Fuzzy segmentation spatiotemporal patterns of cognitive potential into microstates

Fuzzy c-mean algorithm was applied to segment spatiotemporal patterns of br ainwave into microstates and memberships. The optimal clustering number was estimated with both the trends of objective function and the eigenvalue nu mber of microstates. Comparable spatial patterns may occur at different tem poral moments in consideration of fuzzy index that is beyond the Limit of s erial processing. Those techniques were illustrated with multichannel event -related potentials recorded from 9 subjects during Stroop test. Statistica l parametric map of F value suggested that significant task (color decision and word decision) effect involve widespread cortical regions after stimul us onset 280 ms and this result supports the hypothesis that Stroop interfe rence derives from response competition during post-perception stage. As si gnificant stimulus (congruent stimulus and incongruent stimulus) effect onl y involves several separate visual regions within 100 ms after stimulus pre sentation, it may reflect top-down attentional regulation.