Respiration Rate Estimation from Electrocardiogram Signal for Wearable System Application

Abstract

Continuous respiratory monitoring is crucial for the early detection of fatal diseases. Invasive respiratory devices that restrict mobility are used for respiratory monitoring. Implementing these sensors in wearable devices is not always easy. In this study, respiratory rate was estimated from the ECG signal without using an additional sensor. The ECG-derived respiration (EDR) signal was obtained by interpolating the R-peak amplitude values of the ECG. This signal was processed using a bandpass filter and the Fast Fourier Transform (FFT) to estimate respiratory rate. All operations are performed on a fixed-point microcontroller with low power consumption and low processing power for maximum speed and minimum memory usage. The accuracy of the implemented algorithm was demonstrated using data from a BIOPAC device.