A Novel Method for Estimating Respiration Rate based on Ensemble Empirical Mode Decomposition and EKG Slope

The clinical monitor now mostly uses impedance IP (impedance pneumography) to measure respiratory signals. While in breathing, the movement of chest leads to position change of the EKG (Electrocardiogram) electrodes on the skin resulting in a change in impedance which can be used to estimate the respiratory rate. Measuring the EKG's impedance change for estimating the respiratory rate requires some specialized hardware. Other indirect methods for estimating respiratory rate, such as the EDR (EKG Derived Respiration), just simply utilize the EKG signal making use of the inherent variations in respiration wherein the respiratory rate is obtained from the parameter variations within the EKG waveform including RSA (Respiratory Sinus Arrhythmia) and R Peak Amplitude (RPA). This study proposes a new EDR method in which the square of the slope of the EKG waveform is calculated first and then followed by the moving average. The respiratory rate is obtained by the proposed algorithm that employs the modulated time series and compared to the results from RPA and RSA methods. The new method uses EEMD (Ensemble Empirical Mode Decomposition) to remove noise from EKG, reconstructs the respiratory signal by selecting the right IMF (Intrinsic Mode Function) as respiratory signal, and finally compares it with the nasal mouth pressure reference respiratory signal. The new RSS (R-peak Slope Square) method works with adaptive signal processing tool EEMD to obtain the EDR exploring the potential feasibility of clinical application in the future. The results demonstrate that the innovative methods proposed by this study are more accurate than that from RSA in elderly monitoring and nearly same performance as RPA (R-peak Amplitude) as well.