Ateke Goshvarpour, Sahand University of Technology, Tabriz, Iran
Ataollah Abbasi, Sahand University of Technology, Tabriz, Iran
Atefeh Goshvarpour, Sahand University of Technology, Tabriz, Iran

Abstract: The objective of the present study is to investigate the anatomical distribution of the cortical sources of emotional response to music videos by means of electroencephalogram (EEG) analysis. A novel methodology is introduced to determine the nonlinear couplings between different brain regions based on the coherence analysis, nonlinear features of EEG recordings and a source localization method, standard low resolution electromagnetic tomography (sLORETA). 32 channels of EEG time series of 32 subjects available in DEAP database were studied. The Lyapunov exponents and approximate entropy were applied to the EEG. The coherence for Lyapunov exponents and approximate entropy were calculated between each electrode paired to all other electrodes. Considering valence and arousal related effects, the sLORETA was applied to each above mentioned feature to determine emotional processing cortices. Using the proposed methodology, significant differences in sLORETA activity are observed between different emotional states. These changes were dominantly localized in the Brodmann 11 area (frontal lobe). In addition, some evidences provided that the left hemisphere is more activated to valence and arousal-related effects. Results suggest that considering two dimensions of emotions concurrently, a wider brain region was dominated in synchronization: superior frontal gyrus, middle frontal gyrus, and superior parietal lobule. Cooperating nonlinear coupling along with EEG source localization methods could provide an interesting tool for understanding the cortical specialization in emotional processes.

Keywords: approximate entropy, coherence analysis, emotion, Lyapunov exponent, standard low resolution brain electromagnetic tomography (sLORETA)