Ultrasound-based deep learning radiomics with explainable machine learning for predicting benign and malignant thyroid nodules in lymphocytic thyroiditis: a multicenter study
Ji Zhou, Wenwu Lu, Wei Wei, Ziyi Liu, Wang Zhou, Wei Peng, Wenbo Ding, Xin Wu, Chaoxue Zhang · 2026
Read Ultrasound-based deep learning radiomics with explainable machine learning for predicting benign and malignant thyroid nodules in lymphocytic thyroiditis: a multicenter study by Ji Zhou, Wenwu Lu, Wei Wei, Ziyi Liu, Wang Zhou, Wei Peng, Wenbo Ding, Xin Wu, Chaoxue Zhang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13654164.txt.
Description
Ji Zhou and co-authors build a prediction model for a hard ultrasound call: whether a thyroid lump is cancerous when the gland also carries chronic inflammation. The inflammation muddies the scan picture, and few models were tested in exactly this group. The multicenter study enrolled 640 patients whose nodules were confirmed by pathology. Deep learning and hand-crafted image features were fused, trimmed, and fed to ten candidate algorithms, then joined with clinical factors into a single chart.
Logistic regression won the contest. Scores for telling benign from malignant reached 0.90 in training and in an outside test set, and the combined chart lifted them to 0.93, 0.94, and 0.92 across the three sets. An explanation method showed which features drove each call. The authors offer the chart for pre-surgery sorting of these tricky nodules.
Catalog record
| CREATOR | Ji Zhou, Wenwu Lu, Wei Wei, Ziyi Liu, Wang Zhou, Wei Peng, Wenbo Ding, Xin Wu, Chaoxue Zhang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 35 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |