Assessment of multi-parameter dual-energy CT in predicting muscle invasion in bladder cancer: comparison with VI-RADS.

objectivesPreoperative differentiation of muscle-invasive bladder cancer (MIBC) is challenging. This study explores the application of the Dual-Energy CT (DECT) model and the vesical imaging reporting and data system (VI-RADS) in assessing MIBC and compares their diagnostic performance. MATERIALS AND

methodsThis single-center prospective study included 105 patients (33 MIBC and 72 non-MIBC cases). Two radiologists independently performed DECT morphological assessment, and evaluated quantitative DECT parameters and VI-RADS scores, blinded to the pathological findings. A DECT-based model was constructed by integrating independent quantitative predictors with tumor diameter (D). Diagnostic performance was compared using the area under the receiver operating characteristic curve (AUC). Subgroup analyses were performed for equivocal VI-RADS 3-4 cases and small lesions (D < 3 cm).

resultsVI-RADS achieved an AUC of 0.926 (sensitivity 0.970) but low specificity (0.542). The DECT model showed moderate performance (AUC = 0.761). However, the DECT + D model achieved an AUC of 0.925, comparable to VI-RADS (p > 0.05), with improved specificity (0.833). Crucially, in equivocal VI-RADS 3-4 cases, DECT + D (AUC = 0.904) significantly outperformed VI-RADS (AUC = 0.652, p < 0.05), while standalone DECT achieved an AUC of 0.824. For D < 3 cm, DECT + D (AUC = 0.895) and DECT (AUC = 0.857) were comparable to VI-RADS (AUC = 0.866), but DECT maintained high specificity (0.934). A nomogram and web-based risk calculator were developed.

conclusionThe DECT-based model achieves diagnostic accuracy comparable to VI-RADS and demonstrates superior stability in clinically challenging subgroups. It serves as a robust, objective alternative for preoperative bladder cancer staging, particularly beneficial for patients with MRI contraindications. CRITICAL RELEVANCE STATEMENT: This study demonstrates that a quantitative DECT-based model provides an accurate, objective alternative to MRI for preoperative bladder cancer staging, mitigating diagnostic ambiguity in clinically equivocal cases and ensuring accurate risk stratification in small lesions. KEY POINTS: Preoperative differentiation of muscle invasion is critical for bladder cancer management. The DECT-based model yields diagnostic accuracy statistically comparable to VI-RADS. The DECT-based model improves diagnostic confidence in VI-RADS 3-4 lesions.