Analyzing Emotional Dynamics: Transition Network Insights from Electrodermal Activity

Emotion is a complex dynamic process that involves cognitive, affective, physiological, and expressive elements. Electrodermal Activity (EDA) is a technique used to study the skin's electrical properties. These signals exhibit nonlinear and non-stationary characteristics. In this work, an attempt is made to utilize the transition network approach for the analysis of emotional states. For this, EDA signals are obtained from a public dataset and are decomposed into phasic and tonic components. The phasic component of the EDA is transformed into a binary sequence and is subjected to a transition network algorithm. Further, twelve symbolic features are extracted from the symbolic sequence. The proposed approach is able to differentiate emotional states using EDA signals. Results show that the word probability, Renyi entropy and frequent pattern symbol varied significantly (p < 0.05) in differentiating the arousal dimension. These findings suggest potential clinical relevance for this approach, particularly in facilitating diagnoses, monitoring treatment efficacy, and providing objective measures of emotional response in diverse clinical populations. Clinical Relevance-The proposed method could be used to understand the neurological, psychiatric, and bio-behavioral mechanisms associated with arousal and valence emotional states.