CN109567799A - EMG Feature Extraction based on smooth small echo coherence - Google Patents

EMG Feature Extraction based on smooth small echo coherence Download PDF

Info

Publication number
CN109567799A
CN109567799A CN201811603107.XA CN201811603107A CN109567799A CN 109567799 A CN109567799 A CN 109567799A CN 201811603107 A CN201811603107 A CN 201811603107A CN 109567799 A CN109567799 A CN 109567799A
Authority
CN
China
Prior art keywords
wavelet
coherence
smooth
electromyography signal
small echo
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201811603107.XA
Other languages
Chinese (zh)
Inventor
席旭刚
杨晨
罗志增
张启忠
佘青山
林树梁
华仙
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hangzhou Dianzi University
Original Assignee
Hangzhou Dianzi University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hangzhou Dianzi University filed Critical Hangzhou Dianzi University
Priority to CN201811603107.XA priority Critical patent/CN109567799A/en
Publication of CN109567799A publication Critical patent/CN109567799A/en
Pending legal-status Critical Current

Links

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316—Modalities, i.e. specific diagnostic methods
    • A61B5/389—Electromyography [EMG]
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316—Modalities, i.e. specific diagnostic methods
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235—Details of waveform analysis
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235—Details of waveform analysis
    • A61B5/725—Details of waveform analysis using specific filters therefor, e.g. Kalman or adaptive filters
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235—Details of waveform analysis
    • A61B5/7253—Details of waveform analysis characterised by using transforms

Abstract

The invention discloses a kind of EMG Feature Extractions based on smooth small echo coherence, the present invention acquires the electromyography signal of human body related muscles by electromyographic signal collection instrument, and it is pre-processed using band-pass filtering method, wavelet transformation is carried out to the electromyography signal after filtering processing, then the cross wavelet analysis of two-way electromyography signal is calculated, and carries out the smooth operation on time shaft and scale axis respectively to cross wavelet analysis.Finally, calculating the smooth wavelet coherence of two-way electromyography signal, and examine the correlation between different behaviors with the presence or absence of statistical difference using T inspection, obtains 32 grades of smooth wavelet coherences of different muscle combinations as feature vector.The smooth small echo coherence that the present invention uses has very big innovation on feature extracting method, identifies discrimination with higher and reliability for follow-up mode, can better meet the requirements for extracting features in multimode recognition task, have broad application prospects.

