Abstract
BACKGROUND: Precise identification of papillary thyroid carcinoma (PTC) is crucial in clinical practice to prevent unnecessary treatment. This research aimed to combine metabolic function parameters and inflammatory markers to establish a multimodal diagnostic model using the Chinese Thyroid Imaging Reporting and Data System (C-TIRADS) for enhancing risk evaluation and improving clinical decision-making for PTC. METHODS: A total of 314 patients with thyroid nodules were retrospectively enrolled, consisting of 193 cases of PTC and 121 cases of benign thyroid nodules (BTNs). These participants were randomly divided into a training set with 222 cases and a validation set with 92 cases at a ratio of 7:3. Univariate analysis and multivariate logistic regression were utilized to identify independent predictors for PTC diagnosis, leading to the creation of prediction models, including both baseline models and an integrated model. The discriminative ability of the integrated model was assessed using the area under the curve (AUC) of the receiver operating characteristic (ROC) curve. Subsequently, the diagnostic performance of the integrated model was compared with that of the baseline model. The calibration of the integrated model was evaluated using calibration curves in combination with the Hosmer-Lemeshow test. RESULTS: Multivariate logistic regression identified C-TIRADS high-risk classification, body mass index (BMI), thyroid stimulating hormone (TSH), lymphocyte-to-monocyte ratio (LMR), age, and thyroglobulin (Tg) as independent predictors of PTC diagnosis. The integrated model exhibited significantly higher diagnostic efficiency compared to the baseline model in the training set (AUC: 0.903 vs. 0.827, 0.878, 0.874, P<0.05). However, there was no statistically significant difference between the model and the baseline model in the validation set (AUC: 0.845 vs. 0.816, 0.837, 0.829, P>0.05). The calibration curve demonstrated a high level of consistency between the predicted probability of the integrated model and the actual risk probability (Hosmer-Lemeshow test P>0.05). CONCLUSIONS: The integrated model based on C-TIRADS classification combined with metabolic function parameters and inflammatory indicators has good efficacy in risk assessment of PTC, and can provide an objective quantitative tool for individual diagnostic evaluation and treatment decisions.