimulation signals that generate waveforms that vary with time include: sine waves and square waves. The forms of electrical stimulation in the prior art include two forms of stimulation, one is monophasic electrical stimulation and the other is bidirectional electrical stimulation, but since the direction of the current of monophasic stimulation is single, the monophasic pulse it produces The current has a direct current component, which may have a certain electrolytic phenomenon on the target of the electrical stimulation, so the electrical stimulation in the present invention adopts the form of biphasic electrical stimulation. In biphasic electrical stimulation, one phase is used for stimulation, and the other phase is used to balance the charge accumulation generated by the electrical stimulation phase, thereby greatly reducing the generation of electrolysis. The human-computer interaction module in the present invention is used for the display and modification of the electrical stimulation scheme and the display and processing of the electromyography signal.

When the human-computer interaction module is used to modify the electrical stimulation scheme, the user judges whether the electrical stimulation generated by the current electrical stimulation scheme is satisfied with the current electrical stimulation scheme or feeds back the new electric stimulation scheme to the offline data training module, The off-line data training module trains the classifier (support vector machine) again according to the new idea electrical stimulation scheme, retriggers the electrical stimulation module to set the waveform, frequency, pulse width or amplitude, thereby revises the electrical stimulation scheme, and transfers the modified electrical stimulation Stimulus programs are stored in the database module for the next call. While training offline, the data processing and display device also processes the myoelectric signal in real time to obtain multiple myoelectric signal features of the myoelectric signal, and then the data processing and display device will The electrical state classification classifies the multiple myoelectric signal features, so as to update the generation timing of electrical stimulation in the new electrical stimulation scheme. The database module in the present invention is also used for storing information such as the electromyographic signal information within at least 12 hours, the classification result of the classifier, and the electrical stimulation scheme adopted.

As shown in FIG. 2 , it is a flowchart of the control method of the present invention. The method includes: step 001, monitoring and electrical stimulation equipment collects myoelectric signals, and sends them to the card controller; step 002, the card controller receives the myoelectric signals, preprocesses them and sends them to the The data processing and display device; step 003, the data processing and display device further processes the electromyographic signal, and judges whether the current electromyographic signal feature is greater than the electrical stimulation threshold; step 004, if the electromyographic signal feature is greater than the set threshold, Then the data processing and display device sends a control command to the card controller; step 014, if the myoelectric signal feature is less than or equal to the set threshold, then continue to maintain the monitoring state; step 005, the card controller according to the control Command to send the corresponding electrical stimulation signal to the monitoring and electrical stimulation equipment; step 006, the monitoring and rehabilitation equipment discharges through the electrical stimulation electrodes according to the electrical stimulation signal, and sends the myoelectric signal to the card controller in real time. The present invention will be further described below in conjunction with the work flow chart of the present invention:

