Time-frequency spectral analysis of electric field plethysmography signals.

Cardiac activity contaminates respiratory electric field plethysmography (EFPG) signals, and vice versa, and is usually treated as an artefact. On the other hand, simultaneous information on respiration and cardiac activity is often required in clinical practice. Hence, we use time-frequency spectral analysis (TFSA) by means of the spectrogram (SG) to investigate the instantaneous cardio-respiratory information contained in EFPG-signals and show that it is a practical means of extracting this information out of a single EFPG-signal. Comparisons with a Wigner distribution reveal sufficient time-frequency resolution of the SG for detection of physiological events in EFPG-signals. SG-analysis indicates spectral peaks related to respiration and cardiac activity (depending on the electrode configuration, up to the third and fifth harmonic, respectively) and the possibility to detect respiratory sinus arrhythmia by means of EFPG-measurements. Intermodulation products cause overlaps of respiratory and cardiac spectra pointing out nonlinear relationships. Additionally, SG-analysis supplies essential information for the separation of EFPG-signals into respiratory and cardiac components, by means of filter techniques.