Correcting temperature related diffusivity drift for postmortem diffusion MRI

Motivation: In diffusion MRI, large post-mortem samples reach steady-state temperature and diffusivity only after a very long scan time. Goal(s): Our goal was to use data during steady-state temperature adaptation to model changing diffusivity and still obtain reliable tissue orientations. Approach: We estimated the diffusivities at steady-state temperature with a subset of the data and computed the temperature correction coefficients from the difference between predicted and measured data. Results: The corrected data has greatly increased test-retest agreement and angular accuracy of fiber orientation estimates. Impact: Accelerate post-mortem diffusion MRI by replacing dummy scans with our new data-driven temperature modeling.