Related Experiment Video
Updated: Jul 10, 2026
08:41
Novel In Vivo Micro-Computed Tomography Imaging Techniques for Assessing the Progression of Non-Alcoholic Fatty Liver Disease
Published on: March 24, 2023
Leveraging a vendor-neutral deep learning reconstruction algorithm to reduce scan time and enhance image quality in
Boryeong Jeong1, Seung-Seob Kim2
1Department of Radiology and Research Institute of Radiological Science, Center for Clinical Imaging Data Science, Severance Hospital, Yonsei University College of Medicine, 50-1 Yonsei-ro, Seodaemun-gu, Seoul, 03722, Republic of Korea.
Scientific Reports
|July 8, 2026
Summary
Vendor-neutral deep learning reconstruction (DLR) improved accelerated liver MRI T2-weighted imaging (T2WI) quality. Single breath-hold DLR enhanced sharpness and signal-to-noise ratio (SNR) compared to routine scans.
Area of Science:
- Radiology
- Medical Imaging
- Artificial Intelligence in Medicine
Background:
- Accelerated T2-weighted imaging (T2WI) in liver MRI aims to reduce scan time.
- Routine T2WI often requires multiple breath-holds, potentially impacting patient comfort and workflow.
- Deep learning reconstruction (DLR) offers a potential solution for improving accelerated MRI quality.
Purpose of the Study:
- To evaluate if vendor-neutral deep learning reconstruction (DLR) can enhance image quality in accelerated single breath-hold T2WI for liver MRI.
- To compare the image quality of accelerated T2WI reconstructed with DLR against routine T2WI acquired with multiple breath-holds.
Main Methods:
- Retrospective analysis of 86 patients undergoing nonenhanced 3-T liver MRI.
- Comparison of three T2WI protocols: routine two breath-holds (Routine-2BH), accelerated single breath-hold with DLR (SwiftMR-1BH), and accelerated two breath-holds with DLR (SwiftMR-2BH).
- Assessment of quantitative SNR and qualitative image parameters including sharpness, motion artifacts, spatial mismatch, and lesion conspicuity by two radiologists.
Main Results:
- SwiftMR protocols significantly reduced acquisition time by ~9 seconds.
- Both SwiftMR protocols demonstrated significantly higher sharpness and SNR compared to Routine-2BH (p < 0.001).
- SwiftMR-1BH showed significantly lower spatial mismatch scores than Routine-2BH and SwiftMR-2BH (p < 0.05), with generally improved lesion conspicuity.
Conclusions:
- Vendor-neutral DLR using SwiftMR effectively restores and enhances image quality in accelerated single breath-hold liver MRI T2WI.
- Accelerated T2WI with DLR achieves higher SNR and overall image quality compared to conventional multi-breath-hold T2WI.
- DLR holds promise for improving efficiency and diagnostic performance in liver MRI.