PRACTICO-CM Computational Psychiatry and Integral Models of Behavior

Human behavior is understood in many different ways from different fields and sciences, and therefore different tools are used to answer different questions. A first notion of behavior has to do with the physical actions carried out by a person in a certain context, a framework within which we would include mobility and other physical activity, sleep, intellectual activity, etc. On a second level, people, as belonging to an ultra-social species like ours, interact with each other in a social context, in situations ranging from cooperation to trust others. The main difference with the previous notion is that the decision making in a social context is influenced by the fact that each person follows their own objectives that might not align with ours; this is called strategic interaction. Finally, for a psychologist or a psychiatrist, behavior, and especially its alterations, is linked to manifestations of mental disorders, which are usually studied with reference to behavioral patterns considered “normal” in a certain sense.

The project is based on the hypothesis that these three notions of human behavior are the projection onto different domains of the same entity, and therefore there is a connection between them that allows explaining and predicting to a certain extent what is observed in one domain from the others. Our goal is to test this hypothesis and, above all, to advance its application by means of a multidisciplinary approach and team. To achieve this, we will build physical behavior models based on the digital fingerprint of the individuals, strategic behavior models (both in cooperative and other relevant aspects of mental disorders) using data obtained from experiments specifically designed for this purpose, and models in mental disorder behavior (particularly adaptive disorders). Next, we will analyze the relationships between them, evaluating the mutual predictive capacity and obtaining a description of the mechanisms involved in the relationship between the different behavioral traits. This will allow us to automatically generate representative behavioral traits and, in the specific case of adaptive disorders, propose a personalized treatment for patients to accelerate their recovery (thus reducing direct and indirect costs). Specifically, these models will be based on a probabilistic description using latent (hidden) variables, hierarchically ordered and, in the case of physical behavior, based on deep architectures. The project will combine telecommunication engineering and signal theory techniques, to obtain and process data obtained from mobile devices, with behavioral sciences, with which the design and analysis of data from specific experiments will be addressed to test our hypothesis. These results will be carried out with a team of psychiatrists and their patients, and their applicability and efficacy will be tested with a population of people affected by mental disorders in different situations and receiving different therapies.