A Flow-based Truncated Denoising Diffusion Model for super-resolution Magnetic Resonance Spectroscopic Imaging

Siyuan Dong, Zhuotong Cai, Gilbert Hangel, Wolfgang Bogner, Georg Widhalm, Yaqing Huang, Qinghao Liang, Chenyu You, Chathura Kumaragamage, Robert K. Fulbright, Amit Mahajan, Amin Karbasi, John A. Onofrey, Robin A. de Graaf, James S. Duncan

Abstract: Highlights•Designed a novel diffusion model for performant and efficient super-resolution MRSI.•Facilitated multi-scale super-resolution with network conditioning.•Introduced a temperature parameter for diffusion models to adjust image sharpness.•Developed the first in vivo MRSI dataset for training deep learning networks.•Neuroradiologists assessed super-resolved metabolic maps from clinical perspectives.