Abstract: Objective This study aimed to compare the efficacy of computed tomography (CT) and magnetic resonance imaging (MRI) in the diagnosis of patients with hepatocellular carcinoma (HCC) based on liver imaging report data system (LI-RADS) v2018. Methods We performed a retrospective study in 72 patients with high risk of HCC (proven pathologically HCC in 53, non-HCC malignant tumors in 10 and benign lesions in 9), and all patients underwent CT and MRI scan between September 2017 and July 2020. Two independent radiologists assigned a LI-RADS category for all hepatic foci in consensus. The intraclass agreement for the LI-RADS category between CT and MRI was evaluated by using Kappa statistic. The area under receiver operating characteristic curve (AUC) was analyzed for diagnostic accuracy of LI-RADS v2018. Results The intraclass agreement for the LI-RADS category between CT and MRI was substantial, and the kappa value was 0.693 (95% CI: 0.545-0.841), P＜0.001]; the AUCs of the two modalities were 0.827 (95% CI: 0.708-0.946) and 0.856 (95% CI: 0.761-0.952), respectively, without significant difference(P＞0.05); the sensitivities of CT and MRI in diagnosing HCC based on the LR-5 were significantly different, while the specificities of CT and MRI were comparable; among the major features defined by LI-RADS, the display rate of enhanced capsule of tumors by MRI was 40.3%, significantly higher than 5.5% by CT (P ＜0.001). Conclusion Based on the LI-RADS v2018, the CT and MRI show a considerable efficacy in diagnosing patients with HCC, and the sensitivity of MRI is high with obvious presentation of enhanced capsule of tumors.

Key words: Hepatocellular carcinoma, Tomography, Magnetic resonance imaging, Liver imaging reporting data system, Diagnosis