Description

EMG Feature Extraction based on smooth small echo coherence
Technical field
The invention belongs to electromyography signal process fields, are related to a kind of EMG Feature Extraction, in particular to a kind of EMG Feature Extraction based on the smooth small echo coherence of myoelectricity.
Background technique
Electromyography signal (Electromyogram, EMG) is the superposition of muscle fibre action potential, is that one kind is faint, non-linear, Random non-stable bioelectrical signals.Medically, electromyography signal can be used to diagnose and analyze Parkinson's disease, amyotrophia funiculus lateralis The diseases such as hardening.In engineer application field, the bioelectrical signals such as myoelectricity are used frequently as control signal source.Neural activation bit can To obtain from electromyography signal, according to the approach for obtaining electromyography signal, it is generally divided into plug-in type electromyography signal and surface myoelectric letter Number (surface EMG, sEMG).Surface electromyogram signal is not due to needing insertion skin layer hereinafter, being got over to human zero damage It is used come more researchers, is also had a wide range of applications.
The feature extraction of surface electromyogram signal is the premise and committed step of pattern-recognition and regression forecasting, its purpose exists In therefrom obtaining useful information and removing the information of redundancy, so that information has better ga s safety degree.According to feature The mode of extraction can be divided into time domain, frequency domain, time-frequency domain, nonlinear dynamic analysis method.Time-frequency domain method combines time domain And the characteristics of frequency domain, the characteristic information of description signal that not only can be local, but also will appreciate that the frequency domain information of signal, wavelet transformation is A kind of main method of time-frequency domain.
Wavelet transformation has a wide range of applications in surface electromyogram signal processing, it is by Decomposition Surface EMG at many Subband comprising precise information.And coherent analysis, and number are carried out to the relationship in time frequency space between two kinds of echo signals Common method in word signal analysis especially wavelet transformation.RyotaroImoto et al. uses electromyography signal small echo coherence analysis The coordinated movement of various economic factors mechanism for having studied agonistic muscle and Opposing muscle has obtained under stable condition there is more high correlation than instability condition Conclusion.In general, small echo coherence can be used to analyze nonstationary random signal, such as electromyography signal and EEG signals.Surface myoelectric The coherence of signal can provide a seed coat layer muscle coupling information.Currently based on the research of surface electromyogram signal coherent analysis It is less, there is more wide research space.
Summary of the invention
The present invention proposes a kind of based on smooth small echo coherence point for deficiency existing for existing myoelectricity feature extracting method The EMG Feature Extraction of analysis can effectively extract the feature of surface electromyogram signal.It is adopted by electromyographic signal collection instrument Collect the electromyography signal of human body related muscles, and pre-processed using band-pass filtering method, to the electromyography signal after filtering processing Wavelet transformation is carried out, the cross wavelet analysis of two-way electromyography signal is then calculated, and the time is carried out respectively to cross wavelet analysis Smooth operation on axis and scale axis.Finally, calculating the smooth wavelet coherence of two-way electromyography signal, and examine using T It examines the correlation between different behaviors with the presence or absence of statistical difference, obtains 32 grades of wavelet coherences of different muscle combinations As feature vector.
In order to achieve the goal above, the method for the present invention mainly comprises the steps that
Step 1 obtains related electromyography signal sample data, acquires human body correlation flesh by electromyographic signal collection instrument first Then the electromyography signal of meat is pre-processed using band-pass filtering method.
Step 2 carries out wavelet transformation to the electromyography signal after filtering processing.
Wx(a, b) is wavelet coefficient, and x (t) is electromyography signal to be analyzed,It is Morlet mother wavelet function, a is small Wave scale factor, b are time shifts, and t is local time's origin.
P is frequency parameter, and σ is the parameter for controlling small wave attenuation.
Step 3 calculates two-way electromyography signal x (t), the cross wavelet analysis of y (t):
Step 4, to Wxy(a, b) is smoothed.First calculate the smooth operation on time shaft.
c1It is normalisation coefft, * represents convolution algorithm.Then the smooth operation on scale axis is calculated.
Sa(Wx(a, b))=Wx(a,b)*c2Π(0.6a)
c2It is also normalisation coefft, Π is rectangular function.
Step 5 calculates the smooth wavelet coherence of two-way electromyography signal.
S is smooth function, S (W)=Sa[St(W)]。
Step 6 obtains maximum average wavelet coherence R when taking 32 scalexy(32,b).Then it is examined using T to check Whether there were significant differences in different behaviors for the wavelet coherence of different muscle combinations, will finally have maximum average small echo phase 32 multi-scale wavelet coherence factors of the muscle combination of responsibility number are as feature vector.
The EMG Feature Extraction based on myoelectricity small echo coherence that the present invention designs, has a characteristic that
The smooth small echo coherence that the present invention uses has significantly validity in terms of feature extraction, and propose Feature extracting method is broken down into 32 scales using wavelet transformation to electromyography signal, wavelet coefficient is recycled to calculate small echo Coherence factor examines correlation between different behaviors with the presence or absence of statistical difference using T inspection later, then by different fleshes Method of 32 grades of wavelet coherences as feature vector of meat combination, for follow-up mode identify discrimination with higher with Reliability can better meet the requirements for extracting features in multimode recognition task, have broad application prospects.
Detailed description of the invention
Fig. 1 is implementation flow chart of the invention；
Fig. 2 (a) is original semitendinosus surface electromyogram signal figure；
Fig. 2 (b) is the semitendinosus surface electromyogram signal figure after noise reduction；
Fig. 3 (a) is that the coherence of rectus femoris and semitendinosus over time and frequency schemes；
Fig. 3 (b) is that the coherence of rectus femoris and gastrocnemius over time and frequency schemes；
Fig. 3 (c) is that the coherence of rectus femoris and tibialis anterior over time and frequency schemes；
Fig. 3 (d) is that the coherence of semitendinosus and gastrocnemius over time and frequency schemes；
Fig. 3 (e) is that the coherence of semitendinosus and tibialis anterior over time and frequency schemes；
Fig. 3 (f) is that the coherence of gastrocnemius and tibialis anterior over time and frequency schemes；
Fig. 4 (a), which falls, occurs the wavelet coherence average value of all muscle combinations under different scale；
Fig. 4 (b) moves the wavelet coherence average value of all muscle combinations under different scale upstairs；
The wavelet coherence average value of all muscle combinations under Fig. 4 (c) road-work different scale；
Fig. 4 (d), which goes downstairs, moves the wavelet coherence average value of all muscle combinations under different scale；
The wavelet coherence average value of all muscle combinations under Fig. 4 (e) stance different scale；
The wavelet coherence average value of all muscle combinations under Fig. 4 (f) walking movement different scale；
Specific embodiment
As shown in Figure 1, the present embodiment includes the following steps:
Step 1 obtains related electromyography signal sample data, specifically: passing through DELSYS TrignoWireless The electromyography signal of related muscles when System electromyographic signal collection instrument acquires human body lower limbs movement, the experiment movement taken are station It stands, walk, run, go upstairs, go downstairs and falls, the related muscles taken are gastrocnemius, tibialis anterior, rectus femoris and half tendon Flesh.After being collected data, Signal Pretreatment is carried out using band-pass filtering method.Semitendinosus surface myoelectric letter after original and noise reduction Number such as Fig. 2 (a), shown in 2 (b).
Step 2 carries out wavelet transformation to the electromyography signal after filtering processing.
Wx(a, b) is wavelet coefficient, and x (t) is electromyography signal to be analyzed,It is Morlet mother wavelet function, a is small Wave scale factor, b are time shifts, and t is local time's origin.
P is frequency parameter, this example is that 6, σ is the parameter for controlling small wave attenuation.
Step 3 calculates two-way electromyography signal x (t), the cross wavelet analysis of y (t):
Step 4, to Wxy(a, b) is smoothed.First calculate the smooth operation on time shaft.
c1It is normalisation coefft, * represents convolution algorithm.Then the smooth operation on scale axis is calculated.
Sa(Wx(a, b))=Wx(a,b)*c2Π(0.6a)
c2It is also normalisation coefft, Π is rectangular function.
Step 5 calculates the smooth wavelet coherence of two-way electromyography signal.
S is smooth function, S (W)=Sa[St(W)]。
Step 6 obtains maximum average wavelet coherence R when taking 32 scalexy(32,b).Then it is examined using T to check Whether there were significant differences in different behaviors for the wavelet coherence of different muscle combinations, will finally have maximum average small echo phase 32 multi-scale wavelet coherence factors of the muscle combination of responsibility number are as feature vector.Different muscle combinations are over time and frequency Shown in coherence such as Fig. 3 (a)-(f)；Size distribution of the wavelet coherence in different daily behaviors such as Fig. 4 (a)-(f) institute Show.

