Heart rate variability: Analysis and classification of healthy subjects for different age groups

Heart rate and its variability contain subtle information about cardiac health, which are very much difficult to extract without a suitable computer based computation technique. The objective of this study is to evaluate the HRV analysis of healthy subjects of different age groups by features computed from time domain, frequency domain and non-linear methods and classify those using SVM, KNN and PNN multi classifier modules with PCA as a feature space reduction technique. For short term HRV analysis, 60 normal and healthy male volunteers were chosen for ECG recording. 20 subjects in each age group of 18 to 30 years: named as young, 30 to 45 years: named as middle and 45 to 60 years: named as old were considered. In this study, analysis results show that, young age group has higher values in all features than other two groups and PCA-PNN classifier module classifies with a better accuracy of 70% and sensitivity 70% with all age groups.