Great news! Our “ML in appe” proposal got approved, first shot! Initially conceived as a way for our lab to gain expertise in AI, this is now a grown-up, 5-year, $400k funded grant proposal aiming to move the diagnosis of perforated appendicitis from the operating room to the ED.

Appendicitis is like that – an odd, common, condition where a one-day delay can make the difference between a simple operative procedure with same-day discharge and quick recovery to a challenging intervention followed by a week-long hospitalization often complicated by septic complications and further interventions… If only we could now which case it is (simple or perforated) right from the ED, so we can council the family and prepare for the resources likely needed…

… which is exactly what our grant project will hopefully achieve, filling a significant gap in the literature by using machine learning to develop an AI model including all patient data available in the ED to predict perforation, and even grade (severity) of perforation. And, unlike most of the ML models available out there in health care, we plan to validate the winning algorithm in an external, prospectively-collected data set. The ultimate goal would be to embed the algorithm in the standard ED workflow, hence generating the prediction live for each patient.

Now the real work starts!