DECT and Radiomics Models Predict Muscle Invasion in BCa

Muscle-invasive bladder cancer (BCa) carries a poorer prognosis, often requiring more radical treatment than non-muscle invasive disease. Preoperative staging is crucial to guide management decisions. Dual-energy CT urography (DECTU) is a promising imaging modality for BCa evaluation.

Mengting Hu and the team aimed to develop and evaluate a model combining quantitative DECT parameters and radiomics to predict muscle invasion in BCa preoperatively.

The study included 126 patients with BCa who underwent dual-energy CT urography (DECTU) at the hospital. Patients were randomly divided into training and test cohorts in a 7:3 ratio. Quantitative parameters from DECTU were identified through univariate and multivariate logistic regression analysis to develop a DECT model. Radiomics features were extracted from 40, 70, and 100 keV images, as well as iodine-based material-decomposition (IMD) images in the venous phase.

These features were used to create radiomics models from individual and combined images using a support vector machine classifier, with the best-performing model selected as the final radiomics model. A fusion model was then established, combining the DECT parameters and the radiomics model.

The results demonstrated that the normalized iodine concentration (NIC) in DECT was an independent factor in diagnosing muscle invasion of BCa. The optimal multi-image radiomics model achieved predictive performance with an area under the curve (AUC) of 0.867 in the test cohort, outperforming NIC, which had an AUC of 0.704.

The fusion model showed enhanced performance, with an AUC of 0.893, although this difference was not statistically significant. Additionally, the fusion model demonstrated superior performance in decision curve analysis (DCA).

For lesions smaller than 3 cm, the fusion model exhibited high predictive capability, reaching an AUC of 0.911. There was a slight improvement in model performance, but this difference was also not statistically significant, as evidenced by the AUC values of the DECT and radiomics models, which were 0.726 and 0.884, respectively.

The study highlighted the potential of combining quantitative DECT parameters and radiomics in a fusion model to improve the preoperative prediction of muscle-invasive BCa. This approach may enhance treatment planning and patient outcomes.

No funding was provided.

Source: https://pubmed.ncbi.nlm.nih.gov/39285354/

Hu M, Wei W, Zhang J, et al. (2024). “Assessing muscle invasion in bladder cancer via virtual biopsy: a study on quantitative parameters and classical radiomics features from dual-energy CT imaging.” BMC Med Imaging. 2024;24(1):245. Published 2024 Sep 16. doi:10.1186/s12880-024-01427-w