Poster 247. Bot vs. Doc – Who is Better at Reading Proximal Humerus Fracture X-Rays?
Kyle K Obana, Mark Ren, Andrew J Luzzi, Matthew R LeVasseur, Hasani W Swindell, William N Levine · 2026
Read Poster 247. Bot vs. Doc – Who is Better at Reading Proximal Humerus Fracture X-Rays? by Kyle K Obana, Mark Ren, Andrew J Luzzi, Matthew R LeVasseur, Hasani W Swindell, William N Levine (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/PMC13530548.txt.
Description
Obana and colleagues present in 2026, as conference poster 247, a head-to-head test of ChatGPT-5 against orthopedic surgeons at reading shoulder X-rays. The images were checked for proximal humerus fractures, breaks near the top of the arm bone. The playful title, Bot versus Doc, frames a serious question about trusting generative AI in clinical reading.
Seventy images, half fractured and half normal from a public Stanford dataset, were read independently by the chatbot and by surgeons from junior resident to fellowship-trained attending. The chatbot reached only about two-thirds accuracy overall, missing nearly a quarter of the isolated fractures by calling them normal and inventing fractures in a quarter of the normal films. Its agreement with reference reads was slight to poor on fracture details, while human trainees agreed moderately to almost perfectly. The authors warn against leaning on the tool.
Catalog record
| CREATOR | Kyle K Obana, Mark Ren, Andrew J Luzzi, Matthew R LeVasseur, Hasani W Swindell, William N Levine |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 2 KB |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |