Development and evaluation of AI model with deep learning for segmentation of extraocular muscles in thyroid eye disease
Yusuke Haruna, Mizuki Tagami, Ryo Kurosaki, Mizuho Nishio, Gen Kinari, Mami Tomita, Norihiko Misawa, Yoshihiro Muragaki · 2026
Read Development and evaluation of AI model with deep learning for segmentation of extraocular muscles in thyroid eye disease by Yusuke Haruna, Mizuki Tagami, Ryo Kurosaki, Mizuho Nishio, Gen Kinari, Mami Tomita, Norihiko Misawa, Yoshihiro Muragaki (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/plos-pone-0349074.txt.
Description
Yusuke Haruna and colleagues built an artificial intelligence model that outlines the eye muscles on MRI scans of people with thyroid eye disease. The single-center study, from a hospital in Osaka, used scans from fifty-two patients whose muscles had been traced by hand. The model learned from those tracings and was tested against them.
Accuracy improved as the training set grew from twelve to fifty-two cases, and the model's brightness measures inside the muscles matched hand measures closely. Swollen eye muscles drive much of the harm in this disease, so fast and consistent outlining could help doctors stage it and follow treatment. The gains with more training cases show data size matters.
Catalog record
| CREATOR | Yusuke Haruna, Mizuki Tagami, Ryo Kurosaki, Mizuho Nishio, Gen Kinari, Mami Tomita, Norihiko Misawa, Yoshihiro Muragaki |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 19884 |
| LICENSE | CC_BY |
| RIGHTS | CC BY |
| SOURCE | PLOS |