Heart rate detection: Fractional approach and empirical mode decomposition

The comparison of two methods used to separate electrocardiography and electromyography signals: Empirical Mode Decomposition (EMD) and Fractional Order Calculus (FC) is presented. For quantitative evaluation, crest factor (CF) and signal-to-noise ratio (SNR) values were calculated, where higher CF and SNR values were preferable. Results suggest that FC and EMD performance can be partially assessed by CF indicating better artifact cancellation with EMD approach by 9.72%. For FC and EMD, automatic fractional order application and intrinsic mode function selection were suggested and discussed, respectively. In summary, the results suggested that proposed filtering assessment should be enhanced for more appropriate FC and EMD method's assessment.