To explore the predictive value of dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) combined with spectral CT for microvascular invasion (MVI) of hepatocellular carcinoma.

A retrospective analysis was conducted on the clinical data of 90 patients with hepatocellular carcinoma admitted to Affiliated Hospital of Yangzhou University from January 2022 to December 2024. All patients underwent DCE-MRI combined with spectral CT scan examination. According to the MVI situation in pathological examination, the patients were divided into two groups: the MVI group (n=41) and the non-MVI group (n=49). Two groups of data were collected, the receiver operating characteristic (ROC) curve was constructed, and the AUC was calculated to evaluate the predictive efficacy of DCE-MRI and spectral CT alone and in combination for MVI.

The apparent diffusion coefficient (ADC) and the percentage of extracellular space outside blood vessel (V_e) in the MVI group were lower than those in the non-MVI group, while the rate constant (K_ep), the volume transfer constant (K^trans), the standardized iodine concentration (NIC) in the arterial phase, the slope of the energy spectrum curve (λHU) in the arterial phase, and λHU in the portal vein phase were higher than those in the non-MVI group, with statistically significant differences (P<0.05). Multivariate Logistic analysis showed that ADC (OR=0.099, 95% CI: 0.028-0.348) and V_e (OR=0.145, 95% CI: 0.045-0.469) were independent protective factors for MVI in hepatocellular carcinoma. While K_ep (OR=6.524, 95% CI: 2.341-18.181), K^trans (OR=8.197, 95% CI: 2.594-25.900), and arterial phase NIC (OR=24.915, 95% CI: (5.386-115.274), λHU in the arterial phase (OR=4.678, 95% CI: 1.915-11.429), and λHU in the portal vein phase (OR=2.440, 95% CI: 1.157-5.145) were independent risk factors for MVI in hepatocellular carcinoma (P<0.05). Among them, the OR value of NIC in the arterial phase was the highest, suggesting that it had the greatest predictive value for MVI. ROC curve analysis showed that among each single parameter, the one with the highest predictive efficacy for MVI was arterial phase NIC (AUC=0.873, 95% CI: 0.809-0.937), followed by K^trans (AUC=0.834, 95% CI: (0.761-0.907) and ADC value (AUC=0.812, 95% CI: 0.732-0.892). After constructing a joint prediction model with the above-mentioned independent predictive factors of statistical significance, its predictive efficacy was significantly improved (AUC=0.902, 95% CI: 0.886-0.978), with a sensitivity of 90.24% and a specificity of 87.76%.

DCE-MRI combined with spectral CT is helpful for preoperative prediction of MVI in hepatocellular carcinoma, its diagnostic efficacy is superior to any single mode, providing key information support for the selection of clinical surgical strategies and prognosis evaluation.