Paulo Lopez-Meyer, Brian Cheung and Edward Sazonov, Proceedings of the 6th International Conference on Sensing Technology, Dec. 18th-21st, 2012, Kolkata, India, pp. 811-814.



The capability of three different sensors was analyzed in this study with the aim to build a simple electronic, wearable and non-invasive system to accurately capture human respiratory signals in free-living conditions. Two of the sensors were based in the piezoelectric effect and the third one was resistive in nature. These sensors were compared to the Respiratory Inductive Plethysmograph (RIP) technique used as a gold standard. RIP utilizes elastic belts to capture the expansion and contraction of the lungs based on changes in the belt inductance. The possibility of using less expensive sensors was explored in this study and the compared to the RIP reference respiration signal. Results suggest that data collected using a single piezoelectric belt correlates significantly with the reference signal (ρ>0.81), regardless of the position of the sensor for passive activities like sitting and reading. However, lower correlation was observed for activities (such as like walking and eating) that  involve substantial motion of body parts.