Application of the empirical mode decomposition method to the analysis of respiratory mechanomyographic signals

The study of the mechanomyographic (MMG) signals during dynamic contractions requires a criterion to separate the low frequency (LF) component (basically due to gross movement of the muscle or of the body) and the high frequency (M) component (related with the vibration of the muscle fibers during contraction). In this study, we propose to use the Empirical Mode Decomposition method in order to analyze the Intrinsic Mode Functions of MMG signals of the diaphragm muscle, acquired by means of a capacitive accelerometer applied on the costal wall. This signal, as the MMG signals during dynamic contractions, has a LF component that is related with the movement of the thoracic cage, and a HF component that could be related with the vibration of diaphragm muscle fibers during contraction. The method was tested on an animal model, with two incremental respiratory protocols performed by two non anesthetized mongrel dogs. The results show that the proposed EMD based method provides very good results for the cancellation of low frequency component of MMG signals. The obtained correlation coefficients between respiratory and MMG parameters were higher than the ones obtained with a Wavelet maltiresolution decomposition method utilized in a previous work.