Genetically Aligned Patient Representations Improve Hematological Diagnosis
Muhammed Furkan Dasdelen, Fatih Ozlugedik *, Ilaria Looser, Rao Muhammad Umer, Christian Pohlkamp, Carsten Marr · 2026
Read Genetically Aligned Patient Representations Improve Hematological Diagnosis by Muhammed Furkan Dasdelen, Fatih Ozlugedik *, Ilaria Looser, Rao Muhammad Umer, Christian Pohlkamp, Carsten Marr (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/arxiv-2605-29980.txt.
Description
Muhammed Furkan Dasdelen and five colleagues line up single white blood cell images with karyotypes and gene-panel mutations for blood-cancer diagnosis. Training runs in two stages: vision-only pretraining of a transformer aggregator on over fifteen hundred patients, then genetic lining up on acute myeloid leukemia patients. The lined patient encoder beats slide-level histopathology models on blood diagnosis tasks. It also offers retrieval for diseases and genetic changes. Folding genetic data into patient encoders lifts quality and matches clinical workflow, opening the way to blood-specific AI.
Catalog record
| CREATOR | Muhammed Furkan Dasdelen, Fatih Ozlugedik *, Ilaria Looser, Rao Muhammad Umer, Christian Pohlkamp, Carsten Marr |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 21606 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.29980