Related Experiment Video
Updated: Apr 4, 2026
A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
Published on: April 21, 2023
Integrating Ultrasound and Clinicopathologic Characteristics to Predict the Invasive Papillary Thyroid Carcinoma
Minfang Yao1,2, Zheng Zhang1, Zheng Hua1
1Department of Medical Ultrasound, Affiliated Hospital of Jiangsu University, Zhenjiang, 212000, People's Republic of China.
Purpose:
The management of indeterminate thyroid nodules (ITNs; Bethesda III-V) poses significant clinical challenges. This study sought to identify preoperative ultrasound and clinicopathologic predictors of invasive papillary thyroid carcinoma (PTC) among ITNs and to develop a corresponding risk prediction model.
Patients And Methods:
In this retrospective study, 494 patients with FNA-confirmed ITNs and postoperative PTC diagnosis were included. Based on pathology confirming extrathyroidal extension and/or lymph node metastasis, patients were classified as invasive PTC (n=141) or non-invasive PTC (n=353). Univariate and multivariate logistic regression analyses identified independent risk factors, and a predictive nomogram was developed and validated.
Results:
Male (odds ratio [OR]=2.91, 95% confidence interval [CI]:1.78-4.76), age ≤45 years (OR=1.93, 95% CI:1.23-3.03), abundant nodule vascularity (OR=4.60, 95% CI:2.42-8.75), and capsule proximity ≤2 mm (OR=3.63, 95% CI:2.15-6.14) were independent risk factors for invasive PTC, while abnormal thyroglobulin antibody (TgAb) levels reduced risk (OR=0.38, 95% CI:0.18-0.77). The prediction model achieved an AUC of 0.776 (95% CI:0.728-0.825) in the training set and 0.759 (95% CI:0.643-0.874) in validation, with decision curve analysis confirming clinical utility.
Conclusion:
An integrated model incorporating sex, age, vascularity, capsule distance, and TgAb status effectively predicts invasive PTC risk in ITNs. The nomogram provides preoperative risk stratification to guide personalized treatment, potentially reducing unnecessary aggressive surgery in low-risk cases while ensuring optimal management of high-risk patients. An interactive online version is available for clinical implementation.
