L.A. Potulova, R.A. Maraguei, G.B. Milovanova

In order to optimize human operator performance dynamics of EEG spectra and coherence of a human operator and his/her recognition efficiency during monotonous activity was analyzed under conditions of different informational uncertainty set by the number of choices (hypotheses). Subjects were examined during their sensorimotor activity within 3.5 hours in a dark soundproof room. They had to recognize two or eight figures presented with 15-20% visual noise. The EEG was recorded in 12 derivations of both hemispheres (10-20% International System). Latencies of correct responses (long and short) and erroneous performance (signal omissions, false alarms, and recognition errors) were analyzed. Psychometric analysis revealed certain difference in recognition under conditions of choice of a correct image from a set of small and large numbers of figures. Under conditions of a small number of choices, signal omissions prevailed over the errors of other types. Increase in the degree of the informational uncertainty eliminated this predominance. It was shown that the increase in the number of choices enhanced the level of emotions and CNS activation (spectral density over all frequencies has decreased, the peak in the alpha band was lower or absent, and the beta activity was higher). These changes were accompanied by a certain deceleration of performance impairment. Under conditions of the small number of choices spectral and coherence EEG characteristics in certain cortical areas were different. In most of the cases changes in the spectral density in the theta- and beta-frequencies during signal omissions and false alarms were oppositely directed. Increase in the informational uncertainty led to elimination of these differences in the EEG linear parameters.