Non Invasive Cardiovascular Health Assessment in Post-adolescent Age Group Using Augmentation Index

This paper presents a non-invasive method of classifying a subject’s health into either one of two classes depending on the condition of the subjects’ cardiovascular health. Novelty of the work lies in segregating the subjects who had good cardiovascular health and the subjects who were at risk. In the proposed work, VPG signals and other general information such as weight, height, BP were collected from 38 individuals in the post-adolescent age group. Furthermore, a signal from a person who was known to have good cardio vascular health was collected. This signal became the reference that was used to compare with other signals. After the initial pre-processing, the PPG and APG were obtained from the VPG. Then a total of 7 features of the wave contour from the APG and PPG signals were extracted. Based on the Augmentation Index the signals were classified into two classes using SVM and ELM classifiers. Where one class represented healthy individuals and the other class represents the individuals at risk of CVD. The average values of the extracted features were used and the final accuracy obtained was also an average value. The accuracy obtained using ELM classifier with K-fold cross validation was 77%. Whereas the efficiency achieved using SVM was 94.59%. Hence the proposed method can be used to assess the vascular health analysing PPG signals in the post-adolescent age group.