Electroencephalographic scalp-energy analysis as a tool for investigation of cognitive performance

In a new method of using electroencephalography (EEG) to monitor cognitive events, multichannel event-related potentials (ERPs) were used to estimate scalp distributions of surface energy densities of cortically generated electrical fields. Cross-subject regression analyses were then used to map sites and post-stimulus latencies, for which there is a high correlation of energy densities with subjects' performances. In a preliminary study, five right-handed young men were presented mental arithmetic tasks via a computer screen. A 21-channel ERP was computed for each subject and converted to show the scalp distribution of energy density at each sampling period (every 7.8 msec). Indices of subjects' task performances were regressed upon these estimates of potential energy at each electrode site and post-stimulus latency. High correlations were found at four distinct regions and latencies. These corresponded to the stages and sites of cortical localization for mental arithmetic inferred from studies of patients with localized cortical lesions. This correspondence suggests that the EEG procedure may offer an inexpensive, noninvasive method of investigating cortical localization of cognitive function in healthy subjects.