Deep Learning for Wound Tissue Segmentation: A Comprehensive Evaluation using A Novel Dataset Journal: ’
Muhammad Ashad Kabir, Nidita Roy, Md. Ekramul Hossain, Jill Featherston, Sayed Ahmed · 2025
Read Deep Learning for Wound Tissue Segmentation: A Comprehensive Evaluation using A Novel Dataset Journal: ’ by Muhammad Ashad Kabir, Nidita Roy, Md. Ekramul Hossain, Jill Featherston, Sayed Ahmed (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/arxiv-2502-10652.txt.
Description
Muhammad Ashad Kabir and co-authors release a labeled wound image dataset and grade the leading segmentation models on it. Teaching software to outline wound tissue types has stalled for lack of public, carefully labeled data. Their 2025 dataset holds one hundred forty-seven wound images marked for six tissue kinds, from healthy granulation to dead necrosis, plus exposed bone and tendon.
Labels come in three formats: whole image, patches, and superpixels. The team then runs a sweep of architectures — the classic U-Net, adversarial networks, and hybrid designs — alongside classification approaches. Results sort the field on identical data for the first time. Medical imaging researchers get both the missing dataset and a fair scoreboard for it.
Catalog record
| CREATOR | Muhammad Ashad Kabir, Nidita Roy, Md. Ekramul Hossain, Jill Featherston, Sayed Ahmed |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 83 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-10652