Using Electrocardiogram and Photoplethysmography Data to Assess Human Emotions

Human emotions are linked to mental well-being and physical health, making emotion recognition via physiological signals increasingly important. Although recent studies show promise, the combined use of electrocardiogram (ECG) and pho-toplethysmography (PPG) data for emotion assessment remains underexplored. This study examines the feasibility of using joint ECG and PPG signals for emotion evaluation within the Affective Dimensional Model (ADM) framework. Morphological features extracted from these signals are used to classify felt arousal and valence with Support Vector Machines (SVM) and Neural Networks (NN). On a per-participant basis, SVM achieved average valence and arousal accuracies of 72.69% (p < 0.05) and 72.30% (p < 0.05), while NN reached 72.48% (p < 0.05) and 73.01% (p < 0.05). The findings suggest that the morphological features of ECG and PPG encode emotion-dependent information, enabling accurate prediction of emotional states.