Emotional Stimulus Classification from Brain Electrical Activity using Multivariate Empirical Mode Decomposition

Emotions play a crucial role in shaping various aspects of our daily lives, influencing our psychology, perspectives, feelings, and behaviors. Investigating the relationship between visual stimuli and emotions has become a prominent focus in neurophysiological studies. This study offers an overview of emotional responses elicited by different types of arousal-inducing pictures, specifically utilizing the Nencki Affective Picture System (NAPS). The chosen pictures aim to evoke three basic affects: positive, neutral, and negative emotions. Visual stimuli are presented, and emotional data are captured through multichannel Electroencephalogram (EEG) recordings. Additionally, we employ a MEMD-based iterative feature extraction method to decompose the raw signals into sets of oscillations, referred to as intrinsic mode functions (IMFs). Eight reduced IMFs for each visual stimulus are subjected to statistical analysis to assess the emotional state and bolster the understanding of emotional stimulation. The experimental findings indicate that visual stimuli amplify the emotional experience triggered by affective pictures. The results from the collected emotional EEG data demonstrate interdependence between the IMFs and emotional pictures. Furthermore, brain topographs support the statistical analysis by revealing that brain activation is more neuroactive for neutral-based visual stimuli compared to other types of visual stimuli.