Brain electrical activity mapping (BEAM) in schizophrenic patients.

A new topographic approach, brain electrical activity mapping, summarizes EEG and evoked potential data as color maps. Eleven drug-free and 14 medicated schizophrenic patients and 11 normal controls were studied with this technique. Compared with controls, both schizophrenic groups had more slow activity (delta, 0 to 3.5 Hz), greatest in frontal regions, and more fast activity (beta, 20 to 31.5 Hz) in postcentral regions. Both schizophrenic groups were different from controls late in the visual evoked potential and in the middle of the auditory evoked potential. Overall, using multivariate discriminant analysis, a very significant group separation effect was found, and retrospective classification was 95% successful in discriminating both schizophrenic groups from controls. To estimate the ability of our features to assign subjects prospectively to control or patient groups, we employed "jackknifing," which resulted in overall classification success rates of 84.0% for medicated schizophrenic patients v controls and 81.8% for drug-free schizophrenic patients v controls.