q3-MuPa: Quick, Quiet, Quantitative Multi-Parametric MRI using Physics-Informed Diffusion Models

Graphical Abstract q3-MuPa: Quick, Quiet, Quantitative Multi-Parametric MRI using Physics-Informed Diffusion Models Shishuai Wang, Florian Wiesinger, Noemi Sgambelluri, Carolin Pirkl, Stefan Klein, Juan A. Hernandez-Tamames, Dirk H.J. Poot arXiv:2512.23726v1 [physics.med-ph] 19 Dec 2025 Highlights q3-MuPa: Quick, Quiet, Quantitative Multi-Parametric MRI using Physics-Informed Diffusion Models Shishuai Wang, Florian Wiesinger, Noemi Sgambelluri, Carolin Pirkl, Stefan Klein, Juan A. Hernandez-Tamames, Dirk H.J. Poot • The proposed qMRI mapping method leverages diffusion model and physics information. • We tailored our method for a novel 3D silent qMRI sequence (MuPa-ZTE) • Our method is trained on synthetic data and gener- alises well to real data. • Our method provides high accuracy and superior visual performance. • It is feasible to get high-quality 3D qMRI maps from approximate 1-minute scan. q3-MuPa: Quick, Quiet, Quantitative Multi-Parametric MRI using Physics-Informed Diffusion Models Shishuai Wanga, Florian Wiesingerb,c, Noemi Sgambelluria, Carolin Pirklb, Stefan Kleina, Juan A. Hernandez-Tamamesa,d, Dirk H.J. Poota aDepartment of Radiology and Nuclear Medicine, Erasmus MC, Rotterdam, The Netherlands bGE HealthCare, Munich, Germany cDepartment of Neuroimaging, Institute of Psychiatry, Psychology & Neuroscience, King’s College London, London, United Kingdom dDepartment of Imaging Physics, TU Delft, Delft, The Netherlands Abstract The 3D fast silent multi-parametric mapping sequence with zero echo time (MuPa-ZTE) is a novel quantitative MRI (qMRI) acquisition that enables nearly silent scanning by using a 3D phyllotaxis sampling scheme. MuPa-ZTE im- proves patient comfort and motion robustness, and generates quantitative maps of T1, T2, and proton density using the acquired weighted image series. In this work, we propose a diffusion model-based qMRI mapping method that leverages both a deep generative model and physics-based data consistency to further improve the mapping perfor- mance. Furthermore, our method enables additional acquisition acceleration, allowing high-quality qMRI mapping from a fourfold-accelerated MuPa-ZTE scan (approximately 1 minute). Specifically, we trained a denoising diffusion probabilistic model (DDPM) to map MuPa-ZTE image series to qMRI maps, and we incorporated the MuPa-ZTE forward signal model as an explicit data consistency (DC) constraint during inference. We compared our mapping method against a baseline dictionary matching approach and a purely data-driven diffusion model. The diffusion models were trained entirely on synthetic data generated from digital brain phantoms, eliminating the need for large real-scan datasets. We evaluated on synthetic data, a NISM/ISMRM phantom, healthy volunteers, and a patient with brain metastases. The results demonstrated that our method produces 3D qMRI maps with high accuracy, reduced noise and better preservation of structural details. Notably, it generalised well to real scans despite training on syn- thetic data alone. The combination of the MuPa-ZTE acquisition and our physics-informed diffusion model is termed q3-MuPa, a quick, quiet, and quantitative multi-parametric mapping framework, and our findings highlight its strong clinical potential. Keywords: Quantitative MRI, Diffusion Models, Multi-parametric Mapping, Deep Generative Models 1. Introduction Magnetic Resonance Imaging (MRI) is a widely used non-invasive imaging technique that provides superior soft-tissue contrast and detailed anatomical or func- tional information. However, conventional MRI pro- duces weighted images whose voxel intensities have no standardised physical meaning, complicating compar- isons across different scanner settings or sites. In con- trast, quantitative MRI (qMRI) aims to measure the in- trinsic tissue properties (e.g. T1 and T2 relaxation times or proton density), yielding qMRI maps that are com- parable across scans and promising for use as imaging biomarkers. Recent research has focused on 3D fast multi- parametric mapping sequences [1, 2, 3, 4], aiming to in- fer multiple relevant tissue properties within clinically acceptable additional scan time. However, many 3D fast qMRI sequences generate loud acoustic noise due to rapidly switched gradients, which can be problematic for certain patient groups (e.g. children or individuals with hyperacusis). In this work, we instead focus on a novel 3D fast silent multi-parametric mapping sequence with zero echo time (MuPa-ZTE) [5, 6, 7]. By employ- ing a nominal zero echo time and a 3D radial phyllotaxis readout, MuPa-ZTE minimises gradient switching noise (enabling nearly silent scanning) and is motion-robust, while also capturing signal from ultrashort T2 compo- nents normally invisible at longer echo times [6]. The baseline of qMRI mapping for MuPa-ZTE re- constructs a series of weighted images (with varying T1, T2, proton density weightings) and then performs dictionary matching to the underlying

…(Full text truncated)…

This content is AI-processed based on ArXiv data.