Frequency parameters from myographic signals for the evaluation of respiratory muscle activity during an increased ventilatory effort

Reliable estimation of respiratory muscle activity is especially important in patients with obstructive sleep apnea syndrome (OSAS). Changes in power spectral density (PSD) of electromyographic (EMG) and vibromyographic (VMG) signals are related with muscle activity and fatigue. We propose alternative frequency parameters to characterize these signals. They are based on the cumulative PSD and are more robust than classical parameters (mean and central frequencies) in the presence of white random and cardiac noise. A comparative study between OSAS patients and normal subjects during an increased respiratory effort is performed with both types of frequency parameters (FP) calculated for myographic signals from three respiratory muscles: genioglossus, sternomastoid and diaphragm. All the FP are obtained in auto and cross-spectral analysis. Both classical and alternative FP provide interesting information about VMG and EMG signals, respectively, in order to differentiate muscle function between normal subjects and patients.