Hearables: AI for in-ear doctorless cardiovascular care in the community

Lead Research Organisation: Imperial College London
Department Name: Dept of Computing

Abstract

Wearable technology is envisaged to revolutionise the way we approach the monitoring of health and wellbeing, however, for this to become a reality, physiological sensors should be integrated into a discreet, comfortable, unobtrusive and non-stigmatizing device. Our earlier studies introduced a novel in-ear device for the monitoring of cardiac activity from sensors embedded on a one-fits-all earplug that is worn inside the ear, so-called Hearables. Cardiac activity is of particular interest since it can serve as a keyhole into the broader picture of the general health status of a person. At an extreme, an early prediction of abnormal heart rate may even save the life of a person, but for this to become reality 24/7 continuous ECG monitors are a prerequisite. Our novel in-ear ECG sensor has already been evaluated against Chest-ECG in terms of the accuracy of the acquired ECG waveform, and has for the first time demonstrated the ability of Ear-ECG to detect an abnormal heartbeat caused by ventricular bygeminy. This project will further comprehensively examine medically relevant parameters, such as the timings of the characteristic waves within Ear- ECG, to reliably capture both normal and abnormal ECG waveform morphologies. The corresponding AI platform will employ deep models for data conditioning, finding patterns and detecting anomalies in ear-ECG.

Publications

10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
EP/S023283/1 01/04/2019 30/09/2027
2449184 Studentship EP/S023283/1 05/10/2020 04/10/2024 Edoardo Occhipinti