OncoSynth: Synthetic data generation for treatment effect estimation in oncology
Octavia-Andreea Ciora, Julian Welzel, Dennis Frauen, Maresa Schröder, Marie Brockschmidt, Harry Amad, Thomas Callender, Mihaela van der Schaar, Stefan Feuerriegel · 2026
Read OncoSynth: Synthetic data generation for treatment effect estimation in oncology by Octavia-Andreea Ciora, Julian Welzel, Dennis Frauen, Maresa Schröder, Marie Brockschmidt, Harry Amad, Thomas Callender, Mihaela van der Schaar, Stefan Feuerriegel (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-2606-25762.txt.
Description
Cancer researchers cannot freely share patient records. Synthetic stand ins are tempting. But ordinary generators scramble the causal links between traits, treatment choice, and survival. Any effect estimate then comes out biased. Octavia-Andreea Ciora and colleagues built OncoSynth in 2026 to keep those links intact. Its diffusion generator works sequentially. First it models how traits steer treatment assignment. Then it models how treatment shapes survival. The result is a fake cohort that behaves causally like the real one.
Validation used large lung and breast cancer cohorts. They held roughly thirty seven thousand and seventeen thousand patients. The synthetic cohorts matched real distributions of traits, treatments, and outcomes closely. Treatment effect estimates improved sharply over existing generators. Population level error fell by up to two thirds. Patient level error fell by well over half. That supports evidence generation where data sharing is blocked.
Catalog record
| CREATOR | Octavia-Andreea Ciora, Julian Welzel, Dennis Frauen, Maresa Schröder, Marie Brockschmidt, Harry Amad, Thomas Callender, Mihaela van der Schaar, Stefan Feuerriegel |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 92 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.25762