Brain's capacity to detect abstract regularities from visual stimuli under different attentive conditions- an ERP study

Abstract

Many previous studies have applied oddball paradigm to study change detection. Although changes within single features have been investigated a lot, the changes in multiple feature conjunctions have not. The aim of our study was to investigate with event-related potentials by applying oddball paradigm, whether the brain can detect abstract regularities in visual stimulus stream when two different features are combined - semantic meaning and color. Participants were shown adjective words written in red and blue print in quasi-random order on a computer screen. In an oddball paradigm, 90 % of the words (‘standard’) followed the rule “words printed in red have a negative meaning and the words printed in blue have a positive meaning”. In the rest 10 % of the words (‘deviant’), the rule was reversed: “words printed in red have a positive meaning and the words printed in blue have a negative meaning”. In the half of the subjects the rules in standards and deviants were reversed. The effect of attention directed to the stimuli was explored by using two groups of subjects. In one group, the subjects were not informed of the regularities in stimuli (non informed, n=20), and they were instructed to count the six-letter words. In the other group participants were informed about the rule for the regularity and they were instructed to count the rule violations (informed, n=20). During the experiment, participants’ electroencephalography (EEG) was recorded with electrodes attached to their scalp. In non informed group the deviants elicited differential ERPs (event-related potential) compared to standards at two different latencies. At 320-360 ms post-stimulus a negative shift was observed in posterior electrodes and a positive shift in fronto-central electrodes. At 410-450 ms post-stimulus a negative shift was observed in left temporal electrodes and a positive shift in right temporal electrodes. In the attended informed group no significant differences were found between standards and deviant responses at any latency range. The results are discussed in the context of the visual mismatch negativity (vMMN).