An Optimized Multi-Component Imaging Method on a Homebuilt 0.5 T MRI System: Combing Intra- and Inter-Voxel Constraints

Motivation: Applying the component imaging method to low-field MRI systems will face a main problem: low-SNR image data. Goal(s): An optimized inversion method is proposed, aiming to give better results for image data with low SNR. Approach: This paper proposes an optimized inversion method with the formula of the optimization problem combining intra- and inter-voxel constraints. Results: The optimized method shows a better convergence rate avoiding the fragmentation of component images and the appearance of pseudo peaks in the spectrum. Impact: This multi-component imaging approach can provide sub-voxel characterization and be applied to numerous applications of popular portable low-field MRI systems.