Neural Indexes of Attention Extracted from EEG Correlate with Elderly Reaction Time in response to an Attentional Task

Abstract

In the present paper, we analyze electroencephalogram (EEG) signals recorded by a single frontal channel from 105 elderly subjects while they were responding to an attention-demanded task (Stroop color test). The first objective is to discover how post-cue frequency band oscillations of EEG, as neural index of attention, are correlated with elderly response time (RT), as behavioral index of attention. Furthermore, we aim to detect the most informative period of brain activity (EEG) in which the strongest correlations with reaction time exist. Our results show that 1) there is significant negative correlation between alpha gamma ratio (AGR) and response time (p<0.0001), 2) theta beta ratio (TBR) is positively correlated with subjects' response time (p<0.0001) and 3) these correlations are stronger in a 500ms period right after triggering the cue (question onset in Stroop test). Our study provides an insight into the research on analysis and prediction of subject behavior from EEG. Moreover, it has potential to be used in implementation of feasible and efficient single channel EEG-based brain computer interface (BCI) training systems for elderly.