Physics informed guided diffusion for accelerated multi-parametric MRI reconstruction

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Mayo, Perla, Pirkl, Carolin M., Achim, Alin, Menze, Bjoern, Golbabaee, Mohammad
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866913917615210496
author Mayo, Perla
Pirkl, Carolin M.
Achim, Alin
Menze, Bjoern
Golbabaee, Mohammad
author_facet Mayo, Perla
Pirkl, Carolin M.
Achim, Alin
Menze, Bjoern
Golbabaee, Mohammad
contents We introduce MRF-DiPh, a novel physics informed denoising diffusion approach for multiparametric tissue mapping from highly accelerated, transient-state quantitative MRI acquisitions like Magnetic Resonance Fingerprinting (MRF). Our method is derived from a proximal splitting formulation, incorporating a pretrained denoising diffusion model as an effective image prior to regularize the MRF inverse problem. Further, during reconstruction it simultaneously enforces two key physical constraints: (1) k-space measurement consistency and (2) adherence to the Bloch response model. Numerical experiments on in-vivo brain scans data show that MRF-DiPh outperforms deep learning and compressed sensing MRF baselines, providing more accurate parameter maps while better preserving measurement fidelity and physical model consistency-critical for solving reliably inverse problems in medical imaging.
format Preprint
id arxiv_https___arxiv_org_abs_2506_23311
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Physics informed guided diffusion for accelerated multi-parametric MRI reconstruction
Mayo, Perla
Pirkl, Carolin M.
Achim, Alin
Menze, Bjoern
Golbabaee, Mohammad
Image and Video Processing
Machine Learning
Medical Physics
We introduce MRF-DiPh, a novel physics informed denoising diffusion approach for multiparametric tissue mapping from highly accelerated, transient-state quantitative MRI acquisitions like Magnetic Resonance Fingerprinting (MRF). Our method is derived from a proximal splitting formulation, incorporating a pretrained denoising diffusion model as an effective image prior to regularize the MRF inverse problem. Further, during reconstruction it simultaneously enforces two key physical constraints: (1) k-space measurement consistency and (2) adherence to the Bloch response model. Numerical experiments on in-vivo brain scans data show that MRF-DiPh outperforms deep learning and compressed sensing MRF baselines, providing more accurate parameter maps while better preserving measurement fidelity and physical model consistency-critical for solving reliably inverse problems in medical imaging.
title Physics informed guided diffusion for accelerated multi-parametric MRI reconstruction
topic Image and Video Processing
Machine Learning
Medical Physics
url https://arxiv.org/abs/2506.23311