A support vector machine classifier using digital volume pulse waveforms predicted high or low arterial stiffness with >85% accuracy compared to aortal pulse wave velocity.
Cross-Sectional
Does digital volume pulse measurement combined with an SVM classifier accurately predict arterial stiffness compared to aortal pulse wave velocity?
A novel method using digital volume pulse and SVM classification can accurately predict arterial stiffness, offering a rapid, non-invasive tool for cardiovascular disease risk assessment.
Effect estimate: >85% accuracy
A method for rapidly assessing a patient's arterial stiffness and hence risk of developing cardiovascular disease (CVD) without resorting to laborious blood tests is presented. Simple measurement of a patient's volume pulse measured at the finger-tip (digital volume pulse) using an infrared light absorption detector placed on the index finger is sufficient to predict their CVD risk. Suitable features are extracted from the waveform and a support vector machine (SVM) classifier has been found to make accurate (>85%) prediction of high or low arterial stiffness as indicated by the aortal pulse wave velocity (PWV). This would otherwise require an extensive and time consuming procedure, and hence this new method is promising as a tool to help health professionals prevent cardiovascular diseases.
Alty et al. (Thu,) conducted a cross-sectional in Cardiovascular disease / Arterial stiffness. Support vector machine (SVM) classifier on digital volume pulse vs. Aortal pulse wave velocity (PWV) was evaluated on Prediction of high or low arterial stiffness (>85% accuracy). A support vector machine classifier using digital volume pulse waveforms predicted high or low arterial stiffness with >85% accuracy compared to aortal pulse wave velocity.