Real-time Online Visualization of Patient Severity in Intensive Care Units

Description: ICU dashboards often overload physicians by displaying an enormous amount of patient's information but fail to provide a snapshot or a summarized patient's story that physicians can understand at a glance. Visual analytics based on Artificial Intelligence can be a powerful way to simplify complexity in clinical decision making. We have recently developed an unsupervised learning algorithm to create insightful visualizations of patient severity in real time. The student will engage in our ongoing effort in building better visualization systems for massive health data. This project aims to create an online visual analytic system based on our developed algorithm for the MIMIC II ICU database.

Contact Info: rjoshi@mit.edu