WindowNet: Learnable Windows for Chest X-ray Classification
Alessandro Wollek, Sardi Hyska, Bastian Sabel, Michael Ingrisch, Tobias Lasser · 2023
Read WindowNet: Learnable Windows for Chest X-ray Classification by Alessandro Wollek, Sardi Hyska, Bastian Sabel, Michael Ingrisch, Tobias Lasser (2023) 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-10743662.txt.
Description
Alessandro Wollek and co-authors teach chest X-ray models to adjust image windows like radiologists do in a 2023 paper. Shared X-ray sets shrink image depth to save space, which can hide faint signs. Human readers stretch contrast windows to reveal those signs. The team tests whether windowing lifts machine scores on chest films.
Windowing raised classification scores by a clear margin. Their WindowNet model learns several window settings on its own. It averaged well above a standard rival build on a large hospital dataset. Learned windows mimic a reader habit that costs little compute. The trick may lift other medical image tasks where faint contrast carries the diagnosis.
Catalog record
| CREATOR | Alessandro Wollek, Sardi Hyska, Bastian Sabel, Michael Ingrisch, Tobias Lasser |
|---|---|
| YEAR | 2023 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 16 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |