A modified frequency slice wavelet transform for physiological signal time-frequency analysis

As a new signal processing tool, a Modified frequency slice wavelet transform (MFSWT) is proposed for physiological signal time-frequency analysis in this study. The transform generates time-frequency representation from the frequency domain, and the reconstruction is independent of frequency slice function (FSF). To realize accurate time-frequency location of signal components, a bound signal-adaptive FSF was introduced to serve as a dynamic frequency filter for the transform. This method avoids troublesome parameter selection, is signal-adaptive and easy to use. The results of two case studies demonstrate the validity of the proposed method, which has good interpretability with high time-frequency resolution, and hence has significant potential for bio-signal time-frequency analysis.