Quantitative Assessment of Stress Through EEG During a Virtual Reality Stress-Relax Session

Recent studies have addressed stress level classification via electroencephalography (EEG) and machine learning. These works typically use EEG-based features, like power spectral density (PSD), to develop classifiers. Nonetheless, these classifiers are usually limited the discrimination of two (stress no stress) or three (low, medium, high) levels. In this study we propose an alternative for quantitative assessment based on EEG regression algorithms. To aim, conducted a group 23 participants (mean age 22.65 ± 5.48) over stress-relax experience while monitoring their EEG. First, stressed Montreal imaging task (MIST), then led them through 360-degree virtual reality (VR) relaxation experience. Throughout session, reported self-perceived (SPSL) surveys. Subsequently, extracted features from developed individual models algorithms predict SPSL. We evaluated performance in terms mean squared percentage error (MSPE) correlation coefficient ( R 2 ). The results yielded evaluation (MSPE = 10.62 2.12, 0.92 0.02) suggest that our approach predicted with remarkable performance. may positive impact diverse areas could benefit prediction. include research fields neuromarketing, training professionals such as surgeons, industrial workers, firefighters, often face stressful situations.