Volume Conduction Effects On The Wavelet Coherence Analysis Of EEG Recorded From Subjects With Extreme Low Birth Weight, Change Blindness And Epilepsy

Abstract

The magnitude square coherence (MSC) function is a useful measure to detect the linear correlation between two signals. However estimation of true MSC function is a major issue in EEG signal processing. Apart from the statistical problems, the volume conduction effects from various uncorrelated sources (VCUS) present in the brain introduce the biased estimates into its true value. Based on the assumption that VCUS effects are additive and therefore can be ignored while comparing two or more different EEG activities, various EEG-based studies have ignored the issue of VCUS effects. MSC analysis of epilepsy, MSC analysis of change blindness, and MSC analysis for extreme low birth weight effects on brain are among such studies which have been frequently reported in literature without keeping in account the issue of VCUS effects. This study using EEG and scalp Laplacian-based wavelet MSC functions has provided the substantial evidence that VCUS effects are not additive and they produce significant effects on the results of these studies. Since scalp Laplacian is a useful measure under some conditions for minimizing the VCUS effects, the difference in SL-based MSC function between control and non-control subjects was compared to the same study but based on conventional EEG based MSC function. And then the results of statistical analysis obtained from both studies were compared to each other in order to examine the effects of VCUS effects. This study was repeated for above three mentioned neurological disorders.