Deep learning-based ordinal classification overcomes subjective assessment limitations in intraoral free flap monitoring
Hyounmin Kim, Dongwook Kim, Juho Bai · 2025
Read Deep learning-based ordinal classification overcomes subjective assessment limitations in intraoral free flap monitoring by Hyounmin Kim, Dongwook Kim, Juho Bai (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/pmc-12847850.txt.
Description
Hyounmin Kim, Dongwook Kim, and Juho Bai build an AI model to grade the health of tissue flaps inside the mouth, in a 2025 study. Free flaps rebuild the mouth after cancer or injury surgery, but checking them relies on subjective judgement and has no standard. The model sorts flap photos into three ordered grades: normal, suspicious, and compromised.
Close to nineteen hundred photos from more than one hundred patients were used. The model was trained with a new loss method that respects the order of the grades. On held-out photos, it reached about ninety-six percent accuracy, well above a baseline model. It was strongest at spotting normal flaps and weaker, though still useful, on compromised ones. Scores that separate each grade from the others were all high. The authors present the tool as a way to make flap checks more standard and timely.
Catalog record
| CREATOR | Hyounmin Kim, Dongwook Kim, Juho Bai |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 38 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |