Impact of speech on non-invasive sleep metric measurements using an FSR sensor placed under a mattress
Master
Semester programme:Master of Applied IT
Tom Verlinde
Project description
This study has tried to increase the understanding of FSR-based systems by investigating the effects of speech. It aims to create a setup for future research related to snoring and sleep talking.
Context
Collecting reliable sleep metrics requires people to go to specialised places where current standard methods can be used.
Various studies have recently been performed on measuring heart rate (HR) and respiratory rate (RR) using noninvasive methods.
A subset of these studies has focused on low-cost and reliable methods.
This subset has explicitly taken an interest in FSR sensors.
FSR sensors are cheap and have favourable characteristics.
However, these studies have only been performed in recent years, and because of that, the test environments that have been set up have been rather ideal.
Results
Speech significantly negatively impacts the estimation of both HR and RR.
HR estimation suffers the most from speech.