BIDS Special Seminar

Perioperative healthcare data is vast and heterogenous and provides a rich resource for the development of clinically meaningful predictive models. Several data modalities exist - from structured electronic health record data, imaging, and clinical notes. As we move toward designing predictive models in healthcare that incorporate multimodal data, it is important to understand different approaches in harmonizing data while maintaining scalability.

The lecture will provide an overview of Gabriel's research in using AI to predict actionable outcomes for the intensive care unit, acute pain services, and preoperative optimization with anesthesia preoperative clinics. How we can turn these predictive models from academic work to EHR deployment will also be discussed.