Dimensional complexity of the EEG indicates a right fronto-cortical locus of attentional control

The application of non-linear systems theory to the EEG has been shown to provide information beyond that provided by traditional EEG measures. The present study aimed at discriminating different attentional states of human brain activity based on changes in the dimensional complexity of the EEG. Via headphones subjects (n = 15) were confronted with a dichotic listening task consisting of frequent standard pips and rare deviant pips to each ear. Demands of controlled attention were varied by instructing the subject: (1) to divide attention and count deviants in both ears at once; (2) to selectively attend to deviants in only one ear, but ignore input to the other ear; and (3) to ignore stimuli in both ears. In addition, stimulus load was varied For the ignore and divided-attention conditions by delivering the stimuli either at a slow (0.93/s) or fast rate (1.22/s). Dimensional complexity, traditional power spectra, and event-related potential measures were derived From the EEG record. Beta power at F3 and F4 was higher during selective attention than during both divided-attention and the ignore condition, suggesting a bilateral frontocortical involvement in the focusing of attention. By contrast, dimensional complexity was highest during divided attention, intermediate during selective attention, and lowest during the ignore condition. This effect was maximal at F4 and suggests a particular role of the right frontal cortex in the controlled (i.e., attentive) processing of stimuli once selected. In the ignore condition (but not during divided attention) the increased stimulus rate - presumably enhancing automatic processing at a preattentive state - was accompanied by a reduced dimensionality of the EEG over the left frontal cortex. Yet EEG activity picked up from the left supratemporal areas of the auditory association cortex could have contributed to this effect. Results indicate that the dimensional complexity of the EEG may characterize specific aspects of the brain's attentional state not accessible by traditional EEG measures and ERPs.