Claims (1)

1. the EMG Feature Extraction based on smooth small echo coherence, it is characterised in that: this method comprises the following steps:
Step 1 obtains related electromyography signal sample data, acquires human body related muscles by electromyographic signal collection instrument first Then electromyography signal is pre-processed using band-pass filtering method；
Step 2 carries out wavelet transformation to the electromyography signal after filtering processing；
Wx(a, b) is wavelet coefficient, and x (t) is electromyography signal to be analyzed,It is Morlet mother wavelet function, a is small echo ruler The factor is spent, b is time shift, and t is local time's origin；
P is frequency parameter, and σ is the parameter for controlling small wave attenuation；
Step 3 calculates two-way electromyography signal x (t), the cross wavelet analysis of y (t):
Step 4, to Wxy(a, b) is smoothed；First calculate the smooth operation on time shaft；
c1It is normalisation coefft, * represents convolution algorithm；Then the smooth operation on scale axis is calculated；
Sa(Wx(a, b))=Wx(a,b)*c2Π(0.6a)
c2It is also normalisation coefft, Π is rectangular function；
Step 5 calculates the smooth wavelet coherence of two-way electromyography signal；
S is smooth function, S (W)=Sa[St(W)]；
Step 6 obtains maximum average wavelet coherence R when taking 32 scalexy(32,b)；Then difference is checked using T inspection Whether there were significant differences in different behaviors for the wavelet coherence of muscle combination, will finally have maximum average small echo phase responsibility 32 multi-scale wavelet coherence factors of several muscle combinations are as feature vector.
CN201811603107.XA 2018-12-26 2018-12-26 EMG Feature Extraction based on smooth small echo coherence Pending CN109567799A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811603107.XA CN109567799A (en) 2018-12-26 2018-12-26 EMG Feature Extraction based on smooth small echo coherence

