Brain age gap estimated from MRI based deep learning model was associated with cognitive impairment in Parkinson’s disease

Motivation: About 10-20% of newly diagnosed Parkinson's disease (PD) patients have cognitive deficient. Objectively evaluate the degree of cognitive impairment is important for PD management. Goal(s): Investigate the association between cognitive impairment and brain age gap (BAG) estimated from MRI-based deep learning model in PD patients. Approach: 53 PD and 46 healthy controls (HCs) were enrolled. BAG and other MRI-based whole brain anatomical features, including whole brain volume, gray matter volume, white matter volume, and brain parenchymal fraction were analyzed. Results: BAG and gray matter volume were significantly associated with MoCA. BAG was independent predictor of cognitive impairment in PD patients. Impact: BAG can be used to objectively estimate cognitive impairment in PD patients. The application of deep learning model to accurately and robustly predict brain age would be helpful for the management of PD patients.