Colorectal Lesion diagnosis using transformer and deep learning with multiscale feature interface
Dhirendra Prasad Yadav, Bhisham Sharma, Julian L Webber, Abolfazl Mehbodniya · 2026
Read Colorectal Lesion diagnosis using transformer and deep learning with multiscale feature interface by Dhirendra Prasad Yadav, Bhisham Sharma, Julian L Webber, Abolfazl Mehbodniya (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/plos-pone-0357664.txt.
Description
Dhirendra Prasad Yadav and fellow engineers present a computer vision system that sorts growths in the colon from medical images. Finding these lesions early matters because bowel cancer ranks among the most common cancers worldwide, yet expert review takes time and skill. Their network mixes two designs: one that reads fine local detail and one that weighs the whole image at once, with a bridge part that lets the two share what they learn.
Tested on two public image sets, the combined system scored high on precision, recall, and agreement measures, missing few true cases. Yadav and colleagues argue that joining local and global views helps a model notice both small surface changes and wider shape cues. Hospitals short on specialists could use such tools as a first reader.
Catalog record
| CREATOR | Dhirendra Prasad Yadav, Bhisham Sharma, Julian L Webber, Abolfazl Mehbodniya |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 41929 |
| LICENSE | CC_BY |
| RIGHTS | CC BY |
| SOURCE | PLOS |