Yan Leyfman, Medical Oncologist, Co-Founder and Executive Director of MedNews Week, shared on LinkedIn, by Arturo LoAIza-Bonilla, Systemwide Chief of Hematology and Oncology at St. Luke’s University Health Network, adding:
“Quick question for the hematology community:
Should AI be judged on straightforward guideline-based questions-or on the cases where even experts thoughtfully disagree?
We believe the latter.
That’s why we’re inviting hematologists and oncologists to contribute their clinical judgment to help establish a physician benchmark for evaluating AI in hematologic malignancies.
Your expertise is the reference standard.
Thanks to Dr. Bonilla for sharing this initiative with the community. If you’re interested in participating, the survey links are available in the shared post below.”
Quoting Arturo LoAIza-Bonilla:
“Most AI-in-medicine papers test the easy questions.
We’re testing the ones that keep hematologists up at night.
In AML, CLL and mantle cell lymphoma there are decisions where NCCN, ELN and iwCLL genuinely stop short, and two excellent clinicians can defend opposite choices.
We built a key set of them into a preregistered benchmark, and we’re running frontier AI models and practicing physicians through the same items.
Here’s the part that matters: the physicians aren’t the control arm. They’re the reference standard. If you treat these diseases, your judgment is the thing the models get measured against.
5-10 minutes. Anonymous. No patient data. Take one or take all three:
Closes in August.
Questions: Yan Leyfman,
Please tag a hematologist who should weigh in.”
Other articles about AI on OncoDaily.