Identificação de estados emocionais complexos com base em ressonância magnética e eletroencefalografia: uma abordagem com cenário de realidade virtual

Abstract

During the last decades, a significant effort has been made to identify objective physiological markers of subjective emotional states. This work presents an experimental protocol for the induction of tenderness and anguish emotions using a multimodal stimulation with the combination of a virtual reality (VR) scene with excerpts of music. The protocol was applied in neurofeedback acquisitions using functional magnetic resonance imaging (fMRI) that demonstrated to be possible activate brain regions related to the emotions studied. Subsequently, a new study on the protocol using electroencephalographic (EEG) signals was done, where data were used in offline analysis with artificial neural networks. The results demonstrated that it is possible to distinguish complex emotional states through EEG features with maximum accuracy above 90%. In addition, they allowed a better understanding of the mechanisms of voluntary emotional regulation and its neural correlates, and could contribute to the development of new tools and/or neurotechnological equipment.