Evaluation of the Time Difference Method in Identifying Hepatocellular Carcinoma in Current CEUS LR-M Category Nodules

Abstract

ObjectiveTo explore a potential method for improving the diagnostic detection hepatocellular carcinoma (HCC) based on the contrast-enhanced ultrasound (CEUS) liver imaging reporting and data system (LI-RADS) version 2017. MethodsWe retrospectively evaluated 585 liver nodules in 427 patients with HCC risk from December 2020 to March 2023. The nodules were categorized as LR-1 to LR-M based on CEUS LI-RADS version 2017 and were randomly subclassified into a developmental cohort (DC) and a validation cohort (VC) at 3:1. In the DC, the cutoff value of the time difference (∆T) for differentiating HCC from other malignancies in LR-M was calculated, which was used to reclassify nodules in the VC. The diagnostic effect on HCC detection before and after reclassification was further assessed. ResultsAccording to the current CEUS LI-RADS, 140 of 426 (32.9%) confirmed HCC nodules were misclassified as LR-M. In the DC (439 nodules), the receiver operating characteristic (ROC) curve showed that the cutoff value of ∆T (washout onset time minus contrast arrival time) recommended for HCC diagnosis was greater than 21 s. In the VC (146 nodules), 34 HCCs were correctly categorized as LR-5 according to the cutoff value and after reclassification LR-5 had higher accuracy (67.1% vs. 89.0%, p < 0.001) and sensitivity (56.0% vs. 87.2%, p < 0.001) for HCC diagnosis with high specificity (100% vs. 94.6%, p = 0.500) remained. ConclusionUsing the time difference method could identify HCC nodules misdiagnosed as LR-M and improve the diagnostic performance for current CEUS LI-RADS.