ICA-Derived Respiration Using an Adaptive R-Peak Detector

Breathing Rate (BR) plays a key role in health deterioration monitoring. Despite that, it has been neglected due to inadequate nursing skills and insufficient equipment. ECG signal, which is always monitored in a hospital ward, is affected by respiration which makes it a highly appealing way for the BR estimation. In addition, the latter requires accurate R-peak detection, which is a continuing concern because current methods are still inaccurate and miss heart beats. We describe a systematic approach for robust and accurate BR estimation based on the respiration-modulated ECG signal. Increased accuracy is obtained by a time-varying adaptive threshold for QRS complex identification, and discriminating high amplitude Q-peaks as false R-peaks. Improved robustness derives from the use of an Empirical Mode Decomposition (EMD) approach to R-peak detection and an Independent Component Analysis (ICA) used to separate out the respiration signal in the frequency domain as opposed to the more usual time domain approaches. The performance of our system, tested on real data from the Capnobase dataset, returned an average mean absolute error of less than 0.7 breaths per minute compared with up to 15 breaths per minute produced by some of the best time domain analysis approaches. Additionally, the QRS detector component part of our system is competitive with the best current published methods, achieving a detection rate of over 99.80% on real data.