Research Areas
Predictive Analytics
Predictive analytics uses historical data to predict future events. Predictive analytics has growing applications in healthcare to predict disease risk, assist in clinical decision making, and improve patient outcomes.
Natural Language Processing
Natural language processing (NLP) aims to understand the "meaning" of free text and speech. This use of unstructured data can uncover patterns and features that is otherwise difficult to find using structured data within the EHR. Several of our projects use NLP models to draw insights from large amounts of clinical notes.
Decision Science
Description
Knowledge Graphs
Description
Computational Phenotypes
Computational phenotyping is a biomedical informatics method that identifies patient populations and features from electronic health record data. Our work uses unsupervised learning techniques to discover novel phenotypes that may not be found using preconceived groupings.
Healthcare Data Engineering
As healthcare grows digitized, electronic health data is being generated and stored in increasingly larger amounts. Data engineering helps us ensure this data is usable and accessible to end-users. Our work aims to build data pipelines for retrieving emergency medicine data for researchers.