Motivation: Current cardiac T1 mapping technique suffers from long acquisition time and its sensitivity to noise or motion artifacts. Goal(s): To reduce the acquisition time and to improve the robustness against motion or noise artifacts in cardiac T1 mapping. Approach: A deep learning framework integrated with Trans-Unet and a fully connected network was developed to realize T1 mapping with T1 weighted images acquired within three cardiac cycles. Results: The method combined the advantages of current deep learning methods and achieved comparable accuracy to MOLLI, with shortened image acquisitions and enhanced robustness. Impact: The approach achieves cardiac T1 mapping within three cardiac cycles, clinically reducing the breath-hold time and causes less discomfort to patients. Meanwhile, the integrated network takes advantages of current deep learning methods and significantly improves the reconstruction quality.
Li et al. (Tue,) studied this question.