Fast hepatobiliary phase gadoxetate-enhanced imaging under breath-holding utilizing DL reconstruction (Sonic DL): preliminary experience

Motivation: DL Speed (DLS), a modified Sonic DL, is a new deep learning technology, reconstructing high-quality MRI images from under-sampled k-space data, providing minimal image degradation compared to standard methods. Goal(s): To assess the performance of DLS-LAVA for breath-holding hepatobiliary phase (HBP) imaging in comparison with conventional LAVA. Approach: In 20 liver MRI cases, DLS-LAVA and conventional LAVA were evaluated by two radiologists using qualitative and quantitative measures, including a liver-spleen intensity ratio (LSR). Results: DLS-LAVA provided superior overall image quality, sharpness, and aliasing though graininess was slightly less favorable in some highly undersampled settings. DLS-LAVA achieved high-quality HBP images with shorter breath-holding times. Impact: DLS-LAVA enables radiologists to obtain high-quality HBP images with reduced scan time, enhancing patient comfort and diagnostic precision compared to conventional LAVA. This advancement is especially valuable for patients with limited breath-holding capacity.