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811603107.XA CN109567799A (en) 2018-12-26 2018-12-26 EMG Feature Extraction based on smooth small echo coherence

Publications (1)

Publication Number Publication Date
CN109567799A true CN109567799A (en) 2019-04-05

Family

ID=65932752

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811603107.XA Pending CN109567799A (en) 2018-12-26 2018-12-26 EMG Feature Extraction based on smooth small echo coherence

Country Status (1)

Country Link
CN (1) CN109567799A (en)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110151176A (en) * 2019-04-10 2019-08-23 杭州电子科技大学 A kind of continuous method for estimating of upper limb elbow joint based on electromyography signal
CN111563581A (en) * 2020-05-27 2020-08-21 杭州电子科技大学 Method for constructing brain muscle function network based on wavelet coherence
CN112036357A (en) * 2020-09-09 2020-12-04 曲阜师范大学 Upper limb action recognition method and system based on surface electromyogram signal
CN112932508A (en) * 2021-01-29 2021-06-11 电子科技大学 Finger activity recognition system based on arm electromyography network
CN116189909A (en) * 2023-03-06 2023-05-30 佳木斯大学 Clinical medicine discriminating method and system based on lifting algorithm
CN116595290A (en) * 2023-07-17 2023-08-15 广东海洋大学 Method for identifying key factors affecting chlorophyll change of marine physical elements

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101833671A (en) * 2010-03-30 2010-09-15 西安理工大学 Support vector machine-based surface electromyogram signal multi-class pattern recognition method
CN105411565A (en) * 2015-11-20 2016-03-23 北京理工大学 Heart rate variability feature classification method based on generalized scale wavelet entropy
WO2016110804A1 (en) * 2015-01-06 2016-07-14 David Burton Mobile wearable monitoring systems
CN105997064A (en) * 2016-05-17 2016-10-12 成都奥特为科技有限公司 Method for identifying human lower limb surface EMG signals (electromyographic signals)
CN107506579A (en) * 2017-08-14 2017-12-22 西南大学 Cerebral hemorrhage forecast model method for building up and system based on integrated study
CN107997741A (en) * 2018-01-08 2018-05-08 北京大学 A kind of coupling analytical method of Fetal Heart Rate and uterine contraction signal
CN108538388A (en) * 2018-03-06 2018-09-14 上海逸动医学科技有限公司 A kind of motion of knee joint obstacle function evaluation method

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101833671A (en) * 2010-03-30 2010-09-15 西安理工大学 Support vector machine-based surface electromyogram signal multi-class pattern recognition method
WO2016110804A1 (en) * 2015-01-06 2016-07-14 David Burton Mobile wearable monitoring systems
CN105411565A (en) * 2015-11-20 2016-03-23 北京理工大学 Heart rate variability feature classification method based on generalized scale wavelet entropy
CN105997064A (en) * 2016-05-17 2016-10-12 成都奥特为科技有限公司 Method for identifying human lower limb surface EMG signals (electromyographic signals)
CN107506579A (en) * 2017-08-14 2017-12-22 西南大学 Cerebral hemorrhage forecast model method for building up and system based on integrated study
CN107997741A (en) * 2018-01-08 2018-05-08 北京大学 A kind of coupling analytical method of Fetal Heart Rate and uterine contraction signal
CN108538388A (en) * 2018-03-06 2018-09-14 上海逸动医学科技有限公司 A kind of motion of knee joint obstacle function evaluation method

