Wearable vital signs monitoring device with machine learning

Abstract

Cardiovascular diseases (CVDs) is the leading cause of death in the world . According to the World Health Organization, about 17.5 million lives are lost to CVDs in 2019 which is around a 1/3 of the total number of deaths in the world. Examples of CVDs are stroke, coronary artery disease and heart failure. One of the simplest ways to measure a person’s heart rate is through the use of a wearable heart rate monitor device such as an electrocardiogram of a fitness tracker.

This project aims to collect heart rate data from a cheap and reliable fitness tracker to develop machine learning techniques. An affordable fitness tracker is required as it should be readily available to as many people as possible. As the project requires accurate data, repeated testing would be required as well as cross checking with more reliable sources of data. The random forest algorithm has been found to be the most reliable in all situations while the k nearest and naïve bayes can also be used at higher ranges of heart rates.