Subtyping Early Parkinson’s Disease by Mapping Cognitive Profiles to Brain Atrophy with Visual MRI Ratings

TL;DR

This study identifies distinct cognitive subtypes in early Parkinson’s disease by linking brain atrophy patterns to cognitive profiles, offering a practical classification model for better patient management.

Contribution

A novel classification model for early Parkinson’s disease subtyping based on cognitive profiles and visual MRI ratings is proposed.

Findings

Eight anatomocognitive subtypes were identified, each reflecting distinct patterns of brain vulnerability.

MRI markers correlated with cognitive deficits in domains like executive function, memory, and language.

Cluster analyses validated subtype reliability with AUC scores ranging from 0.68 to 0.95.

Abstract

Background: Cognitive heterogeneity in Parkinson’s disease (PD) remains a diagnostic and prognostic challenge, particularly in early stages. In this cross-sectional study, we aimed to identify clinically relevant cognitive subtypes in early PD by integrating neuropsychological profiles with regional brain atrophy assessed via visual MRI scales. Methods: Eighty-one de novo PD patients (≤36 months from diagnosis) and twenty healthy controls underwent 3T MRI with visual atrophy ratings and completed an extensive neuropsychological battery. Results: Using a mixed a priori–a posteriori approach, we defined eight anatomocognitive subtypes reflecting distinct patterns of regional vulnerability: frontosubcortical, posterior cortical, left/right hippocampal, global, and preserved cognition. Specific MRI markers correlated with cognitive deficits in executive, visuospatial, memory, and language domains. Cluster analyses supported subtype validity (AUC range: 0.68–0.95). Conclusions: These results support a practical classification model linking cognitive performance to brain structural changes in early PD. This scalable approach may improve early patient stratification and guide personalized management strategies. Longitudinal studies are needed to assess progression patterns and therapeutic implications.

Figures