Related Experiment Video
Updated: May 1, 2026
A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
Published on: April 21, 2023
Construction of a nomogram predictive model for thermal ablation efficacy in benign thyroid nodules based on
Mengyao Zhao1,2, Zeyang Dong1,2, Jie Chen2
1Department of Ultrasound, Zhejiang Hospital, Hangzhou, Zhejiang Province, China.
Background:
This study aimed to develop a nomogram for predicting the 1-year volume reduction ratio (VRR) of the ablated area after thermal ablation of benign thyroid nodules (BTN) based on preoperative clinical, ultrasound, and serological features, and to explore the role of the ablation volume ratio (AVR).
Methods:
This retrospective study included 243 BTN patients who underwent ultrasound-guided radiofrequency or microwave ablation between January 2020 and December 2024. Clinical, ultrasound, serological, and procedural data were collected. Contrast-enhanced ultrasound was performed immediately after ablation to assess the ablation zone and calculate AVR. Patients were followed at 1, 3, 6, and 12 months and classified as favorable (VRR ≥80%) or unfavorable. The cohort was randomly divided into training (70%) and validation (30%) sets. Logistic regression identified independent predictors, and a nomogram was constructed and evaluated using ROC, calibration, and decision curve analyses.
Results:
Among 243 patients, 142 (58.4%) achieved VRR ≥80%. Significant differences were observed in BMI, nodule size, vascularity, morphology, FT3 level, and location within the dangerous triangle (all p < 0.05). AVR was higher in the favorable group but was not an independent predictor. Multivariate analysis identified BMI, maximum diameter, FT3, morphology, vascularity, and dangerous triangle location as independent predictors. The model showed good performance (AUC 0.904 training, 0.815 validation).
Conclusion:
A nomogram based on preoperative factors provides a noninvasive tool for predicting 1-year outcomes after BTN ablation. AVR reflects procedural adequacy but has limited prognostic value. Further multicenter validation is needed.
