Usefulness of the Contrast‐Enhanced Ultrasound Liver Imaging Reporting and Data System in Diagnosing Focal Liver Lesions by Inexperienced Radiologists

OBJECTIVES: To evaluate the usefulness of the contrast-enhanced ultrasound (CEUS) Liver Imaging Reporting and Data System (LI-RADS) in diagnosing focal liver lesions (FLLs) by inexperienced radiologists. METHODS: Images and clinical data from 258 patients at risk for hepatocellular carcinoma who underwent CEUS were collected retrospectively. Two trained inexperienced radiologists and 2 experienced radiologists reviewed all CEUS clips. Each inexperienced radiologist assigned a CEUS LI-RADS category for each observation and labeled it benign or malignant independently. Each experienced radiologist labeled each lesion malignant or benign independently using a conventional diagnostic method. Interobserver agreement of CEUS LI-RADS was analyzed by the kappa test. The overall diagnostic accuracy of the LI-RADS category and conventional diagnosis was described by the sensitivity, specificity, positive predictive value, and negative predictive value. All test results were considered significant at P < .05. RESULTS: A kappa value of 0.774 indicated that the CEUS LI-RADS algorithm resulted in substantial consistency between the inexperienced radiologists. For the diagnosis of hepatocellular carcinoma, the sensitivity, specificity, positive predictive value, and negative predictive value were improved significantly in inexperienced radiologists using the CEUS LI-RADS compared to conventional methods. The overall diagnostic accuracy of the experienced radiologists was almost equal to that of CEUS LI-RADS categories assigned by the inexperienced radiologists. CONCLUSIONS: The CEUS LI-RADS algorithm can not only obtain substantial consistency among inexperienced radiologists but also have excellent diagnostic efficacy in the differentiation of benign from malignant FLLs compared to conventional methods. As a comprehensive algorithm, the CEUS LI-RADS can act as a guide for trainees in learning how to diagnose FLLs.