MIT EECS - DENSO Undergraduate Research and Innovation Scholar
Applying Machine Learning to ECG Datasets to Predict Cardiac Risk
Collin M. Stultz
The goal of this research will be to find other computergenerated metrics such as morphologic variability that identify high-risk cardiac patients and also to try and improve, using machine learning, the way morphologic variability can be used as a predictive tool. A large picture goal of the research will be to be able to assign a patient a single number based off of the analysis of their ECG data that communicates how at-risk they are for death in the immediate future from cardiac causes.
My background in Cognitive Science when merged with my interests in Computer Science draws me toward machine learning and artificial intelligence opportunities. My past research involving using computer generated metrics to understand patterns of search behavior and human motor learning lends itself nicely to machine learning applications. Trying to better predict how at-risk cardiac patients are using machine learning complements my interests well.