Adaptive collaborative feature fusion and shape-aware optimization for multi-scale chest lesion detection
Yubo Yuan, Chengkang Liu, Qingwen Feng, Li Zhang, Kun Yang, Shengyue Cheng · 2026
Read Adaptive collaborative feature fusion and shape-aware optimization for multi-scale chest lesion detection by Yubo Yuan, Chengkang Liu, Qingwen Feng, Li Zhang, Kun Yang, Shengyue Cheng (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/PMC13518927.txt.
Description
Yubo Yuan and five colleagues present a 2026 deep learning method for spotting chest lesions on X-rays across many sizes and shapes. Lesions vary wildly in scale and outline, and small ones are easily missed inside complex anatomy. Their system first enriches features with both whole-chest context and local lesion detail.
An adaptive focusing stage then pools features at several scales to catch small lesions, while a shape-aware training rule using a normalized distance measure steadies box placement for irregular shapes. On the VinDr-CXR dataset the method gained 6.3 percent in average precision and 8.6 percent in recall over leading rivals. Gains on ChestX-ray8 confirmed the approach travels across datasets.
Catalog record
| CREATOR | Yubo Yuan, Chengkang Liu, Qingwen Feng, Li Zhang, Kun Yang, Shengyue Cheng |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 141856 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |