Analysis of EEG signals to assess emotionality and well-being

Abstract

The purpose of this project is to look into the neural correlates of trait emotional intelligence. This study uses statistical analysis and machine learning to evaluate the relationship between EEG data and the psychological constructs emotionality and well being, two components of the Trait Emotional Intelligence Questionnaire. This project's data is derived from the paper "A mind-brain-body dataset of MRI, EEG, cognition, emotion, and peripheral physiology in young and elderly individuals" published in Scientific Data no6 (Article number: 180308) (Mikolajczak, Bodarwé, Laloyaux, Hansenn, & Nelis, 2010) .There is a hyperlink in this article to a publicly available database called "LEMON database". The LEMON dataset includes 224 subjects who were subjected to various tests and brain analysis methodologies. The context, hypothesis, objectives, development, outcomes, and conclusions are the six phases of this research. In the developed program, the data was sorted into 12 brain regions. The activity of each brain region was segmented into 5s intervals, which were subsequently used to characterize the band power corresponding to 5 different brain waves (i.e. delta, theta, alpha, beta and gamma). Theta is the band with the most relevant differential activation, with beta coming in second. Notably, stronger correlations are found in well-being than in emotionality, with significant p-values being less than 0.01. In general, however, we cannot discriminate between high-scores and low-scores for such constructs (i.e. emotion/well-being) on the basis of the characterized EEG activity since no apparent separation between the groups emerges from the machine learning techniques.