Uncertainty quantification of parenchymal tracer distribution using random diffusion and convective velocity fields

Area of Science:

• Neuroscience
• Biophysics
• Computational Biology

Background:

• Brain substance transport involves diffusion and convection, with unclear relative importance.
• Accurate modeling is hindered by parameter uncertainty and wide literature value variations.
• This study quantifies tracer distribution variability due to diffusion and convection model parameter uncertainty.

Purpose of the Study:

• To rigorously quantify tracer distribution variability in the brain parenchyma.
• To assess the impact of diffusion and convection model parameters on tracer transport.
• To compare simulation results with experimental MRI data.

Main Methods:

• Simulated intrathecal tracer distribution using the convection-diffusion-reaction equation.
• Tested various diffusion models and velocity fields to assess uncertainty and magnitude effects.
• Compared simulation outcomes with experimental MRI tracer enhancement data.

Main Results:

• Pure diffusion models showed slow white matter penetration (peak >24h).
• Glymphatic system models qualitatively mimicked diffusion but reduced variability.
• Directional glymphatic flow reduced peak time to 11h; direct parenchymal drainage to 6-8h.

Conclusions:

• Diffusion alone cannot explain deep white matter tracer transport observed experimentally.
• Gmphatic velocity fields, especially with directional structure, can enhance brain substance transport.
• Accurate modeling requires considering convective transport mechanisms beyond simple diffusion.