Real-time sleep prediction using a virtual sensor to estimate Heart Rate Variability through Respiratory Rate

Abstract

One of the most important causes of death while driving is sleepiness. To solve this problem, different kinds of technologies are needed. A recent work presented an approach based on Photoplethysmogram (PPG) analysis to predict the sleep onset. As PPG is not always available, especially in the case of commercial of the shelf wearable devices that provide features such as heart beat and respiration rate, in the paper we present a novel approach to predict sleep onset, which leverages a virtual sensor able to provide an estimation of the PPG-related Heart Rate Variability (HRV) through Respiration Rate (RR) analysis. The experimental results show 100% sensitivity and specificity in the collected data.