As shown in Figure 3, it is a work flow diagram of the present invention. The working process of the device of the present invention is as follows. The card controller initializes each module under the control of the control module; signal, and send it to the card-type controller through the first wireless communication module, and the card-type controller performs preprocessing on the electromyography signal through the preprocessing module, such as collecting, amplifying, and denoising and analog-to-digital conversion, and then the card-type controller sends the preprocessed myoelectric signal to the data processing and display device for off-line data training analysis and myoelectric signal processing; wherein, the off-line data training analysis is through Extract the myoelectric signal features of the myoelectric signal within a period of time, and after obtaining a plurality of myoelectric signal features, use the support vector machine to classify the myoelectric signal features, obtain the classification of the myoelectric state, and determine the source of electrical stimulation. The threshold value of the EMG signal feature, this process can be triggered again through the human-computer interaction module, thereby modifying the electrical stimulation scheme and the threshold value; the EMG signal processing is to obtain a specific EMG signal through the EMG signal Signal feature (the peak of myoelectric signal under different feature extraction methods); according to the classification of myoelectric state obtained in the offline data analysis, the real-time myoelectric signal is carried out to myoelectric state classification, and determine the electrical stimulation threshold, if said myoelectric state classification If the characteristic value of the electromyographic signal is less than or equal to the asynchronous control threshold, the electromyographic signal is returned for collection, and the equipment continues to maintain the monitoring state; if the characteristic value of the electromyographic signal is greater than the asynchronous control threshold, the data processing and display device Send a control command to the card-type controller, the card-type controller generates an electrical stimulation signal according to the control command, and sends it to the monitoring and electrical stimulation device, and finally the array in the monitoring and electrical stimulation device The electrical stimulation electrodes are discharged according to the electrical stimulation signal; after that, the user can check whether the electrical stimulation at this time meets the user's requirements, and if not, the electrical stimulation can be controlled by the human-computer interaction function in the data processing and display device. The scheme is adjusted, and the adjustment method is to change the SVM classification function in the offline data training analysis, reclassify the myoelectric state and re-determine the asynchronous control threshold. If yes, the electrical stimulation control command is continued, and the device of the present invention also has a timing function, and automatically ends the electrical stimulation discharge when the timing time is reached. In the above-mentioned process, the device of the present invention will also store information such as electromyographic signal information, classifier classification results, and electrical stimulation schemes taken within 12 hours through the database module in the data processing and display device. The stored electrical stimulation scheme can be directly selected. In addition, the device of the present invention can also make statistics on the frequency of use of the electrical stimulation scheme, so that the commonly used scheme can be preferentially selected when the device is turned on next time.

As shown in Fig. 4, it is a flowchart of the electrical stimulation of the present invention. When a data processing and display device such as a mobile phone or a pad sends a control command to the card controller, the encoding module is stimulated to work, and then the decoding circuit is stimulated to work, and the timer is triggered to work at the same time. The control command includes The digital signal of the electrical stimulation signal and the specific control instruction, after the digital signal of the electrical stimulation signal is digital-to-analog converted by the myoelectric signal preprocessing module in the card controller, the low/intermediate frequency is modulated according to the control instruction The selection of the circuit is amplified by the power amplifier circuit at a certain frequency (low frequency or intermediate frequency), and the amplified electrical stimulation signal is displayed by the data processing and display device (mobile phone or pad) on the one hand; The wireless communication module sends to the monitoring and electrical stimulation equipment, so as to perform discharge stimulation through the array electrical stimulation electrodes.

As shown in FIG. 5 , it is a flow chart of myoelectric signal acquisition and feature extraction in the present invention. The myoelectric signal preprocessing module in the card-type controller obtains the myoelectric signal received by the second wireless communication module through the acquisition circuit in it, then amplifies it through the amplification circuit, and then filters and removes noise through the filter circuit, thereby After the preprocessed EMG signal is amplified by the amplifier and then filtered, the effective information in the signal can be better preserved. At this time, the EMG signal is still an analog signal, and it needs to be converted into an analog signal by a digital-to-analog conversion circuit. After the conversion, it is then sent to the data processing and display device, and feature extraction is performed through the muscle signal processing module. In the present invention, multiple feature extractions are performed on the myoelectric signal, thereby obtaining multiple myoelectric signal features, thereby according to multiple Each myoelectric signal feature evaluates the myoelectric state at this time, so that the monitoring of the myoelectric state by the device of the present invention is more accurate.

Among the above, the various feature extraction methods include:

(1) Time-domain feature extraction, including: root mean square amplitude (Root Mean Square, RMS), integrated EMG (Integrated EMG, iEMG).

Both the root mean square value and the integrated EMG value can reflect the variation characteristics of the sEMG signal amplitude in the time dimension. RMS represents the energy of the signal, which can determine the strength of the muscle force. RMS is the effective value of the discharge, which can describe the average change characteristics of the discharge over a period of time, and better reflect the discharge volume of a specific exercise for a period of time. Its characteristics are defined as:

Among them, |EMG(t)| is the time change function of the EMG curve; x _i is the i-th sampling sample of |EMG(t)|; N is the number of sampling points in one cycle; T is the test time length; t ₀ is the start time of the test, and the unit of RMS is mv.

