Investigations on certain cardiovascular risk level detection techniques using PPG signals

Abstract

for new techniques to constantly diagnose diseases by analysing the biosignals recorded from the patients. Cardiovascular disease (CVD) is one of the most important non-communicable diseases causes large number of deaths worldwide. The CVDs leads to severe disability and also increases medical expenditure. Due to the critical nature of cardiac monitoring, it is important to spend effort and resources to strengthen the current techniques or to create new ones. Accessibility, low cost of use, non-invasiveness would be some desirable traits for a new system. Such a device should be ideal for domestic use, so that people may use it. The non-invasive approach would help to reduce the risks associated with pathogens. Photoplethysmography (PPG) signal has all the requisite features. Finger PPG is a widely used method in the medical field research and it is based on a clear assessment of the optical properties of a particular skin region. Study of the PPG waveform has recently drawn more interest particularly in respiratory and circulatory monitoring. The proposed methods have been validated using several PPG recordings with a wide variety of waveform morphologies from capnobase database which is available online. newlineThis research work investigates to analyze the dimensionality reduction techniques, non-linear classifiers, evolutionary algorithms and hybrid classifiers for abnormality detection of PPG signals. Also, CVD risk prediction system is built to discriminate the various risk stages of cardiovascular disease using photoplethysmography signals. newline