A comparison of intravoxel incoherent motion (IVIM) fitting models in the liver
Alexander D. Cohen 1 , Mark D. Hohenwalter 1 , and Kathleen M. Schmainda 1,2
Radiology, Medical College of Wisconsin,
Milwaukee, WI, United States,
Medical College of Wisconsin, Milwaukee, WI, United
The best model for fitting the IVIM signal is still
unknown. Previous studies have used several different
models and achieved different results. This study
compared three models: full, segmented, and Bayesian, in
terms of parameter value and repeatability in normal
liver parenchyma. For pseudodiffusion, all three models
significantly differed with each other. The Bayesian
model resulted in the highest pseudodiffusion, followed
by the full model and then the segmented model.
Repeatability was comparable for the full and segmented
models. The Bayesian model had worse repeatability for
fractional perfusion and pseudodiffusion, despite less
noisy parametric maps.
This abstract and the presentation materials are available to members only;
a login is required.