Removing electroencephalographic artifacts by blind source separation

• DOI: 10.1111/1469-8986.3720163
• Publication Date: 2004-12-28T14:36:57Z
• Authors: Tzyy‐Ping Jung; Scott Makeig; Colin Humphries; Te‐Won Lee; Martin J. McKeown; Vicente Iragui; Terrence J. Sejno...
• Keywords: Artifact (error); Electrooculography

Eye movements, eye blinks, cardiac signals, muscle noise, and line noise present serious problems for electroencephalographic (EEG) interpretation analysis when rejecting contaminated EEG segments results in an unacceptable data loss. Many methods have been proposed to remove artifacts from recordings, especially those arising movements blinks. Often regression the time or frequency domain is performed on parallel electrooculographic (EOG) recordings derive parameters characterizing appearance spread of EOG channels. Because ocular activity mix bidirectionally, regressing out inevitably involves subtracting relevant signals each record as well. Regression become even more problematic a good channel not available artifact source, case artifacts. Use principal component (PCA) has multichannel EEG. However, PCA cannot completely separate brain they comparable amplitudes. Here, we propose new generally applicable method removing wide variety records based blind source separation by independent (ICA). Our collected normal autistic subjects show that ICA can effectively detect, separate, contamination artifactual sources with comparing favorably obtained using methods. also be used analyze blink-related activity.