Differentiation of common histological lesions in bovine liver using convolutional neural network-based deep learning
Susumu Iwaide, Kumiko Kimura, Ryo Sugiura, Yu Oishi, Tomoyuki Shibahara · 2026
Read Differentiation of common histological lesions in bovine liver using convolutional neural network-based deep learning by Susumu Iwaide, Kumiko Kimura, Ryo Sugiura, Yu Oishi, Tomoyuki Shibahara (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/pmc-13314324.txt.
Description
Susumu Iwaide and co-authors train, in a 2026 study, a learning model to sort cattle liver lesions from slide images. Pathologists face lymphoma, necrosis, and fibrosis in bovine livers, and an automatic classifier could speed their work. The team assembled ten cases for each of the four classes, normal liver included. They scanned the slides and cut them into fifty patches per slide, then trained a DenseNet-based convolutional network with cross-validation and judged it on a test set.
The model classified test patches with 84.5 percent accuracy and an F1 score of 83.5 percent across labels. It handled necrosis and fibrosis confidently, with recall above ninety percent for both. It caught half the lymphoma patches, despite precision when it did commit. Normal liver showed the opposite problem, catching everything while over-calling. The authors call the performance promising for a support tool, with lymphoma detection the target for improvement.
Catalog record
| CREATOR | Susumu Iwaide, Kumiko Kimura, Ryo Sugiura, Yu Oishi, Tomoyuki Shibahara |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 25 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |