Application of wavelets based multiresolution analysis to detect relevant points of interest from finger-tip photoplethesmography and pressure signal from the radial artery

The purpose of this study is to apply and evaluate performance of wavelets and select an appropriate wavelet to detect the peaks and dicrotic notch from the waveform captured from infrared finger-tip photoplethesmography (PPG) and detect the significant peaks and valleys from the signal captured from the radial artery. Accurate detection of dicrotic notch is important as it finds its usage in various medical applications viz. pulse oximeters, vascular diagnostics, digital beat-to-beat blood pressure measurement systems and autonomic function analyzers. The method was extended to detect points of interest in pressure signal from radial artery. Radial artery signals were captured from a robust system called `Nadi Yantra' [1] developed for the validation of ancient method of non-invasive pulse diagnosis of ailments. The system uses piezoelectric based pressure system to capture the palpations from the radial artery. To characterize the signals, wavelets were applied and the percussion wave, tidal wave and dicrotic notch were detected. The techniques developed were executed on a set of 120 data samples from healthy subjects. Daubechies wavelet was found to be suitable for PPG and radial artery pulse analysis. Overall efficiency of 98.45% for PPG signal and 97.89% for radial artery signal was achieved.