Roupen Odabashian, Hematology/Oncology Fellow at the Karmanos Cancer Institute, shared a post on LinkedIn:
“I’m an oncologist. In the last few months I shipped two tools and open-sourced both.
One is for patients: you describe your cancer, your city, and the language you speak, and five ‘helpers’ (a movement specialist, a dietician, a swallowing specialist, an emotional wellbeing specialist, and a patient navigator) pull from public sources and hand you a plain-language summary, in your language, to bring to your oncology team. Not medical advice. A better set of questions for your next appointment.
The other is for clinicians: paste a de-identified case and it convenes an AI ‘tumor board’ and surfaces matching clinical trials.
I didn’t build these to sell them. I built them because the gap between what’s possible and what reaches an actual patient in an actual clinic is enormous, and most of it isn’t a technology problem anymore. It’s a ‘nobody bothered’ problem. So I open-sourced both. If they’re useful, use them. If they’re wrong, fix them.
Both are early (the tumor board is literally v0.5). That’s the point. I’d rather ship something honest and improvable than wait for perfect.
Both repos are on my GitHub. Try them, break them, or send a pull request.”
Other articles featuring Roupen Odabashian on OncoDaily.