MIT EECS — Angle Undergraduate Research and Innovation Scholar
Inferring Patient Activity Level from Long Term ECG Signals
Collin M. Stultz
The goal of this project is to be able to infer what the patient’s activity level is based on their daylong electrocardiographic (ECG) signals. The hypothesis is that the analysis of such longterm ECG signals during everyday activity can be complementary to the shorter ECG signals obtained from a cardiac stress test. The project utilizes a wearable device that simultaneously collects both ECG and accelerometer signals of healthy volunteers doing a variety of different physical activities. We plan to use these data to build models to predict the activity level of a person from a ECG signal, and will further apply these models to existing ECG data from heart patients, to build risk models to predict patients’ risk of future adverse outcomes, such as death.
I’m excited to finally try my hands at some serious bioEECS research! I have wet-lab research experience from high school, and have taken a bunch of course 6 classes in college. It would be great to see these two different backgrounds come together.