Calculation of the cyclic characteristics of the electroencephalogram for investigation of the electrical activity of the brain

The purpose of the study is experimental verification of the proposed EEG analysis method based on the construction of a connectivity graph of the analyzed signal, in which the amplitudes are displayed by vertices, and their relative position relative to each other by arcs. The display of the EEG signal in the graph structure causes the appearance of cyclic structures with the possibility of calculating their numerical characteristics. As a result of the study, criteria for initialization of the initial conditions of the counting algorithm have been developed. The following parameters were calculated: the number of cycles and the Euler number in the EEG recording. Coil representations of graphs are given. The proposed algorithm has a scaling parameter, the choice of which affects the final results. The second free parameter of the proposed algorithm is the degree of artificial signal coarsening. Variants of the algorithm application for multichannel EEG signals with multichannel signal processing by channel-by-channel detection of semantic units and construction of a generalized semantic connectivity graph are considered. An example of an analyzed multichannel EEG signal, which was pre-processed with reduction of all amplitudes to natural numbers in accordance with the calculated characteristics, is given. An example of an EEG of a subject with closed eyes during quiet wakefulness and an EEG of a subject with open eyes is given. In Conclusion, it is shown that the final indicators can vary significantly (from zero to tens of thousands or more) depending on the particular derivation of the EEG channel. Analysis of the cyclic structures of the electroencephalogram seems to be a potential way to assess various human states due to the possibility of distinguishing them using the proposed method. The study has a limited, pilot character.