Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
ECG Artifact Removal from Surface EMG Signal Using an Automated Method Based on Wavelet-ICA
Mälardalen University, School of Innovation, Design and Engineering, Embedded Systems. (Biomedical Engineering)ORCID iD: 0000-0001-8294-861X
Mälardalen University, School of Innovation, Design and Engineering, Embedded Systems.ORCID iD: 0000-0003-1940-1747
Auckland University of Technology, New Zealand.
2015 (English)In: Studies in Health Technology and Informatics, Volume 211, 2015, 91-97 p.Conference paper, Published paper (Refereed)
Abstract [en]

This study aims at proposing an efficient method for automated electrocardiography (ECG) artifact removal from surface electromyography (EMG) signals recorded from upper trunk muscles. Wavelet transform is applied to the simulated data set of corrupted surface EMG signals to create multidimensional signal. Afterward, independent component analysis (ICA) is used to separate ECG artifact components from the original EMG signal. Components that correspond to the ECG artifact are then identified by an automated detection algorithm and are subsequently removed using a conventional high pass filter. Finally, the results of the proposed method are compared with wavelet transform, ICA, adaptive filter and empirical mode decomposition-ICA methods. The automated artifact removal method proposed in this study successfully removes the ECG artifacts from EMG signals with a signal to noise ratio value of 9.38 while keeping the distortion of original EMG to a minimum.

Place, publisher, year, edition, pages
2015. 91-97 p.
Series
Studies in Health Technology and Informatics, ISSN 0926-9630 ; 211
National Category
Signal Processing
Identifiers
URN: urn:nbn:se:mdh:diva-27873DOI: 10.3233/978-1-61499-516-6-91Scopus ID: 2-s2.0-84939229104ISBN: 978-1-61499-515-9 (print)OAI: oai:DiVA.org:mdh-27873DiVA: diva2:806945
Conference
12th International Conference on Wearable Micro and Nano Technologies for Personalized Health, Västerås, Sweden, June 2-4, 2015
Available from: 2015-04-22 Created: 2015-04-22 Last updated: 2015-08-28Bibliographically approved
In thesis
1. Proposing Combined Approaches to Remove ECG Artifacts from Surface EMG Signals
Open this publication in new window or tab >>Proposing Combined Approaches to Remove ECG Artifacts from Surface EMG Signals
2015 (English)Licentiate thesis, comprehensive summary (Other academic)
Abstract [en]

Electromyography (EMG) is a tool routinely used for a variety of applications in a very large breadth of disciplines. However, this signal is inevitably contaminated by various artifacts originated from different sources. Electrical activity of heart muscles, electrocardiogram (ECG), is one of sources which affects the EMG signals due to the proximity of the collection sites to the heart and makes its analysis non-reliable. Different methods have been proposed to remove ECG artifacts from surface EMG signals; however, in spite of numerous attempts to eliminate or reduce this artifact, the problem of accurate and effective de-noising of EMG still remains a challenge. In this study common methods such as high pass filter (HPF), gating method, spike clipping, hybrid technique, template subtraction, independent component analysis (ICA), wavelet transform, wavelet-ICA, artificial neural network (ANN), and adaptive noise canceller (ANC) and adaptive neuro-fuzzy inference system (ANFIS) are used to remove ECG artifacts from surface EMG signals and their accuracy and effectiveness is investigated. HPF, gating method and spike clipping are fast; however they remove useful information from EMG signals. Hybrid technique and ANC are time consuming. Template subtraction requires predetermined QRS pattern. Using wavelet transform some artifacts remain in the original signal and part of the desired signal is removed. ICA requires multi-channel signals. Wavelet-ICA approach does not require multi-channel signals; however, it is user-dependent. ANN and ANFIS have good performance, but it is possible to improve their results by combining them with other techniques. For some applications of EMG signals such as rehabilitation, motion control and motion prediction, the quality of EMG signals is very important. Furthermore, the artifact removal methods need to be online and automatic. Hence, efficient methods such as ANN-wavelet, adaptive subtraction and automated wavelet-ICA are proposed to effectively eliminate ECG artifacts from surface EMG signals. To compare the results of the investigated methods and the proposed methods in this study, clean EMG signals from biceps and deltoid muscles and ECG artifacts from pectoralis major muscle are recorded from five healthy subjects to create 10 channels of contaminated EMG signals by adding the recorded ECG artifacts to the clean EMG signals. The artifact removal methods are also applied to the 10 channels of real contaminated EMG signals from pectoralis major muscle of the left side. Evaluation criteria such as signal to noise ratio, relative error, correlation coefficient, elapsed time and power spectrum density are used to evaluate the performance of the proposed methods. It is found that the performance of the proposed ANN-wavelet method is superior to the other methods with a signal to noise ratio, relative error and correlation coefficient of 15.53, 0.01 and 0.98 respectively.

Abstract [sv]

Elektromyografi (EMG) är ett verktyg som rutinmässigt används för en mängd olika applikationer inom många discipliner. Dock är denna signal oundvikligen kontaminerad av artefakter som kommer från olika källor. Elektrisk aktivitet av hjärtmuskln, elektrokardiogram (EKG), är en av störkällorna som påverkar EMG-signalerna på grund av närheten till hjärtat och som försämrar analysens tillförlitlig. Olika metoder har föreslagits för att ta bort EKG artefakter från yt-EMG-signaler men trots många försök att eliminera eller minska denna artefakt, kvarstår problemet med korrekt och effektivt brusreducering av EMG. I denna studie har vanliga metoder för brusundertryckning undersökts, såsom högpassfilter (HPF), gatingmetod, spikklippning, hybridteknik, subtraktionsmetod, oberoende komponentanalys (ICA), wavelet, wavelet-ICA, artificiella neurala nätverk (ANN), och adaptiv brusreducering (ANC) och adaptiv neuro fuzzy inference system (ANFIS). Metorderna har använts för att avlägsna EKG- artefakter från yt-EMG-signaler och deras noggrannhet och effektivitet har undersökts. HPF, gatingmetod och spikklippning är snabba; men de tar även bort relevant information från EMG-signalen. Hybridteknik och ANC är tidskrävande. Subtraktionsmetoden kräver kännedom om QRS-mönstret.Wavelettransformen lämnade kvar vissa artefakter i signalen, och avlägsnade även endel av den ursprungliga EMG-signalen. ICA kräver flerkanaliga signaler. Wavelet-ICA kräver inte flerkanaliga signaler, men är däremot användarberoende. ANN och ANFIS har bra prestanda, men det är möjligt att förbättra resultaten genom att kombinera dem med andra tekniker. För vissa tillämpningar av EMG-signaler såsom rehabilitering, rörelsekontroll och prediktion, är kvaliteten på EMG-signalerna mycket viktigt. Dessutom måste de artefaktreducerande metoderna vara i realtid och automatiska. Detta innebär att metoderna ANN-wavelet, adaptiv subtraktion och automatiserad wavelet-ICA rekommenderas för effektiv eliminering av EKG-artefakter från yt-EMG-signaler. För att jämföra resultaten av de undersökta och föreslagna metoderna i denna studie, har rena EMG-signaler från biceps och delta-muskler, samt EKG-artefakter från stora bröstmuskeln spelats in från fem friska personer. För att skapa 10-kanaliga brusiga EMG-signaler har de inspelade EKG-artefakterna adderats till de rena EMG-signalerna. De olika artefaktreduceringsmetoderna har även tillämpats på 10 kanaler verkliga EMG signaler med artefakter, från stora bröstmuskeln på vänster sida. Utvärderingskriterier såsom signal-brusförhållandet, relativta felet, korrelationskoefficienten, förfluten tid och effektspektrumstäthet har använts för att utvärdera de föreslagna metoderna. Prestandan hos den föreslagna ANN-wavelet metoden befanns överlägsen de andra metoderna med ett signalbrusförhållande på 15,53, relativt fel på 0,01 och korrelationskoefficient på 0,98.

Place, publisher, year, edition, pages
Västerås: Mälardalen University, 2015
Series
Mälardalen University Press Licentiate Theses, ISSN 1651-9256 ; 204
National Category
Signal Processing
Identifiers
urn:nbn:se:mdh:diva-27878 (URN)978-91-7485-206-6 (ISBN)
Presentation
2015-06-16, Delta, Mälardalens högskola, Västerås, 13:15 (English)
Opponent
Supervisors
Available from: 2015-04-22 Created: 2015-04-22 Last updated: 2015-05-18Bibliographically approved

Open Access in DiVA

No full text

Other links

Publisher's full textScopushttp://www.es.mdh.se/pdf_publications/3894.pdf

Search in DiVA

By author/editor
Abbaspour, SaraLindén, Maria
By organisation
Embedded Systems
Signal Processing

Search outside of DiVA

GoogleGoogle Scholar

doi
isbn
urn-nbn

Altmetric score

doi
isbn
urn-nbn
Total: 618 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf