Abstract: Objective The purpose of this study was to explore the diagnostic performance of FibroScan, magnetic resonance imaging proton density fat-fraction (MRI-PDFF) and FibroScan-AST(FAST)score in judging patients with high risk non-alcoholic steatohepatitis (NASH) from those with non-alcoholic fatty liver diseases (NAFLD). Mthods A total of 107 patients with NAFLD were encountered in our hospital between June 2017 and December 2021, and all patients underwent liver biopsies. FibroScan, MRI-PDFF and serological detection were completed and three non-invasive models of FAST, FIB-4 and APRI were calculated. Univariate and multivariate Logistic regression analysis was used to screen out factors impacting high-risk NASH. The diagnostic performance of relevant parameters and three non-invasive models to identify high risk NASH was analyzed by ROC curve. Result Of 107 patients with NAFLD, the histo-pathological examination showed high risk NASH in 13 cases (12.1%), and simple alcoholic fatty liver and non-high risk NASH in 94 cases (87.9%); liver stiffness measurement (LSM) by Fibroscan, FAST score and aspartate aminotransferase/platelets (APRI) in patients with high risk NASH were significantly higher than in those with non-high risk NASH (P＜0.05);multivariate Logistic regression analysis showed that only the LSM was the independent risk factor impacting high risk NASH(P＜0.05);ROC analysis demonstrated that the LSM, FAST and APRI could identify high risk NASH(P＜0.05), with AUCs of 0.795, 0.713 and 0.682, and the LSM got the optimal diagnostic efficacy, with sensitivity (Se) and specificity (Sp) of 92.3%and 54.3% based on exclusion cut-off-value, and with Se and Sp of 53.8%and 90.4% based on inclusion cut-off-value. Conclusion The simultaneous increased LSM in patients with NAFLD hints existence of NASH, which might help clinicians make appropriate measures to tackle it.

Key words: Nonalcoholic Steatohepatitis, FibroScan, Magnetic resonance imaging proton density fat-fraction, FAST score, Diagnostic test