Detection, prediction and monitoring of compulsions – PREDICTOC

Obsessive-Compulsive Disorder (OCD) is a disabling mental disorder featuring obsessions and compulsions. Current best medical practice for treating OCD involves a combination of medications and behavioral and cognitive therapy, and deep-brain stimulation as an alternative for the most severe cases. However, a precise assessment of the symptoms is currently lacking in the medical practice, despite its potential interest to monitor the evolution of the pathology and improve the quality of treatment. To overcome this issue, we propose to implement a novel process using a combination of portable accelerometer and brain signal recording device to automatically detect, predict and quantify compulsive behaviors. Our methodology, using recent advances in machine learning, will benefit from preclinical studies with animal models that can be readily adapted for humans. We have already collected accelerometric and electroencephalographic data showing the feasibility of the project on OCD patients