Contactless Sleep Monitoring Measurement Setup

Abstract

In this work, a contactless measurement setup based on a low-cost weighing scale sensor is presented. The positioning of the sensor is a key point in our work with ballistocardiographic (BCG) measurements. This was demonstrated using a strain-gauge sensor placed at the head of the mattress to record the BCG signals, while an unsupervised machine learning algorithm was implemented to detect respiratory and cardiac cycles from patients laying in different postures. Comparison of BCG with electrocardiography (ECG) has shown the ability to detect, at least, 75% of every single heartbeat with the suggested setup and algorithm irrespective of patient’s postures.

Keywords

• balliscardiography
• strain-gauge sensor
• contactless sensing
• machine learning