Low-Rank Representation Enhanced Implicit Neural Representation for Zero-Shot Reconstruction in Highly Accelerated 3D Multi-parametric MRI

Motivation: Existing 3D multi-parametric quantitative MRI techniques encounter notable challenges in reconstruction under highly undersampled scenarios, limiting their potential in applications. Goal(s): To achieve reconstruction for multi-parametric quantitative maps under high acceleration factors (e.g., a 2.5-minute scan for whole-brain), shortening scan times. Approach: With low-rank representation, preliminary weighted images were estimated from highly undersampled, high-dimensional k-space data and utilized to facilitate the reconstruction of quantitative maps. Results: The results demonstrated that, even under high acceleration factors, our proposed method improved the quality of reconstructed quantitative maps and ensured accurate quantitation. Impact: The proposed framework could reconstruct whole-brain T1, T2, T2* maps within just a 2.5-minute scan. This advancement holds clinical promise for tissue characterization and pathological assessment in neuroscience, and enables probability for higher resolution and additional contrasts in MRI sequences.