CLEAR-IT, a framework for contrastive learning to capture the immune composition of tumor microenvironments
Daniel Spengler, Serafim Korovin, Kirti Prakash, Peter Bankhead, Reno Debets, Hayri E. Balcioglu, Carlas Smith · 2026
Read CLEAR-IT, a framework for contrastive learning to capture the immune composition of tumor microenvironments by Daniel Spengler, Serafim Korovin, Kirti Prakash, Peter Bankhead, Reno Debets, Hayri E. Balcioglu, Carlas Smith (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-13260730.txt.
Description
Daniel Spengler and co-authors present CLEAR-IT, a contrastive learning tool for tumor immune mapping, reading cell types from layered tissue images. Sorting cells in the tumor microenvironment usually demands precise cell borders, and that outlining step limits how fast analysis scales. CLEAR-IT learns cell features using only cell spots.
The self-supervised encoders scored strongly in linear tests. Tuning improved them further, and accuracy held across imaging types. Ninety percent fewer labels still yielded strong results, and swapping them into a top sorting model raised scores. Half the labeled data then matched full-data results. One patient's labeled cells predicted survival patterns elsewhere. Labs could sort tumors with far less manual work.
Catalog record
| CREATOR | Daniel Spengler, Serafim Korovin, Kirti Prakash, Peter Bankhead, Reno Debets, Hayri E. Balcioglu, Carlas Smith |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 82 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |