Using Neurite Orientation Dispersion and Density Imaging to Diagnose Data-Driven Parkinson’s Disease Subtypes in Brainstem Nuclei

Motivation: Little is known about whether microstructural changes in the brainstem nuclei can differentiate three novel data-driven PD subtypes defined by a retrospective multidiscipline dataset. Goal(s): To compare various diffusion MRI metrics between patients with three novel data-driven PD subtypes. Approach: Neurite orientation dispersion and density Imaging (NODDI) and diffusion tensor imaging (DTI) metrics were reconstructed to compare three subtypes. Results: Only a NODDI index, free water fraction in the right substantia nigra can distinguish the symptomatically severest subtype from other two subtypes (AUC=0.798). Impact: NODDI assists in diagnosing the data-driven PD subtype with the most severe motor/non-motor comorbidities.