Enhanced Quantitative Susceptibility Mapping of Brain at 0.53T through Locally Low-Rank Denoising

Quantitative susceptibility mapping (QSM), while well-established at high-field strengths (⩾ 1.5T), remains underexplored at lower fields. This study evaluates QSM performance at 0.53T and demonstrates the efficacy of advanced denoising techniques. Using a 3D multi-echo GRE sequence, we acquired in vivo brain images and processed them with Robust Locally Low-Rank (RLLR) and random matrix theory (RMT) denoising methods. Quantitative analysis revealed significant improvements in image quality and reduced noise levels in susceptibility maps derived from denoised data. This optimized low-field QSM approach holds promise for expanding magnetic susceptibility imaging in resource-limited settings and portable MRI applications.