Over the last decade we have deployed a range of two-agent interactive games and used these toward two ends: 1. To track neural responses that correlate with important features/parameters of the game(s), and (2) to parse clinically-important designations in populations of humans diagnosed with various mental or neurological disorders. We have shown that measured patterns of exchange on these games exposes natural clusters that map onto symptom-cluster-defined mental disorders. The possibility of using games to extract low-dimensional descriptions of human mental disease provides one starting point for a computational psychiatry.
RSVP link: Click here.