Assessing the effects of voluntary and involuntary eyeblinks in independent components of electroencephalogram

NEUROCOMPUTING

卷:193

Article

Kanoga, Suguru¹  Nakanishi, Masaki²  Mitsukura, Yasue¹ 

[1] Keio Univ, Sch Integrated Design Engn, Yokohama, Kanagawa 223, Japan

[2] Univ Calif San Diego, Inst Neural Computat, Swartz Ctr Computat Neurosci, San Diego, CA 92103 USA

关键词: Eyeblinks;    Electroencephalographic signals;    Artifacts;    Independent component analysis;    Wavelet transform;   

DOI  :  10.1016/j.neucom.2016.01.057

来源: Elsevier

【 摘 要 】

The effect of voluntary and involuntary eyeblinks in independent components (ICs) contributing to electroencephalographic (EEG) signals was assessed to create templates for eyeblink artifact rejection from EEG signals with small number of electrodes. Fourteen EEG and one vertical electrooculographic signals were recorded for twenty subjects during experiments that prompted subjects to blink voluntarily and involuntarily. Wavelet-enhanced independent component analysis with two markers was employed as a feature extraction scheme to investigate the effects of eyeblinks in ICs of EEG signals. Extracted features were separated into epochs and analyzed. This paper presents following characteristics: (i) voluntary and involuntary eyeblink features obtained from all channels present significant differences in the delta band; (ii) distorting effects have continued influence for 3.0-4.0 s (in the occipital region, 2.0 s); and (iii) eyeblink effects cease to exist after the zero-crossing four (in the occipital region, two) times, regardless of the type. Several characteristics are different between voluntary and involuntary eyeblinks in EEG signals. Therefore, any templates need both types of data for eyeblink artifact rejection if the EEG signals were obtained from small number of electrodes. (C) 2016 The Authors. Published by Elsevier B.V.

【 授权许可】

Free