Roxana Daneshjou, MD, PhD, FAAD
Presented on Wednesday, March 20, 2024 – 4:00 PM MST
Watch this webinar: http://bit.ly/4adPsUX
Dr. Roxana Daneshjou, MD, PhD, FAAD, Assistant Professor of Biomedical Data Science and Dermatology at Stanford School of Medicine, explores the dynamic landscape of large language models (LLMs) within the healthcare domain. This presentation provides an examination of the processes involved in constructing LLMs and the limitations of these models. Furthermore, it illuminates shifts in incorporating LLMs in clinical practice and highlights inherent biases embedded within LLMs that raise deep concerns about their usage in the healthcare setting. She will argue that the lack of standardized frameworks for rigorous testing and validation emerges as a pivotal consideration. As Dr. Daneshjou navigates the frontiers of biomedical data science, she advocates for the development of robust bias mitigation strategies and comprehensive measures to ensure the responsible and equitable deployment of LLMs in clinical contexts.
