Self-supervised 3D deep learning on preoperative contrast-enhanced computed tomography for predicting high pathologic nodal burden in esophageal squamous cell carcinoma: temporal and external multicohort validation
Qian Li, Yongxin Li, Jing Bai, Jinze Zhang, Zhenkai Nie, Yanxin Ren, Zhantao Li, Jin Guo, Meng Li, Tingting Zhang, Xinbo Liu · 2026
Read Self-supervised 3D deep learning on preoperative contrast-enhanced computed tomography for predicting high pathologic nodal burden in esophageal squamous cell carcinoma: temporal and external multicohort validation by Qian Li, Yongxin Li, Jing Bai, Jinze Zhang, Zhenkai Nie, Yanxin Ren, Zhantao Li, Jin Guo, Meng Li, Tingting Zhang, Xinbo Liu (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/PMC13293925.txt.
Description
Qian Li and a large team build a deep learning tool in 2026 to predict heavy lymph node spread in esophageal squamous cell cancer before surgery. Standard CT rules based on node size miss many cases. Their model first learned from 3,200 unlabeled chest scans, then trained on tumor volumes from 1,060 patients across two centers with later and outside test groups.
Accuracy held up well across checks, with area values of 0.881 inside, 0.860 over time, and 0.860 externally. In the outside group, sensitivity beat a guideline-style comparator while specificity also rose. Better calibration adds trust. Such prediction could sharpen staging and treatment choices before an operation.
Catalog record
| CREATOR | Qian Li, Yongxin Li, Jing Bai, Jinze Zhang, Zhenkai Nie, Yanxin Ren, Zhantao Li, Jin Guo, Meng Li, Tingting Zhang, Xinbo Liu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 36684 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |