Coherence analysis between respiration and heart rate variability using continuous wavelet transform.

The continuous wavelet transform (CWT) is specifically efficient in the analysis of transient and non-stationary signals. As such, it has become a powerful candidate for time-frequency analysis of cardiovascular variability. CWT has already been established as a valid tool for the analysis of single cardiovascular signals, providing additional insights into the autonomous nervous system (ANS) activity and its control mechanism. Intercorrelation between cardiovascular signals elucidates the function of ANS central control and the peripheral reflex mechanisms. Wavelet transform coherence (WTC) can provide insight into the transient linear order of the regulatory mechanisms, via the computation of time-frequency maps of the time-variant coherence. This paper presents a framework for applying WTC for quantitative analysis of coherence in cardiovascular variability research. Computer simulations were performed to estimate the accuracy of the WTC estimates and a method for determining the coherence threshold for specific frequency band was developed and evaluated. Finally, we demonstrated, in two representative situations, the dynamic behaviour of respiration sinus arrhythmia through the analysis of the WTC between heart rate and respiration signals. This emphasizes that CWT and its application to WTC is a useful tool for dynamic analysis of cardiovascular variability.