Low Rank Reconstruction Enabled Partial Fourier 2D Dynamic Variable Resolution Hyperpolarized 13C Imaging

Motivation: Low-rank recovery reconstruction enables partial fourier (PF) k-space encoding of 2D dynamic variable-resolution hyperpolarized (HP) [1-13C]pyruvate MRI. Asymmetrical PF encoding reduces T2* weighting, while mitigating the loss of resolution caused by under-sampled k-space. Goal(s): Explore low-rank recovery reconstruction approaches for PF encoding in sparsely sampled matrices (32x32 or 16x16) in 2D dynamic [1-13C]pyruvate MRI. Approach: Retrospective and prospective [1-13C]pyruvate 2D-EPI brain and abdomen datasets were used to evaluate image fidelity of PF low-rank recovery reconstruction approaches. Results: PF encoding enabled by low-rank recovery reconstruction algorithms increased SNR for HP metabolic images, all while maintaining image fidelity. Impact: Asymmetrical PF encoding of 2D dynamic HP [1-13C]pyruvate MRI benefits from low-rank recovery reconstruction methods to ensure image fidelity, and to increase SNR by reducing T2* weighting.