Removal of muscular artefacts for the analysis of brain oscillations: Comparison between ICA and SSD

Irene Winkler 1 Stefan Haufe 1 Klaus-Robert Mueller 1
1 TUB - Technische Universität Berlin
Abstract : The electroencephalogram (EEG) is contaminated by undesired signals of non-neural origin, such as movements of the eyes and muscles. The most common approach for muscle artefact reduction is the linear transformation of EEG signals into source components using Blind Source Separation (BSS) techniques, to separate artefac-tual and neuronal sources. Here we present a case study in which we are interested in clean oscillatory EEG activity. We compare the frequently used Independent Component Analysis (ICA) approach with the recently proposed spatio-spectral decomposition (SSD) method. SSD is designed to extract components that explain oscillations-related variance, and is several orders of magnitude faster than ICA. We investigate EEG data from 18 subjects performing self-paced foot movements with respect to event-related desynchroni-sation (ERD) in the beta band. Results indicate that SSD recovers cleaner signals than ICA on this data set.
Type de document :
Communication dans un congrès
ICML Workshop on Statistics, Machine Learning and Neuroscience (Stamlins 2015), Jul 2015, Lille, France. 2015
Liste complète des métadonnées

Littérature citée [20 références]  Voir  Masquer  Télécharger
https://hal.inria.fr/hal-01225250
Contributeur : Bertrand Thirion <>
Soumis le : lundi 9 novembre 2015 - 19:08:16
Dernière modification le : samedi 16 décembre 2017 - 07:18:04
Document(s) archivé(s) le : mercredi 10 février 2016 - 10:11:00

Fichier

Paper_203.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01225250, version 1

Citation

Irene Winkler, Stefan Haufe, Klaus-Robert Mueller. Removal of muscular artefacts for the analysis of brain oscillations: Comparison between ICA and SSD. ICML Workshop on Statistics, Machine Learning and Neuroscience (Stamlins 2015), Jul 2015, Lille, France. 2015. 〈hal-01225250〉

Exporter

BibTeX TEI DC DCterms EndNote

Partager

Métriques

Consultations de la notice

115

Téléchargements de fichiers

326

Contact


archive-ouverte@inria.fr