Font Size: a A A

Removal of ERP artifacts using independent component analysis

Posted on:2006-07-02Degree:M.AType:Thesis
University:University of LouisvilleCandidate:Wu, JiaFull Text:PDF
GTID:2458390008953846Subject:Psychology
Abstract/Summary:
Within the realm of electrophysiology it is generally understood that there exist two types of electroencephalograph (EEG) recordings: spontaneous waves continually changing with external and internal inputs, and waves that are time-locked to specific stimuli. The latter are known as evoked responses (EPs) or event related potentials (ERPs). While the ERP has the advantages of being sensitive enough to detect minute neuronal changes it also has some disadvantages, one being the distortion by noise. Several relatively recent papers on independent component analysis (ICA) have shown that it is effective in removing sources of interference, and maybe more so than any other method. This report is based on the ERPs of ten 2-year-old children recorded in a syllable listening task. Standard methods of removing noise are compared with ICA using high-density EEG recordings (128 sites). (Abstract shortened by UMI.)...
Keywords/Search Tags:Independent component
Related items