Computational Psychiatry

Computational Psychiatry (CP) combines neuro-sciences, mathematics and computer processing of multiple types of sensory data, Actimetry in general, to better understand and, diagnose and treat psychiatric diseases. The CP is based on artificial intelligence, automatic learning (machine learning) and Deep Learning. These techniques have an explanatory or predictive aim. For example, in the explanatory domain, the CP helps to establish a dimensional diagnostic classification derived from many sources, e.g, genome, behaviors, environments. In the predicable field, the CP helps anticipate the response to antidepressant treatment from multiple clinical and paraclinical variables.

The CP brings the analytical competence necessary to extract clinically useful information from complex signals harvested in the ecological context in which the patient evolves. The CP goes beyond the usual statistical analyzes of the recordings to extract intrinsic characteristics into the temporal evolution of the signals, e.g., fractal temporal dimension. These extractions are indicators of the complexity of underlying psycho physiological processes and, as such, better indicators of (mental) health status.

A complementary point view can be found at Smart Digital Biomarkers or Targeted Digital Diagnostics.

These principles are deeply imbedded into the Cassiopée computational engine Cassiopée

For more infos about the system and its implementation, please, use contact e-mail: Cassiopée support.

From Complex Patterns to Actionable Understanding

Metaphorically, we can sum up the transition from complex signals to actionable ones by Cassiopée acting as an information density filter.

Computational Psychiatry