Performance of Emprical Mode Decomposition in Automated Detection of Hypertension Using Electrocardiography

Hypertension (HPT) is one of the most common chronic diseases presented worldwide. This discomfort causes serious diseases and organ damage. Irreversible health problems may occur due to the failure of early diagnosis. Some known HPT types are difficult to detect. It is vital that HPT can be treated early and adequately. In this study, a research was conducted using ECG signals to automatically and accurately detect HPT patients. HPT detection performance was investigated using the measurements obtained from 5-level intrinsic mode function (IMF) signals with the help of empirical mode decomposition method. Achievements were compared with 9 features obtained for each IMF. As a result, HPT detection performances were obtained with decision trees and Naive Bayes algorithms using 10-fold cross validation method. The highest success rate was achieved with the decision tree algorithm with 99.99% general accuracy. We think that the high performance achieved will benefit the specialist physicians in the detection of HPT patients.