The predictive value of the thyroid nodule benign and malignant based on the ultrasound nodule‐to‐muscle gray‐scale ratio

Abstract Objective To investigate the efficacy of the ultrasonic nodule to muscle gray scale ratio as a predictive tool for distinguishing between benign and malignant thyroid nodules. Methods A retrospective study was undertaken at the First People's Hospital of Hangzhou, affiliated with the Zhejiang University School of Medicine, analyzing ultrasound and pathological data of patients with thyroid nodules between May 2020 and December 2022. The study extracted ultrasound features of nodules and employed univariate and multivariate logistic regression analyses to identify independent risk factors for malignant tumors in the nodules. Subsequently, a predictive model for distinguishing benign and malignant thyroid nodules was developed. Results A total of 466 patients were included in this retrospective study, of which 275 cases were malignant tumors. Univariate and multivariate logistic regression analyses showed that the nodular‐muscle gray‐scale ratio, nodule diameter, margin status, aspect ratio, and calcification were closely related to thyroid malignant tumors. The area under the curve (AUC) of training group was 0.832, with a sensitivity, specificity, and accuracy of 85.5%, 67.4%, and 76.6%, respectively. The AUC of the external validation group was 0.819, with a sensitivity, specificity, and accuracy of 76.4%, 74.5%, and 75.7%, respectively. The calibration and decision curves showed that the model had good diagnostic value. Conclusion The research findings indicate that ratio is significantly associated with the malignant nature of thyroid nodules. The application of a line chart model based on these parameters exhibits a high level of predictive performance.