Comparison of Two Approaches for Diffusional Kurtosis Inference : Synthetic Q-space Learning and DWI Denoising [Presidential Award Proceedings]

Abstract

　We examined the diffusional kurtosis (K) inference by synthetic Q-space learning (synQSL), and the effect of diffusion-weighted image (DWI) denoising. In this study, we compared the results of two approaches for K inference : synQSL and least-squares fitting in denoised DWI. For synSQL, the effect of bias correction was also examined. Through experiments using 3.0T MRI data of healthy volunteers, we confirmed the superiority of the synQSL approach with bias correction.