V7-DiVA: a feature-based Deep MIL colorectal whole-slide histopathology research platform with blinded external image-input validation
Mingchen Sun, Wencong Kong, Shuanglong Qiu, Xinyu Ge · 2026
Read V7-DiVA: a feature-based Deep MIL colorectal whole-slide histopathology research platform with blinded external image-input validation by Mingchen Sun, Wencong Kong, Shuanglong Qiu, Xinyu Ge (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-13491756.txt.
Description
Mingchen Sun, Wencong Kong, Shuanglong Qiu, and Xinyu Ge present a deep learning platform for colorectal whole-slide pathology, in a study from 2026. Slide-level models trained on public data often fail to transfer to real clinics. Their system links classification with quality data and reviewable image regions for trust.
Patch features combine a supervised core with tumor evidence, shape scores, coordinates, and focus quality. The main attention model trained on nearly two thousand public slides from hundreds of cases. It was then frozen and tested blind on three hundred local slides, half malignant. Patch tumor scores were near perfect on a public validation set.
Catalog record
| CREATOR | Mingchen Sun, Wencong Kong, Shuanglong Qiu, Xinyu Ge |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 59 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |