Diagnostic Performance of PDFF, MRE, and T1 Mapping for Detection of NASH and streamline MRI protocol

Motivation: NASH identification is significant, however lacking reliable and concise preoperative assessment. Goal(s): To evaluate performance of MRE, MRI-PDFF and T1 mapping in diagnosing NASH and establish a streamlined MRI protocol to diagnosis NASH more efficiently. Approach: Patients with suspected or diagnosed NAFLD who had liver biopsy and multiparametric MRI/MRE within 30 days were enrolled. Imaging parameters (LSM, PDFF, R2*, T1 mapping) were compared across NAFLD stages using appropriate statistical tests. Prediction models for NASH diagnosis were developed using combinations of these parameters. Results: The two-parameter model using PDFF and MRE-based LSM is sufficient for a streamlined non-contrast liver imaging protocol to NASH diagnosis. Impact: As an active form of NAFLD, NASH can accelerate fibrosis, leading to end-stage liver disease and hepatocellular carcinoma. A streamlined non-contrast liver imaging protocol consisting of PDFF and MRE-based LSM can diagnose NASH satisfactorily and cost-effectively.