Abstract
Human respiratory signal provides important information in modern medical care. In daily life, respiratory signal is usually captured under different motion states with the help of Electrical impedance pneumography (EIP). Consequently, the captured signal is easily corrupted by electronic/electromagnetic noise, internal mechanical vibration of the lung and motion artifacts. Because respiratory signal and interferences co-exist in an overlapping spectra manner, classical filtering method cannot work here. In this paper, we present a new signal processing method for eliminating the noise and interferences included in EIP signal, by separating the correlated motion artifacts from the raw EIP and 3-axis Acceleration (ACC) signals, restoring the rough respiration signal from the mixed signal, and further processing using wavelet analysis approach. Results are compared to traditional denosing algorithms by wiener filter, which indicates that the new signal processing method we presented is suitable for EIP signals under the motion states.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
