Predicting human Decisions in a social Interaction-Scenario using Real-Time Functional Magnetic Resonance Imaging ( rt-fMRI )

Introduction Making decisions in a social context is a fundamental part of our daily life. Is it possible to predict decisions in social interaction scenarios before a subject expresses the own will by investigating distributed activation patterns of the human brain? The presented manuscript addresses this question by using a standard paradigm from economic behavioral research: the Ultimatum Game (UG) [4]. In the UG, two players split a pre-defined sum of money. One player is deemed the proposer and the other, the responder. The proposer makes an offer as to how the money should be split between the two. The second player can either accept or reject this offer. If accepted, the money is split as proposed. If rejected, neither player receives anything. In standard approaches of mind-reading the analysis is done after the measurement [5, 6], prohibiting the usage of the prediction to influence user interaction or dynamic scene representation. In the presented study a real-time fMRI (rtfMRI) system [2] was used to derive the brain activation of the responder. Using pattern classification it was possible to predict the decision of the responder ‘on the fly’ by means of the activation of three distinct brain regions: Anterior Insula (AI), Lateral Praefrontal Cortex (LPFC) and Visual Cortex (VC). The classification result was presented to the operator 1-2 seconds before the subject pressed a button to convey the decision. The classification accuracy reached about 70% averaged over seven subjects.