Cerebral Cortex Atlas of Emotional States Through EEG Processing

ALEJANDRO GOMEZ; OLGA LUCIA QUINTERO; LOPEZ CELANI, NATALIA MARTINA; LUISA FERNANDEZ VILLA

IFMBE PROCEEDINGS

Springer Nature Switzerland

Año: 2019 vol. 75 p. 138 - 144

1680-0737

This paper addresses the cerebral cortex maps constructionfrom EEG signals getting an information simplification methodfor an emotional state phenomenon description. Bi-dimensional densitydistribution of main signal features are identified and a comparisonto a previous approach is presented. Feature extraction scheme is performedvia windowed EEG signals Stationary Wavelet Transform withthe Daubechies Family (1?10); nine temporal and spectral descriptorsare computed from the decomposed signal. Recursive feature selectionmethod based on training a Random forest classifier using a one-vs-allscheme with the full features space, then a ranking procedure via giniimportance, eliminating the bottom features and restarting the entireprocess over the new subset. Stopping criteria is the maximum accuracy.The main contribution is the analysis of the resulting subset featuresas a proxy for cerebral cortex maps looking for the cognitive processesunderstanding from surface signals. Identifying the common location ofdifferent emotional states in the central and frontal lobes, allowing to bestrong parietal and temporal lobes differentiators for different emotions.