Partial Diffusion for Super-Resolution of Quantitative MRI to Accelerate Acquisition

Shishuai Wang, Tessel Huibregtsen, Alireza Samadifardheris, Noemi Sgambelluri, Stefan Klein, Florian Wiesinger, Juan A. Hernandez-Tamames, Dirk Poot

Keywords: Quantitative MRI, Diffusion Models, Partial Diffusion, Super-Resolution

TL;DR: We use partial diffusion for super-resolution of low-quality subsampled qMRI maps, thus allowing further accelration of acquisition.

Abstract: Rapidly acquired 3D quantitative MRI (qMRI) holds great potential in medical imaging. However, accelerating the acquisition process often compromises image quality. In this work, we present a super-resolution method that enhances low-resolution qMRI maps using partial diffusion. We trained and evaluated our approach on in vivo data acquired with a 3D, silent, Multi-Parametric Zero TE MRI sequence (MuPa-ZTE). The results demonstrate that our method improves image quality of the four-fold accelerated acquisition.

Submission Number: 58