Krogmeier C, Coventry BS and Mousas C (2022) Affective Image Sequence Viewing in Virtual Reality Theater Environment: Frontal Alpha Asymmetry Responses From Mobile EEG. Front. Virtual Real. 3:895487. doi: 10.3389/frvir.2022.895487
Date of this Version
virtual reality, mobile EEG, frontal alpha asymmetry, affective images, valence, arousal, emotion, frontal alpha asymmetry burst analysis
Background: Numerous studies have investigated emotion in virtual reality (VR) experiences using self-reported data in order to understand valence and arousal dimensions of emotion. Objective physiological data concerning valence and arousal has been less explored. Electroencephalography (EEG) can be used to examine correlates of emotional responses such as valence and arousal in virtual reality environments. Used across varying fields of research, images are able to elicit a range of affective responses from viewers. In this study, we display image sequences with annotated valence and arousal values on a screen within a virtual reality theater environment. Understanding how brain activity responses are related to affective stimuli with known valence and arousal ratings may contribute to a better understanding of affective processing in virtual reality. Methods: We investigated frontal alpha asymmetry (FAA) responses to image sequences previously annotated with valence and arousal ratings. Twenty-four participants viewed image sequences in VR with known valence and arousal values while their brain activity was recorded. Participants wore the Oculus Quest VR headset and viewed image sequences while immersed in a virtual reality theater environment. Results: Image sequences with higher valence ratings elicited greater FAA scores than image sequences with lower valence ratings (F [1, 23] = 4.631, p = 0.042), while image sequences with higher arousal scores elicited lower FAA scores than image sequences with low arousal (F [1, 23] = 7.143, p = 0.014). The effect of valence on alpha power did not reach statistical significance (F [1, 23] = 4.170, p = 0.053). We determined that only the high valence, low arousal image sequence elicited FAA which was significantly higher than FAA recorded during baseline (t  = −3.166, p = 0.002), suggesting that this image sequence was the most salient for participants. Conclusion: Image sequences with higher valence, and lower arousal may lead to greater FAA responses in VR experiences. While findings suggest that FAA data may be useful in understanding associations between valence and arousal self-reported data and brain activity responses elicited from affective experiences in VR environments, additional research concerning individual differences in affective processing may be informative for the development of affective VR scenarios.