1
2
*Address for correspondence: Dr. Uma Maheswari Krishnaswamy, Associate Professor, Department of Chest Diseases, M.S. Ramaiah Medical Teaching Hospital, MSRIT Post, New BEL Road, Bangalore-560054. e-mail: umamohan99@gmail.com, Phone: 91-80-22183047, 09483702412
Respiratory rate is conventionally recorded using nasal thermocouples, spirometers and plethysmographs and is useful in ambulatory monitoring. As ECG is also a commonly used ambulatory diagnostic tool. An attempt has been made to derive respiratory signals from ECG in this study.
Development of the algorithm was based on the fact that respiration causes modulation of the R peak in the ECG. This algorithm was applied to signals acquired from Physionet, BioPac and Polysomnograms for validation.
When the ECG-derived respiratory (EDR) signal obtained by applying the algorithm to ECG was correlated with traditionally recorded respiratory signals, the correlation coefficient ranged between 0.7 and 0.75. Besides, the plot of fast Fourier transform (FFT) of both signals was found to be similar.
This study can be further extended for real-time processing of signals in PSG. Besides, it can be used to detect apnoeas in patients undergoing Holter ECG.
ECG-derived respiration, sleep apnoea
