Incorporating new criteria can improve the ability of CEUS LI-RADS to differentiate HCC from non-HCC malignancies
摘要
This work aimed to evaluate the diagnostic performance of the CEUS LI-RADS and to further improve its ability to differentiate HCC from non-HCC malignancies by modifying the LR-M criteria.
MethodsThis retrospective study included patients with a high risk of HCC who underwent CEUS examination. A subgroup of LR-M lesions was reassessed by applying different diagnostic criteria, including Criterion 1: early washout < 60 s; Criterion 2: Criterion 1 + marked washout within 4 min; Criterion 3: early washout < 45 s; Criterion 4: washout onset—enhancement onset < 30 s. Diagnostic efficacy was assessed.
ResultsA total of 1681 liver lesions were included. When LR-5 was used as a predictor of HCC, the sensitivities of Criterion 2 (84.3%), Criterion 3 (73.1%), and Criterion 4 (73.5%) were greater than that of Criterion 1 (67.3%; all p < 0.01). The specificities of Criterion 1 (95.1%), Criterion 3 (94.2%), and Criterion 4 (94.6%) were similar and surpassed that of Criterion 2 (73.5%; all p < 0.01). The PPVs of Criterion 1 (97.7%), Criterion 3 (97.5%), and Criterion 4 (97.7%) were similar and surpassed that of Criterion 2 (95.7%; all p < 0.05).
ConclusionThe use of the washout onset time minus the enhancement onset time < 30 s as a new criterion for early washout could help improve the differential ability for HCC and non-HCC malignancies.