Richard Dirven, Otolaryngologist, Head and Neck Surgeon, Facial Plastic Reconstructive Surgery at Radboudumc shared a post on LinkedIn:
“I in the clinic needs needs responsible validation and implementation.
In our recent paper in NPJ Precision Oncology, we propose a framework for responsible use of Large Language Models (LLMs) in clinical decision support for precision oncology.
Interested in reading? It’s open access.
In summary:
While LLMs can structure unstructured patient data, accelerate literature review, and support treatment decisions, very few studies actually use real-world patient data to validate these systems.
We therefore outline 5 key dimensions for responsible AI in clinical practice:
- Adherence to medical standards,
- balancing benefits and risks,
- regulatory compliance,
- ethical considerations,
- measurable clinical and scientific value.
These are translated into 10 practical principles for implementing LLM-driven clinical decision support systems. The real challenge is no longer building AI – it is integrating it responsibly into the clinical workflow.
This is a key step toward the AI-supported consultation room of the future.”
Title: Collaborative framework on responsible AI in LLM-driven CDSS for precision oncology leveraging real-world patient data
Authors: Sonja Mathes, Dyke Ferber, Tobias Dreyer, Kai J. Borm, Luise Modersohn, Theresa Willem, Richard Dirven, Julien Vibert, Simon Kreutzfeldt, Raquel Perez-Lopez, Arsela Prelaj, Fredrik Strand, Richard D. Baird, Martin Boeker, Jakob Nikolas Kather, Maximilian Tschochohei, Jacqueline Lammert
Read The Full Article
Other articles about AI on OncoDaily.