Using language models to predict surgical risk using pre-operative clinical notes

Title:
Using language models to predict surgical risk using pre-operative clinical notes
Description:
In this talk, Dr. Chung will explore the use of language models to perform risk stratification using pre-operative clinical notes. Perioperative risk prediction of patients is a challenge surgeons, proceduralists, and anesthesiologists face every day in clinical practice. If a patient is scheduled for a procedure, clinicians must evaluate the medical fitness of the patient and whether they can safely withstand the stress of the procedure. Traditional clinical risk calculators typically only assess a single facet of the patient’s medical comorbidities and typically use structured clinical data such as medical history, laboratory results, etc. to derive a risk prediction.
Speaker:
Philip Chung, MD, PhD.
Postdoctoral Fellow and Clinical Scholar
Stanford University
Date:
April 16, 2024