Preoperative diagnosis of benign and malignant thyroid nodules is crucial for appropriate clinical treatment and individual patient management. In this study, a double-layer spectral detector computed tomography (DLCT)-based nomogram for the preoperative classification of benign and malignant thyroid nodules was developed and tested.
A total of 405 patients with pathological findings of thyroid nodules who underwent DLCT preoperatively were retrospectively recruited. They were randomized into a training cohort (n=283) and a test cohort (n=122). Information on clinical features, qualitative imaging features and quantitative DLCT parameters was collected. Univariate and multifactorial logistic regression analyses were used to screen independent predictors of benign and malignant nodules. A nomogram model based on the independent predictors was developed to make individualized predictions of benign and malignant thyroid nodules. Model performance was evaluated by calculating the area under the receiver operating characteristic curve (AUC), calibration curve and decision curve analysis(DCA).
Standardized iodine concentration in the arterial phase, the slope of the spectral hounsfield unit(HU) curves in the arterial phase, and cystic degeneration were identified as independent predictors of benign and malignant thyroid nodules. After combining these three metrics, the proposed nomogram was diagnostically effective, with AUC values of 0.880 for the training cohort and 0.884 for the test cohort. The nomogram showed a better fit (all p > 0.05 by Hosmer−Lemeshow test) and provided a greater net benefit than the simple standard strategy within a large range of threshold probabilities in both cohorts.
The DLCT-based nomogram has great potential for the preoperative prediction of benign and malignant thyroid nodules. This nomogram can be used as a simple, noninvasive, and effective tool for the individualized risk assessment of benign and malignant thyroid nodules, helping clinicians make appropriate treatment decisions.