Explainable AI (XAI) in Healthcare- an authentic data-driven design

This project focuses on designing an explainable AI (XAI) application that can predict length of stay (LOS) of patients admitted into hospitals. The LOS of a patient admitted to a hospital can be a pivotal measure in the overall health of the patient. LOS measures are affected by many different factors such as type of admission, type of insurance/ lack of insurance, age, pre-existing conditions, diagnosis, etc. Our goal is to design with a focus on authenticity with a long-term mindset of patients becoming our target users. Currently, we are using Shapley diagrams to visualize our explanations. The explanations from our model can be used as a tool to aid physicians in making decisions that can greatly improve the care that patients receive.