The objective was to evaluate the ability of automated technique based on structural, cross-sectional T1-weighted (T1w) MRI at performing differential classification of idiopathic Parkinson’s disease (IPD) patients against those with either progressive supranuclear palsy (PSP) or multiple systems atrophy (MSA). The results have demonstrated that a classification approach based on quantitative parameters of 3D brainstem morphology has the potential to differentiate IPD versus PSP and MSA with high accuracy, thereby reducing the initial clinical error rate. An additional objective was to develop a cortical morphometry (i.e., shape index, curvedness, and regional bending energy) analysis framework to better understand the alteration on the brain structure due to PD and propose an alternate diagnosis framework differentiating IPD to MSA and PSP. A statistical analysis on these cortical features showed the ability to differentiate these populations from cortical shape differences.