Data visualization is a key component of health data science. Even brilliant data analytics may not end up being useful if the results are not communicated in an effective manner. Modern data visualization techniques provide some of the most powerful ways to interpret health data, information, and knowledge. We strive to not only present our findings in a visually engaging manner but also devise novel visualization methodologies and tools.
Examples of our work can be seen here:
- An investigation of patterns in hemodynamic data indicative of impending hypotension in intensive care (PDF)
- Empirical relationships among oliguria, creatinine, mortality, and renal replacement therapy in the critically ill
- Extensions and Applications of Event Charts
- Event history graphs for censored survival data
We have also developed a web-based data visualization tool for Multiparameter Intelligent Monitoring in Intensive Care II (MIMIC-II) which is freely available to anyone.