Statistical Analysis of the Relationship Between SWI-Derived Features and Brain Stiffness Measured by MRE in Parkinson’s Disease

Parkinson’s disease (PD) is a progressive neurodegenerative disorder characterized by motor impairments and a range of structural, vascular, and biomechanical changes in the brain. While Magnetic Resonance Elastography (MRE) has consistently revealed a progressive softening of brain tissue in PD, the vascular correlates of this phenomenon remain unclear. This study investigates the relationship between brain stiffness and vascular features extracted from Susceptibility Weighted Imaging (SWI), with a specific focus on gray matter (GM) of frontal, parietal, temporal and occipital lobes. To this end, it has been developed a dedicated processing pipeline to generate quantitative venous maps, based on the utilization of Frangi filter method, allowing for the extraction of vascular density, number of branching points, thickness and other vessel-derived features. The results of group-level analysis confirmed the presence of regional brain softening in PD patients more pronounced in the left hemisphere for frontal, temporal and occipital lobes. Through the results of several regression models it was also possible to predict stiffness values of the different regions using SKI-derived features, suggesting a tight interplay between biomechanical and vascular alterations. Notably, patients showed a negative relationship between stiffness and both vascular density and branching points, and a positive one with venous contrast. These findings, together with the reduction of cerebral blood flow (CBF) and the increase of oxygen extraction fraction (OEF) reported in previous studies, suggest that the softening process is accompanied by a potential pathological compensation process (angiogenesis). Interestingly in healthy controls (HC), venous contrast has been found to be consistently negatively related to stiffness, in complete opposition to PD patients. Furthermore, stiffness reductions were associated with higher UPDRS 1 scores, suggesting that structural tissue changes may parallel clinical symptom progression. These findings provide preliminary evidence that vascular remodeling and biomechanical degradation in PD may be interconnected processes. By linking venous architecture metrics to stiffness alterations, the work highlights a strong indication that SKI-derived vascular features could partially explain changes in mechanical properties of PD patients.