Accelerated Deuterium Metabolic Imaging of Liver Metastasis at 7T Using Low-Rank and Subspace Reconstruction with Retrospective Undersampling

Motivation: Long acquisition times limit the clinical use of Deuterium Metabolic Imaging (DMI) for liver metastases assessment, as it reduces efficiency and feasibility in routine clinical practice. Goal(s): This study demonstrates that additional acceleration of DMI can be achieved by combining Hamming weighted-acquisition with retrospective undersampling, potentially reducing acquisition times without compromising data quality. Approach: DMI data were acquired with weighted-acquisition and retrospectively undersampled to R=1.8 from patients (hepatic tumours). Low-rank and subspace reconstruction was applied to evaluate SNR and CRLB compared to FFT. Results: The low-rank and subspace reconstruction achieved an approximately 3-fold SNR increase and lower CRLB, enabling practical and accelerated DMI. Impact: This study enables faster DMI of the abdomen by combining Hamming weighted acquisition with low-rank and subspace reconstruction. The results pave the way for advanced metabolic imaging applications, potentially enhancing liver cancer assessment and treatment planning in clinical settings.