Functional topography of the human nonREM sleep electroencephalogram

The sleep EEG of healthy young men was recorded during baseline and recover y sleep after 40 h of waking. To analyse the EEG topography, power spectra were computed from 27 derivations. Mean power maps of the nonREM sleep EEG were calculated for 1-Hz bins between 1.0 and 24.75 Hz. Cluster analysis re vealed a topographic segregation into distinct frequency bands which were s imilar for baseline and recovery sleep, and corresponded closely to the tra ditional frequency bands. Hallmarks of the power maps were the frontal pred ominance in the delta and alpha band, the occipital predominance in the the ta band, and the sharply delineated vertex maximum in the sigma band. The e ffect of sleep deprivation on EEG topography was determined by calculating the recovery/baseline ratio of the power spectra. Prolonged waking induced an increase in power in the low-frequency range (1-10.75 Hz) which was larg est over the frontal region, and a decrease in power in the sigma band (13- 15.75 Hz) which was most pronounced over the vertex. The topographic patter n of the recovery/baseline power ratio was similar to the power ratio betwe en the first and second half of the baseline night. These results indicate that changes in sleep propensity are reflected by specific regional differe nces in EEG power. The predominant increase of low-frequency power in front al areas may be due to a high 'recovery need' of the frontal heteromodal as sociation areas of the cortex.