Integrated EMG (Integrated EMG, iEMG) refers to the integration of the area contained between the EMG change curve and the time axis on the EMG, and the unit is mv s. The integral EMG reflects the intensity of muscle activity over a period of time, and its calculation formula is:

Among them, _xi is the sampling point of |EMG(t)|; _xi is the sampling point of |EMG(t)|, which is the test time length; T is the test time length; t ₀ is the start time of the test. Integral EMG can reflect changes in the strength of EMG signals over time. (2) Frequency domain feature extraction, feature parameters are: mean power frequency (Mean Power Frequency, MPF), median frequency (Median Frequency, MF).

The mean power (Mean Power Frequency, MPF) is the average value of the product of the frequency value and the corresponding spectral value on the entire power spectrum. Its definition is:

Where P(f) is the power spectral density obtained by the periodogram method of sEMG; f _mean is the average power frequency of the sEMG signal; f is the sampling frequency; f ₀ is the upper limit frequency of the power spectrum, The unit of MPF is Hz.

The median frequency (Median Frequency, MF) is the frequency that divides the power spectrum into two equal-area regions, which is defined as:

Wherein P(f) is the power spectral density obtained by the periodogram method of sEMG; f _med is the median frequency of sEMG signal; f is the sampling frequency; _f0 is the upper limit frequency of the power spectrum, The unit of MF is Hz. (3) Dual-frequency domain feature extraction, including: bispectrum sum (Bispectrum Sum, BS), bicoherence sum (Bicoherence Sum, BC), bispectrum variance (Bispectrum Variance, BSV) and bispectrum variance (Bispectrum Variance, BCV).

The frequency bispectrum analysis method can obtain the third order cumulant and bispectral function by performing Fourier transform on the third order cumulant, as follows:

In the formula, τ ₁ , τ ₂ are delays respectively; f ₁ , f ₂ are τ ₁ , τ ₌ correspondences in frequency domain respectively. The non-Gaussianity of a random process can be quantified using coefficients and estimated as follows:

In the formula, P(f ₁ ) is the power spectrum of the frequency domain part corresponding to the delay τ ₁ , and P(f ₂ ) is the same as P(f ₁ +f ₂ ). The BS, BC, BSV and BCV formulas derived from B(f ₁ ,f ₂ ) and BC(f ₁ ,f ₂ ) are respectively:

BS=∑|B(f ₁ ,f ₂ )|,

BC＝∑|BC(f ₁ ,f ₂ )|,

BSV＝E{[B(f ₁ ,f ₂ )-mean(B(f ₁ ,f ₂ ))] ² },

BCB＝E{[BC(f ₁ ,f ₂ )-mean(BC(f ₁ ,f ₂ ))] ² }(7)

(4) Time-frequency domain feature extraction, including: Short-time Fourier transform and spectrogram.

The features are classified using a support vector machine approach:

Support Vector Machine (SVM) is a machine learning method based on the statistical learning theory VC dimension theory (Vapnik-Chervonenkis Dimension) and the principle of structural risk minimization. Taking the classification of two types of data as an example, given a training sample set ( _xi ,y _i ), i=1,2,…,, x∈ R ⁿ , y∈{±1}, the hyperplane is denoted as (w·x )+b=0, in order to make the classification face all samples correctly classified and have a classification interval, it is required to meet the following constraints: y _i [(w x _i + b)]≥1, i=1, 2,..., . Among them, x _i represents the n-dimensional vector describing object i, y _i is the label of object i, and w represents the slope of the hyperplane (determines the angle of the hyperplane).