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
DMYTROIATSENKO等: "Linear and synchrosqueezed time–frequency representations revisited: Overview, standards of use, resolution, reconstruction, concentration, and algorithms", 《DIGITAL SIGNAL PROCESSING》 *
HUA YI等: "The improvement of the Morlet wavelet for multi-period analysis of climate data", 《COMPTES RENDUS GEOSCIENCE》 *
JOHN ASHMEAD: "Morlet Wavelets in Quantum Mechanics", 《QUANTA》 *
RYOTARO IMOTO等: "Preliminaly Study on Coordinated Movement Mechanism of Multiple Muscle Using Wavelet Coherence Analysis", 《2016 5TH IIAI INTERNATIONAL CONGRESS ON ADVANCED APPLIED INFORMATICS》 *

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110151176A (en) * 2019-04-10 2019-08-23 杭州电子科技大学 A kind of continuous method for estimating of upper limb elbow joint based on electromyography signal
CN111563581A (en) * 2020-05-27 2020-08-21 杭州电子科技大学 Method for constructing brain muscle function network based on wavelet coherence
CN111563581B (en) * 2020-05-27 2023-08-18 杭州电子科技大学 Brain muscle function network construction method based on wavelet coherence
CN112036357A (en) * 2020-09-09 2020-12-04 曲阜师范大学 Upper limb action recognition method and system based on surface electromyogram signal
CN112036357B (en) * 2020-09-09 2023-05-12 曲阜师范大学 Upper limb action recognition method and system based on surface electromyographic signals
CN112932508A (en) * 2021-01-29 2021-06-11 电子科技大学 Finger activity recognition system based on arm electromyography network
CN116189909A (en) * 2023-03-06 2023-05-30 佳木斯大学 Clinical medicine discriminating method and system based on lifting algorithm
CN116189909B (en) * 2023-03-06 2024-02-20 佳木斯大学 Clinical medicine discriminating method and system based on lifting algorithm
CN116595290A (en) * 2023-07-17 2023-08-15 广东海洋大学 Method for identifying key factors affecting chlorophyll change of marine physical elements

Similar Documents

Publication Publication Date Title
CN109567799A (en) EMG Feature Extraction based on smooth small echo coherence
Sangaiah et al. An intelligent learning approach for improving ECG signal classification and arrhythmia analysis
CN107157477B (en) Electroencephalogram signal feature recognition system and method
Benalcázar et al. Hand gesture recognition using machine learning and the Myo armband
CN105997064B (en) A kind of discrimination method for human body lower limbs surface electromyogram signal
Barrios-Muriel et al. A simple SSA-based de-noising technique to remove ECG interference in EMG signals
Aziz et al. Electromyography (EMG) data-driven load classification using empirical mode decomposition and feature analysis
Taghizadeh et al. Finger movements classification based on fractional fourier transform coefficients extracted from surface emg signals
CN109567798A (en) Daily behavior recognition methods based on myoelectricity small echo coherence and support vector machines
Bajaj et al. Features based on intrinsic mode functions for classification of EMG signals
Tang et al. ECG de-noising based on empirical mode decomposition
Xie et al. Signal decomposition by multi-scale PCA and its applications to long-term EEG signal classification
Adochiei et al. ECG waves and features extraction using Wavelet Multi-Resolution Analysis
Veer et al. Wavelet denoising and evaluation of electromyogram signal using statistical algorithm
CN114533089A (en) Lower limb action recognition and classification method based on surface electromyographic signals
Camacho Navarro et al. EMG-based system for basic hand movement recognition
CN113609975A (en) Modeling method for tremor detection, hand tremor detection device and method
Vigneshwari et al. Analysis of finger movements using EEG signal
CN110151176A (en) A kind of continuous method for estimating of upper limb elbow joint based on electromyography signal
Ai et al. Gestures recognition based on wavelet and LLE
Xiaojing et al. Feature extraction and classification of sEMG based on ICA and EMD decomposition of AR model
Yu et al. The research of sEMG movement pattern classification based on multiple fused wavelet function
Winarno et al. Evaluation on EMG electrode reduction in recognizing the pattern of hand gesture by using SVM method
Ge et al. Feature extraction and classification of hand movements surface electromyogram signals based on multi-method integration
Acır Automated system for detection of epileptiform patterns in EEG by using a modified RBFN classifier

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication

Application publication date: 20190405

RJ01 Rejection of invention patent application after publication