A novel computerized algorithm to detect microstructural brainstem pathology in Parkinson’s disease using standard 3 Tesla MR imaging

Abstract

Increased deposition of α-synuclein in Parkinson's disease (PD) is known to be prominent in the brainstem and discussed to be clinically relevant for motor and non-motor features. Whether structural magnetic resonance imaging is capable to detect degraded tissue microstructure caused by increased deposition of α-synuclein at this predilection site in PD remains unclear. We hypothesize that microstructural degradation in the brainstem leads to a reduced T1 contrast provoking standard tissue segmentation engines to misclassify tissue as additional grey matter in regions predominantly composed of white matter. High-resolution T1-weighted three-dimensional magnetization prepared rapid gradient echo (MPRAGE) imaging at 3 Tesla in fifty-two PD patients with mild-to-moderate disease severity and in forty age- and gender-matched healthy controls was performed. A dedicated computerized algorithm that comprises standard tissue segmentation in combination with a statistical test was set up that evaluates grey matter composition on voxel level. The algorithm detected a single significant cluster of voxels with enhanced grey matter (cluster volume is 1,368 mm(3), p < 0.05 corrected for false discovery rate) in the pontomedullary junction of the brainstem in PD patients as compared to healthy controls. Furthermore, absolute grey matter volume was significantly higher in the brainstem of the PD group compared to healthy controls. We conclude that this cluster may reflect α-synuclein induced microstructural brainstem pathology in PD.