Multi-modality connectome-based predictive modeling of individualized compulsions in obsessive-compulsive disorder

Area of Science:

• Neuroscience
• Psychiatry
• Machine Learning

Background:

• Obsessive-compulsive disorder (OCD) diagnosis often relies on classifying patients versus controls.
• Predicting the continuous severity of specific OCD symptoms is crucial for clinical management.

Purpose of the Study:

• To apply machine learning for predicting individualized OCD symptom severity.
• To identify specific brain network features associated with OCD compulsions.

Main Methods:

• Connectome-based predictive modeling was used.
• Functional and structural brain networks were constructed from resting-state fMRI and DTI data from 54 OCD patients.

Main Results:

• Individualized compulsions were successfully predicted using both structural and functional brain network models.
• Key brain regions and network interactions (salience, frontoparietal, default mode) were identified as vital predictors.

Conclusions:

• This study provides the first evidence for the predictive role of the triple network model in OCD compulsions.
• Findings have significant implications for OCD diagnosis, prognosis, and personalized treatment strategies.