Motivation: Parkinson's disease (PD) exhibits clinical heterogeneity, such as brain-first and body-first, with possible distinct neurodegenerative progression patterns. Goal(s): To identify PD subtypes based on spatiotemporal neurodegeneration patterns using cortical thickness and deep gray matter volumes and investigating on rapid eye movement sleep behavior disorder (RBD) prevalence of each subtype. Approach: Applied a machine learning technique to the brain features from PD patients to uncover subtypes and progression stages. Results: Identified two subtypes: a cortex-first subtype and a deep grey-first subtype; the cortex-first subtype showed higher prevalence of RBD. Impact: Identifying PD subtypes with distinct neurodegeneration patterns enhances understanding of disease heterogeneity, potentially guiding personalized therapeutic strategies and improving prognostic predictions for patients with PD.
Park et al. (Tue,) studied this question.