Super-Resolution DL Reconstruction Boosts Edge Sharpness and Lesion Conspicuity in Liver MRI

Super-resolution deep learning reconstruction (SR-DLR) improved edge sharpness, lesion-to-liver contrast, and conspicuity vs conventional/standard DLR in liver MRI (P<0.001), but gains in malignant lesions were modest.

• Liver SNR was similar between SR-DLR and standard DLR.
• SR-DLR had slightly higher image noise and motion artifacts than DLR.
• Improvements in contrast ratio were primarily due to cystic lesions; malignant lesion improvement was modest.

Automated summary

RadPigeon summaries are original and for information only. They are not clinical advice.