The Problem
The project addresses communication breakdown and workflow friction in radiology access, which leads to delayed patient care, clinician burnout, and avoidable costs across Duke Health. Current access methods (via DukeRads, Epic Messaging, and phone calls) are fragmented and often fail to deliver timely, context-specific guidance.
Our Solution
This innovation is a context-aware AI assistant that serves as an intelligent front desk for Duke Radiology. It uses a Large Language Model to interpret natural language questions and instantly provides the correct contact information and integrated clinical guidance based on dynamic factors like the time of day and the specific clinical scenario.
Anticipated Impact
We hypothesize that our solution will substantially reduce the volume of misdirected, non-emergent, and basic informational phone calls by providing instant, accurate self-service triage. While it does not solve every complex, patient-specific consult, it provides a complete solution for triage and appropriateness guidance.
