Estimation of Respiratory Signals using Tracheal Sound and Movement

Abstract

Introduction: Sleep apnea (SA) is clinically assessed using respiratory signals: thoracoabdominal (TA) movements and airflow, recorded during overnight polysomnography (PSG). A simpler portable device with the ability of estimating respiratory signals is more convenient and cost-effective. Tracheal sound and movements can be conveniently recorded with simple instruments. Objectives: To estimate respiratory signals using tracheal sounds and movements. Methods: Participants >18 yrs with suspected sleep apnea underwent overnight PSG. Simultaneously, tracheal sound and movements were recorded using an embedded microphone and accelerator placed over the tracheal notch. Linear regression was used to estimate TA movements from sound energy and the range of movements. Also, the sound energy envelope was extracted as a surrogate of airflow level. Repeated measure correlation was used to calculate the agreement between reference and estimated signals in normal breathing, apneas and hypopneas. Results: Forty participants (21 females, age: 49 ± 16, BMI: 29.5 ± 6.1, and AHI: 17.9 ± 19.6) were included. Estimated movements correlated significantly with TA movements during normal breathing (0.53±0.06*), hypopneas (0.65±0.07*) and apneas (0.71±0.05*). Airflow correlated with PSG-airflow during normal breathing (0.29±0.08*), hypopneas (0.47±0.07*) and apneas (0.63±0.09*). (*P Conclusion: Tracheal sounds and movements are promising tools for SA assessment.