Project Title

Wearable ECG Device for Arrhythmia Detection

Presenters

Grant BurkeFollow

Academic department under which the project should be listed

SPCEET - Electrical and Computer Engineering

Research Mentor Name

Jeffrey Yiin

Abstract (300 words maximum)

Many deaths caused by cardiovascular disease, the leading cause of death worldwide, could be avoided if the underlying causes are detected. This can be done with electrocardiograms (ECGs), which has long been done in clinical settings. In recent years, these devices have been made small and cheap enough to be used by consumers outside of clinical settings. For example, the popular Apple Watch generates a signal that is similar to a single-lead ECG and detects Atrial Fibrillation (AFib), a condition characterized by irregular beating of the heart and is linked to an increased risk of heart failure and stroke, with high accuracy. The purpose of this study is to explore methods of creating wearable ECG devices and develop a machine learning model for detecting AFib. For developing the machine learning model, I used Convolutional Neural Networks on publicly available datasets for training. The expected outcomes are to build a proof-of-concept ECG device to obtain clean and useful heartbeat signals for further analysis and to achieve a high accuracy in detecting irregular heartbeats.

Disciplines

Computer Engineering

This document is currently not available here.

Share

COinS
 

Wearable ECG Device for Arrhythmia Detection

Many deaths caused by cardiovascular disease, the leading cause of death worldwide, could be avoided if the underlying causes are detected. This can be done with electrocardiograms (ECGs), which has long been done in clinical settings. In recent years, these devices have been made small and cheap enough to be used by consumers outside of clinical settings. For example, the popular Apple Watch generates a signal that is similar to a single-lead ECG and detects Atrial Fibrillation (AFib), a condition characterized by irregular beating of the heart and is linked to an increased risk of heart failure and stroke, with high accuracy. The purpose of this study is to explore methods of creating wearable ECG devices and develop a machine learning model for detecting AFib. For developing the machine learning model, I used Convolutional Neural Networks on publicly available datasets for training. The expected outcomes are to build a proof-of-concept ECG device to obtain clean and useful heartbeat signals for further analysis and to achieve a high accuracy in detecting irregular heartbeats.