Breathing signal combining for respiration rate estimation in smart beds

Abstract

One of the non-invasive ways to measure respiratory effort is in-bed pressure sensor arrays. Based on the area of the bed and the sensor array covered by a patient's body, some sensors may not include significant respiratory effort components or may have low signal to noise ratios. When combining signals from the different sensors, this can produce a low quality output signal. Signal combiners can overcome this problem. This paper describes two different methods of signal combining to achieve a good estimation of the respiratory rate and the respiratory signal itself. To assess the performance, a participant was asked to lay on the bed in supine position while having normal breathing. Our results indicate that both methods can perform very satisfactorily when compared to a gold standard signal, and that they can outperform some previously published methods.