Intrathecal contrast-enhanced magnetic resonance imaging (MRI), utilizing the contrast agent gadobutrol as cerebrospinal fluid (CSF) tracer is emerging as a useful method to study glymphatic function in the human brain (Ringstad 2017, Ringstad 2018). A consistent finding with this technique is large inter-individual variability regarding tracer propagation, which complicates the assessment of brain health and the design of intrathecal drug delivery. In this talk, we outline an approach which predicts the distribution of tracer in the brain based only on structural information captured by MRI, addressing one possible explanation for this variability. Our method utilizes diffeomorphic image registration (Avants 2008) to compute deformation fields from a dataset of prior patients to a target subject based on pre-injection MRI. These mappings are applied to post-injection images to perform voxel-wise predictions of tracer enrichment. In a previous work (Solheim 2025), we validated this approach on a heterogeneous dataset of 134 patients with various CSF disorders, focusing on signal enrichment in the parenchyma at the 24 hour peak. In this talk, we extend this work to a new dataset of 30 subjects, including patients with Parkinson's disease and a healthy control group. This extension incorporates longitudinal data across multiple time-points and expands the analysis to include enrichment in the CSF, utilizing an improved T_1 mapping (Storås 2025) to estimate actual tracer concentrations rather than signal intensity changes. We demonstrate that geometry-based predictions correlate strongly with observed enrichment, and investigate the ability of the method to predict tracer distribution across pathologies. This approach provides a verifiable pathway to isolate the role of brain structure in solute transport, complementing existing physics-based and machine learning models.