P2. Spatial and temporal changes of electrical cortical activity induced by self-initiated finger tapping

Abstract

Objective Sustained neuronal activity can be localized with functional magnetic resonance imaging (fMRI). Signal analysis of the time series within and across brain voxels reveal state specific characteristics enriching our understanding of, e.g., different states of consciousness ( Tagliazucchi et al., 2015 , Tagliazucchi et al., 2014 , Tagliazucchi et al., 2013a , Tagliazucchi et al., 2013b ). It should be possible to derive similar information from a more universally available and bed side tool, the electroencephalogram (EEG) ( Tagliazucchi et al., 2012a , Tagliazucchi et al., 2012b ). Because surface EEG is a blurred representation of multiple neuronal sources, source analysis is required to add spatial detail. Here, we validate a source reconstruction method facilitating the analysis of ongoing EEG in source space with methods comparable to those applied to fMRI time series: We test whether we can localize well described surface EEG changes induced by a finger tapping task to motor cortical brain regions. Method 30 subjects underwent a self-paced finger tapping task which induces desynchronization of the surface EEG ( Pfurtscheller et al., 1996 ). We recorded from 29 EEG channels and EMG electrodes over the right and left M. abductor pollicis brevis and created grand average maps for each tapping condition (left hand, right hand, both hands, no tapping) both in sensor and source (minimum variance distortionless response [‘minimum norm’] beamforming) space. Results In sensor space, we observed a drop in 8–12 Hz activity near C3 and C4 ( Fig. 1 ) with an onset a few seconds before the start of each tapping condition. In source space, the desynchronization localized to central regions ( Fig. 2 ). Discussion EEG desynchronization was in agreement with ( Pfurtscheller et al., 1996 , McFarland et al., 2000 ), the drop in power preceding the EMG-defined (self-paced) tapping might reflect motor planning. The minimum norm beamforming localized the EEG changes biologically plausible to central regions. In conclusion, we validated a technique, which transforms ongoing EEG into source space. This is an important prerequisite for further quantitative EEG signal analysis with spatial information similar those applied to fMRI time series.