Neural mechanisms and functional neuroanatomical networks during memory and cue-based task switching as revealed by residue iteration decomposition (RIDE) based source localization.

Task switching processes reflect a faculty of cognitive flexibility. The underlying neural mechanisms and functional cortical networks have frequently been investigated using neurophysiological (EEG) or functional imaging methods. However, task switching processes are subject to strong intra-individual variability, especially when tested under varying levels of working memory demands. This intra-individual variability compromises the reliable estimation of neurophysiological processes and related functional neuroanatomical networks. In this study, we combine residue iteration decomposition (RIDE) of event-related potentials (ERPs) and source localization methods to circumvent this problem. Due to strong intra-individual variability, behavioral effects between memory-based and cue-based task switching were not reflected by classical ERPs, but were so after applying RIDE. Using RIDE, modulations paralleling the behavioral data were specifically reflected by processes related to the updating of internal representations for response selection (reflected by the C-cluster in the P3-component time range) rather than by stimulus and motor-related processes (reflected by the S-cluster and R-cluster). The C-cluster-processes were associated with activation differences in the inferior parietal cortex, including the temporo-parietal junction (TPJ, BA40) and likely reflect mechanisms related to the updating of internal representations and task sets for response selection. The results underline the necessity to use temporal decomposition methods to control the problem of intra-individual signal variability to decipher the neurophysiology and functional neuroanatomy of cognitive processes.