Combined US and MRI model yields significantly higher diagnostic accuracy for predicting microvascular invasion in HCCCombined Ultrasound and MRI Improve Prediction of Liver Cancer Growth

OBJECTIVE: To compare the diagnostic performance of ultrasound (US) and magnetic resonance imaging (MRI) for predicting microvascular invasion (MVI) in hepatocellular carcinoma (HCC) using a systematic review and head-to-head meta-analysis. MATERIALS AND METHODS: PubMed, Embase, Web of Science, and the Cochrane Library were searched to identify studies assessing both US and MRI for the preoperative prediction of MVI in HCC. Diagnostic accuracy parameters were pooled using a bivariate random-effects model. Meta-regression and sensitivity analyses were performed to explore heterogeneity. RESULTS: Five studies with 771 patients were included. Pooled sensitivity, specificity, and AUC were 0.78 (0.61-0.88), 0.83 (0.74-0.89), and 0.87 (0.84-0.90) for US, and 0.90 (0.53-0.99), 0.83 (0.73-0.90), and 0.89 (0.86-0.91) for MRI, respectively. The combined US + MRI model showed a pooled sensitivity of 0.87 (0.78-0.93), specificity of 0.86 (0.79-0.90), and AUC of 0.93 (0.90-0.95), which was significantly higher than US (P = 0.003) and MRI alone (P = 0.027). No significant differences were found in sensitivity or specificity among three modalities (all P > 0.05). Meta-regression revealed that study design and validation cohort contributed to MRI model heterogeneity. CONCLUSION: US and MRI show favorable diagnostic performance for preoperative MVI prediction in HCC. The combined US + MRI model yields significantly higher diagnostic accuracy, supporting multimodal imaging strategies for noninvasive MVI assessment. However, heterogeneity among studies highlights the need for large-scale prospective validation.