Artificial intelligence-based multimodal multitask analysis of thyroid ultrasound image features predicts thyroid cancer: a multicenter study
JNCI Cancer Spectrum · 07.07.2026 · 10 (4)
DOI: 10.1093/jncics/pkag037 ↗Kurz gefasst
Multizentrische Studie in der Zeitschrift JNCI Cancer Spectrum, online erschienen am 09.04.2026. Verfasst von Yu Gui und 23 weiteren Autor:innen. Beteiligte Einrichtung: Department of Breast and Thyroid Surgery, Southwest Hospital, Army Medical University, Chongqing, China. Hauptthemen laut MEDLINE: Thyroid Neoplasms, Thyroid Nodule, Artificial Intelligence. Die Arbeit nennt 35 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Background
Thyroid nodule ultrasound (US) images and their features are of great importance in thyroid nodule diagnosis and can be helpful for radiologists’ clinical decision making. The aim was to evaluate whether an artificial intelligence (AI)-assisted system can accurately characterize thyroid nodule US features and assist radiologists in diagnosing thyroid cancer.
Methods
The AI-assisted system (MDT-TC) was trained and internally validated on B-mode US images from 7204 lesions in 6884 patients in Southwest Hospital. The model performance was validated using 3 independent external validation cohorts.
Results
Echogenicity (ECH) and shape (SHA) are features of high importance for model recognition, and these features lead to excellent model performance. The model achieved up to 87.56% accuracy in determining ECH attributes and 69.21% in identifying shape categories. The area under the receiver operating characteristic curve (AUC) of the internal validation cohort and 3 independent external validation cohorts for MDT-TC were 0.951, 0.837, 0.816, and 0.871, respectively. The sensitivity values were 98.7%, 91.2%, 90.3%, and 85.6%, respectively. The AUC for the accurate diagnosis of radiologists with MDT-TC assistance was significantly higher than that of radiologists without MDT-TC assistance (P < .001). In addition, the AUC for the accurate diagnosis of junior doctors with MDT-TC assistance was significantly higher than that for those without (P < .01).
Conclusion
MDT-TC incorporates radiomic features extracted from thyroid lesion US images and can significantly improve the diagnostic performance of radiologists. This result was particularly strong for junior doctors. Therefore, our data support the idea that MDT-TC can help to identify patients with thyroid cancer and could greatly benefit clinical practice.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Department of Breast and Thyroid Surgery, Southwest Hospital, Army Medical University, Chongqing, China.
- College of Mathematics and Statistics, Chongqing University, Chongqing, China.
- Department of Ultrasound, Southwest Hospital, Army Medical University, Chongqing, China.
- Department of General Surgery, The People's Hospital of Dazu, Chongqing, China.
- College of Mechanical and Transportation Engineering, Chongqing University, Chongqing, China.
- Department of Oncology Surgery, Shaanxi Provincial People's Hospital, Xian, China.
ORCID-Profile: Yu Gui, Longle Liu, Feng Liu, Jianhui Li
Zitieren
APA
Gui, Y., Zhang, X., He, Y., Li, T., Hu, L., Liu, L., Peng, D., Yuan, J., Xiong, X., Li, W., Wu, X., Li, S., Fan, H., Peng, T., Yang, X., Cui, X., Yang, Y., Zeng, L., Song, D., …, & Chen, L. (2026). Artificial intelligence-based multimodal multitask analysis of thyroid ultrasound image features predicts thyroid cancer: a multicenter study. JNCI Cancer Spectrum, 10(4). https://doi.org/10.1093/jncics/pkag037
Vancouver
Gui Y, Zhang X, He Y, Li T, Hu L, Liu L, et al. Artificial intelligence-based multimodal multitask analysis of thyroid ultrasound image features predicts thyroid cancer: a multicenter study. JNCI Cancer Spectrum. 2026;10(4). doi:10.1093/jncics/pkag037
BibTeX
@article{Gui2026,
author = {Gui, Yu and Zhang, Xuerui and He, Yun and Li, Tingting and Hu, Lan and Liu, Longle and Peng, Dan and Yuan, Juan and Xiong, Xiaoyu and Li, Wenhan and Wu, Xin and Li, Shichao and Fan, Haiyan and Peng, Tijie and Yang, Xi and Cui, Xiang and Yang, Ying and Zeng, Lingjuan and Song, Dajiang and Liu, Feng and Li, Jianhui and Wang, Peng and Huang, Zhongquan and Chen, Li},
title = {Artificial intelligence-based multimodal multitask analysis of thyroid ultrasound image features predicts thyroid cancer: a multicenter study},
journal = {JNCI Cancer Spectrum},
year = {2026},
volume = {10},
number = {4},
doi = {10.1093/jncics/pkag037}
}
EUVIMED