Robust estimation of respiratory rate based on linear regression

Among the many parameters of human life, which are subject to intensive monitoring, one can specify the frequency of of respiratory action. The measurement of such physical quantity can be performed directly by tracking the activity of the respiratory organs, as well as indirectly through the breathing frequency estimation based on the ECG signal. This paper presents a method to assess the respiratory rate based on robust estimation of the regression function for QRS electrocardiogram signal in conjunction with the estimation of the spectral density of the time series.