Study and evaluation of denoising and baseline wandering of computerized wrist pulse signal using virtual instrument

Pulse signal contains rich diagnostic information of human body and it reflects the overall status of body which has been proved in Ayurveda for thousands of years. Computerized pulse diagnosis get more and more research concerned in recent years. Computerized pulse signals collected from various sensors, brings in noise inevitably. During the sampling period, baseline of pulse signal waveform always wanders. So, it is necessary to preprocessed the raw signal waveform to get more detailed waveform as it improves the final analysis performance before the feature extraction phase. Pre-processing is an important and crucial step which decides the whole performance of pulse analysis and pattern classification. Pre-processing is classified into three main parts such as denoising, baseline wander, and normalization. The main aim of this paper is to discuss and implement various preprocessing algorithms of pulse waveforms. Different algorithms of denoising and baseline drift removal are implemented by using LabVIEW software and results are evaluated by different parameters.