Applying Neural Imaging and ML to OCD Severity Prediction

Abstract

Obsessive compulsive disorder (OCD) is a mental illness that causes recurrent unwanted obsessions to repeat themselves again and again. In this scientific era of technology and research, we are afraid of these mental disorders. OCD isn't just about nail-biting and negative thinking. Washing your hands several times after touching something contaminated could be a habit. Even though you may not want to think or act in this way, you feel powerless to change your behavior. In the past few years, structural magnetic resonance imaging (MRI) scans of the brain have been used in conjunction with machine learning techniques to distinguish between patients and healthy controls on an individual basis. In spite of this, these methods are rarely used to predict how bad psychiatric symptoms will be. In this study, support vector regression (SVR) is used to look at 37 adults with OCD who hadn't had any treatment yet. The goal was to find out if gray matter (GM) volumes that include cortical-subcortical loops can be used to predict the severity of OCD symptoms. Because of the small sample size, our findings need to be confirmed using samples from other unrelated studies. SVR analysis, a type of machine learning, may be beneficial for discovering neurobiological indicators that predict OCD severity using structural MRI datasets.