Abstract
Measures from diffusion magnetic resonance imaging reflect changes in the substantia nigra of Parkinson's disease. It is the case, however, that partial volume effects from free water can bias diffusion measurements. The bi-tensor diffusion model was introduced to quantify the contribution of free water and eliminates its bias on estimations of tissue microstructure. Here, we test the hypothesis that free water is elevated in the substantia nigra for Parkinson's disease compared with control subjects. This hypothesis was tested between large cohorts of Parkinson's disease and control participants in a single-site study and validated against a multisite study using multiple scanners. The fractional volume of free water was increased in the posterior region of the substantia nigra in Parkinson's disease compared with control subjects in both the single-site and multi-site studies. We did not observe changes in either cohort for free-water-corrected fractional anisotropy or free-water-corrected mean diffusivity. Our findings provide new evidence that the free-water index reflects alteration of the substantia nigra in Parkinson's disease, and this was evidenced across both single-site and multi-site cohorts.