DHEPLab collaborators have published a groundbreaking framework for bias evaluation in large language models in healthcare settings in npj Digital Medicine.
Co-authored by Tara Templin, Sean Sylvia, Junier Oliva, and team members, the paper provides digital health practitioners with concrete methods to identify and mitigate algorithmic bias in clinical AI systems.
As large language models are increasingly deployed in healthcare applications—from clinical documentation to decision support—ensuring these systems perform equitably across patient populations is critical. This framework offers practical guidance for researchers and practitioners seeking to evaluate and improve the fairness of AI systems in clinical settings.
The publication reflects DHEPLab’s commitment to advancing rigorous evaluation methods for clinical AI, ensuring that technological advances benefit all patients.