Differential Diagnosis of Parkinson’s disease, Progressive Supranuclear Palsy and Corticobasal syndromes using machine learning and MRI

In this study we combined machine learning with MRI for the differential diagnosis of three movement disorders: Parkinson’s disease (PD), progressive supranuclear palsy (PSP) and degenerative corticobasal syndrome (CBS). We compared the performance of such approaches when using T1-weighted and diffusion MRI, as well as different methods for feature extraction. Our results suggest that such methods could be used in the future to aid the differential diagnosis of PSP, CBS and PD, in conjunction with clinical assessment, with diffusion MRI data providing the most promising results.