26 April 2018 Optimal spatiotemporal representation of multichannel EEG for recognition of brain states associated with distinct visual stimulus
Author Affiliations +
Abstract
In the paper we propose an approach based on artificial neural networks for recognition of different human brain states associated with distinct visual stimulus. Based on the developed numerical technique and the analysis of obtained experimental multichannel EEG data, we optimize the spatiotemporal representation of multichannel EEG to provide close to 97% accuracy in recognition of the EEG brain states during visual perception. Different interpretations of an ambiguous image produce different oscillatory patterns in the human EEG with similar features for every interpretation. Since these features are inherent to all subjects, a single artificial network can classify with high quality the associated brain states of other subjects.
© (2018) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Alexander Hramov, Alexander Hramov, Vyacheslav Yu. Musatov, Vyacheslav Yu. Musatov, Anastasija E. Runnova, Anastasija E. Runnova, Tatiana Yu. Efremova, Tatiana Yu. Efremova, Alexey A. Koronovskii, Alexey A. Koronovskii, Alexander N. Pisarchik, Alexander N. Pisarchik, } "Optimal spatiotemporal representation of multichannel EEG for recognition of brain states associated with distinct visual stimulus", Proc. SPIE 10717, Saratov Fall Meeting 2017: Laser Physics and Photonics XVIII; and Computational Biophysics and Analysis of Biomedical Data IV, 107171M (26 April 2018); doi: 10.1117/12.2315140; https://doi.org/10.1117/12.2315140
PROCEEDINGS
6 PAGES


SHARE
Back to Top