High-quality Brain MR Fingerprinting Based on Latent Manifold Structure Regularization

Motivation: Estimating high-quality parameter maps from highly undersampled measurements presents one of the major challenges in MR fingerprinting (MRF). Goal(s): To propose a novel MRF reconstruction framework based on manifold structured data priors for high-quality parameter maps estimation. Approach: We propose a novel MRF reconstruction framework leveraging manifold structured data priors to improve the accuracy of the reconstructed parameter maps. Additionally, we integrate a locally low-rank prior into the reconstruction framework to exploit local correlations within each patch and further enhance reconstruction performance. Results: Experimental results demonstrate that our method can achieve significantly improved reconstruction performance with reduced computational time over the state-of-the-art methods. Impact: By exploiting the manifold structure prior of MRF data, our method can better reconstruct detailed textures and provide accurate brain tissue characterization, thereby improving the diagnostic accuracy in clinical applications.