Dual-energy CT quantitative parameters can improve the performance of differential diagnostic between ameloblastomas and odontogenic keratocysts with solid components
摘要
To evaluate whether quantitative parameters derived from dual-energy computed tomography (DECT), combined with conventional imaging features, can effectively distinguish between ameloblastomas and odontogenic keratocysts (OKCs) with solid components.
Materials and methodsThis retrospective analysis evaluated patients with pathologically confirmed ameloblastomas or OKCs who underwent multiphase contrast-enhanced dual-energy CT examinations between January 2020 and July 2024. Quantitative DECT parameters and conventional radiological characteristics were compared. Multivariate analysis identified predictors, and the receiver operating characteristic (ROC) curve assessed performance.
ResultsA total of 28 ameloblastomas and 20 OKCs with solid components were evaluated. Ameloblastomas exhibited significantly higher RHO values in the unenhanced, arterial, and venous phases than those in OKCs (p < 0.001 to 0.002), elevated venous-phase Zeff, IC, NIC, and λHU (p < 0.001), larger lesion dimensions, greater expansion (p < 0.008), and pronounced bone resorption (p < 0.001). Conversely, OKCs demonstrated significantly higher FAT and DEI values in the unenhanced and venous phases (p < 0.001 to 0.043). Multivariate logistic regression analysis identified that λHU (95% confidence interval 1.195–1.807, p < 0.001), NIC (95% CI 1.069–1.262, p < 0.001) in the venous phase, and maximum dimension (95% CI 1.05–1.252, p = 0.002) are independent predictors. A combined model incorporating these parameters demonstrated superior diagnostic performance to differentiate ameloblastomas from OKCs (AUC: 0.963; sensitivity: 85.71%; specificity: 100%).
ConclusionsThe combination of DECT quantitative parameters with conventional imaging features significantly improves differentiating ameloblastomas from OKCs with solid components, offering a potential imaging-based diagnostic tool for clinical decision-making.