Since there is an inverse relationship between the classification interval and the slope w of the hyperplane line, which is 2/||w||, the problem of constructing the optimal hyperplane is transformed into finding the minimum slope w of the hyperplane line under the constraints, and b is a constant:

In order to solve this constrained optimization problem, the Lagrange function is introduced:

In the formula, a _i >0 is the Lagrange multiplier corresponding to the i-th object. Since the constrained optimization problem is determined by the saddle point obtained by the Lagrange function, and the solution of the optimization problem satisfies that the partial derivatives of w and b are 0 at the saddle point, the QP quadratic programming problem is transformed into the corresponding dual problem, namely:

In the formula, maxQ(a) represents the optimal function for obtaining the Lagrangian multiplier.

get the best solution

Calculate the optimal weight vector w ^* and the optimal bias b ^* , respectively:

In the formula, the subscript Therefore, the optimal classification hyperplane (w* x) + b ^* = 0 is obtained, where x is and the optimal classification function is:

This is the support vector machine.

The above is only an exemplary embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present invention. All should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims (10)

1. a kind of equipment for being monitored for electromyography signal and generating corresponding electrical stimulation signal, which is characterized in that including：Monitoring and thorn Swash equipment, card form controller and data processing and display device；Wherein, the monitoring uses adhesive bandage formula with stimulation apparatus Design, including, electromyographic signal collection module and electrical stimulation module, the electrical stimulation module is for generating one or more electric discharge thorns Swash；The data processing can extract the various features in electromyography signal with display device, obtain multiple electromyography signals Feature；The monitoring passes through wireless transmission method and the card with stimulation apparatus and the data processing and display device respectively The communication of formula controller.

2. equipment according to claim 1, which is characterized in that a variety of electric discharges, which stimulate, includes：Low frequency electric discharge stimulation with Intermediate frequency electric discharge stimulation.

3. equipment according to claim 1, which is characterized in that the various features include：Temporal signatures, frequency domain character, Double frequency characteristic of field and time and frequency domain characteristics.

4. equipment according to claim 1, which is characterized in that the monitoring with stimulation apparatus includes：Surface myoelectric electrode, Surface array electric stimulation electrode, electromyography signal feedback display module, the first wireless communication module and battery, wherein the surface Electromyographic electrode and the surface array electric stimulation electrode are designed using array, are respectively used to the acquisition and electro photoluminescence of electromyography signal The electric discharge of signal.

5. equipment according to claim 4, which is characterized in that the myoelectricity state feedback display module includes：Buzzer Element and LED show that modulation element, the myoelectricity state feedback display module are used to feed back display in such a way that acousto-optic combines The status information for including in electromyography signal.

6. equipment according to claim 1, which is characterized in that the card form controller includes：Electromyography signal pretreatment Module, second communication module and control module, wherein the electromyography signal preprocessing module and the second communication module point It is not connected with the control module.

7. equipment according to claim 6, which is characterized in that the electromyography signal preprocessing module includes：Acquisition Circuit, Amplifier circuit, filter circuit and A/D conversion circuit are respectively used to the reception, amplification, denoising of electromyography signal, wherein described A/D conversion circuit is for the modulus of electromyography signal or digital-to-analogue conversion or the digital-to-analogue or digital-to-analogue conversion of the electrical stimulation signal.

8. equipment according to claim 1, which is characterized in that the data processing includes with display device：

Electromyography signal off-line data training module, for being extracted to the characteristic value in the electromyography signal, to the myoelectricity The value of signal is normalized and selects the classification function of the electromyography signal；

Electromyography signal processing module, for being extracted to the characteristic value in the electromyography signal and according to the classification function Classify to the status information in electromyography signal；

Electrical stimulation module, for generating the electrical stimulation signal based on waveform, frequency, pulsewidth and amplitude；

Human-computer interaction module, for generating the display information of electromyography signal and electrical stimulation signal waveform or different electricity thorns being arranged Swash scheme；

Database module, for collecting and storing the electromyography signal in a period of time, the status information in electromyography signal, described point The selection of class function and classification results or the electrical stimulation scheme according to the classification function.

9. equipment according to claim 8, which is characterized in that the classification function includes：Support vector cassification function.

10. equipment according to claim 8, which is characterized in that the electricity based on waveform generated in the electrical stimulation module Stimulus signal includes：Sine wave and square wave.