Toward a Better Understanding of the Cortical Function: Seeing a Full Video

Abstract

An important issue in EEG studies is the size T of the segment of the recorded EEG to be selected for analysis. Because the majority of the studies reported in the literature use tasks of short duration, the value of T is also short, certainly less than 2 seconds. In contrast to this, events of interest in natural life occur in time windows of much larger magnitudes. Many videos seeing in TV and Internet or received by social media have duration in the range of seconds or minutes. The purpose of the present paper is to investigate the understanding of this type o video. EEG was recorded while volunteers saw a 12 seconds video and Event Related Activity (ERA) technique was used to disclose the brain activity correlates of the full video understanding instead of sensory recognition of any its specific scene or scene component. Video decoding was statistically segmented into 6 EEG epochs (ERA(n)) of 2 seconds duration corresponding to identification of a mosquito flying around a sleeping baby (ERA(1)); that may eventually land over the baby stinging him (ERA(2)); what motivates mother to kill the insect (ERA3(2)), using either a spray (ERA(4)); plug device (ERA(5)) or piece of fabrics (ERA(6)). Each ERA(n) was, in turn, segmented in 4 distinct waves Wi of around 500 ms duration. W1 and W2 are proposed to be associated with syntactic and semantic analyses of each video episode, whereas W3 and W4 are correlated with scene meaningfulness and trustiness evaluation. This is one of three papers exploring the hypothesis that cortical processing is supported by a Cortical Oscillatory Modular Processing (COMP).