AI-assisted Early Detection of Pancreatic Ductal Adenocarcinoma on Contrast-enhanced CT
Han Liu, Riqiang Gao, Sasa Grbic · 2025
Read AI-assisted Early Detection of Pancreatic Ductal Adenocarcinoma on Contrast-enhanced CT by Han Liu, Riqiang Gao, Sasa Grbic (2025) 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/arxiv-2503-10068.txt.
Description
Han Liu, Riqiang Gao, Eileen Krieg, and Sasa Grbic present PanDx, an AI aid for spotting pancreatic cancer early on CT scans. Their paper from twenty twenty-five targets pancreatic ductal adenocarcinoma, a swift and deadly cancer. Early signs on scans are faint, so most cases surface late. Their frame works coarse to fine on contrast-enhanced scans.
Two tricks drive it. A distribution-aware ensemble pools models by lesion type to travel across cases. A peak-scaled step then pulls candidate lesions with sharper placement. In the PANORAMA challenge the system took first place on the held-out test set. Scores on ranking gauges were strong. A radiologist helped sort the failure cases afterward. The authors map where such aids still miss and what to build next. Code and models are public.
Catalog record
| CREATOR | Han Liu, Riqiang Gao, Sasa Grbic |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 16 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-10068