A new method of processing the electroencephalogram signals (EEG) is proposed which reveals the characteristic signs of pathological activity of the brain based on computer processing of electroencephalographic data of patients with epilepsy for detection in the software environment MatLab. It has been shown that the bioactivity of the frequency rhythms of the brain varies with neuropathology and in a normal state, which makes it possible to automate the procedure for finding pathological sites on the EEG and to reduce the degree of subjectivity of the patient’s brain neuronal evaluation. The article illustrates the work of the method on the example of treatment EEG records of 10 patients with epileptic attacks and in a state of rest. For each entry, percentages of alpha, beta, gamma and delta frequencies are calculated for total bioactivity.