Quinten Collier1, Arnold Jan den Dekker1,2, Ben Jeurissen1, and Jan Sijbers1
1iMinds Vision Lab, University of Antwerp, Antwerp, Belgium, 2Delft Center for Systems and Control, Delft University of Technology, Delft, Netherlands
Diffusion kurtosis imaging (DKI) suffers from partial volume
effects caused by cerebrospinal fluid (CSF). We propose a DKI+CSF model
combined with a framework to robustly estimate the DKI parameters. Since the
estimation problem is ill-conditioned, a Bayesian estimation approach with a
shrinkage prior is incorporated. Both simulation and real data experiments
suggest that the use of this prior leads to a more accurate, precise and robust
estimation of the DKI+CSF model parameters. Finally, we show that not
correcting for the CSF compartment can lead to severe biases in the parameter
estimations.
