A Region of Interest Selection for Vision-Based Respiratory Rate Measurement in Sleeping Scenario

Bing-Fei Wu,Kuan-Hung Chen,Chun-Han Lin,Da-Hong He,Po-Wei Huang

doi:10.1109/access.2020.3012676

Bing-Fei Wu, Kuan-Hung Chen + Show 3 more

Open Access

https://doi.org/10.1109/access.2020.3012676

Copy DOI

Journal: IEEE Access	Publication Date: Jan 1, 2020
Citations: 35	License type: CC BY 4.0

Affiliation: National Yang Ming Chiao Tung University

Abstract

Vision-based respiratory measurement can remotely measure respiratory information without affecting the sleep quality of the user. However, several scenarios during sleeping, such as when blankets cover areas of the face or body, can be a challenge when estimate respiratory rate (RR) due to the uncertainty of the region of interest (ROI). In this study, we first investigate the metrics and physical meanings of ROI selection from the perspectives of array signal processing and the concept of linear combining. Then, we propose an ROI detection algorithm based on both temporal and spatial consistency (TSC), which aims to extract the representative characteristics of respiration with fewer computational resources. Furthermore, a sleeping database, containing more than $50~hours$ of data, was built to investigate the performance of benchmarked ROI methods during long-term sleeping. The experimental results demonstrate that TSC attains the highest signal-to-noise ratio (SNR) at $20.6~dB$ and a relatively low elapsed time of 100.2 frames per second (fps). For the RR estimation accuracy, TSC reduces the mean absolute error from 1.65 breaths per minute ( $bpm$ ) to $0.97~bpm$ .

Highlights

In the field of health care applications, vital signs are critical indicators to identify the health status of human beings
The model can work without region of interest (ROI), the results showed that the mean of absolute error is larger when accurate ROI is not available
The contributions of this paper two-fold: (1) we analyzed the influence mechanism of ROI selection and established a benchmark for evaluating ROI selection strategies, and (2) we proposed an ROI combining method based on temporal and spatial consistency (TSC)

Summary

Introduction

In the field of health care applications, vital signs are critical indicators to identify the health status of human beings. Respiratory rate (RR) is one of the primary vital signs. One well-established adult respiratory issue is obstructive sleep apnea (OSA). Despite the high incidence of OSA [1], many patients are unaware of their symptoms until reminded by their partners [2]. This phenomenon puts the patient in a long-term state of poor sleep quality, which interferes with concentration during work or study. If the RR during sleeping situations could be monitored, OSA symptoms would be detected and treated earlier [3], [4]

Objectives

Methods

Results

Conclusion