FIELD: medicine.
SUBSTANCE: parameters are recorded: spectral analysis of electroencephalograms (EEG), EEG cross correlational function, visual and auditory evoked potentials, cognitive evoked potentials, the amplitude of that conditionally negative deviation wave, average time of simple visual-motor reactions, the average interval between impacts for a tap test, the amplitude of maximum F-response, ration of the maximum F-response and the M-response, latency of the maximum F-response, indicators of heart rate variability (HRV), characteristics of the average values of pulmonary ventilation, respiratory rate, oxygen level in the exhaled air, partial pressure of carbon dioxide in the exhaled air, energy consumption level. Using an artificial neural network with a completed learning procedure, represented by a multilayer perceptron with 72 input neurons, 8 neurons of the intermediate layer and 1 output neuron, the distribution of subjects to the group of practically healthy individuals or patients with epilepsy takes place.
EFFECT: method allows selection and rapid diagnosis of epilepsy.
2 tbl
