Arrhythmia Detection Using Empirical Mode Decomposition and Boosted Trees in Electrocardiography Signals

Nowadays, heart diseases that cause death have become widespread. Electrocardiography is a biomedical signal commonly used in the diagnosis of these diseases. In this study, a technique which can be used for detecting arrhythmia as a result of ECG examination is proposed. In order to detect arrhythmia, Empirical Mode Decomposition and Singular Value Decomposition were used. Empirical Mode Decomposition is an appropriate technique for analysis of the stationary, non-linear series and uses oscillation signals at the local levels. It separates the signals into oscillation structures called Intrinsic Mode Functions. Singular Value Decomposition is an algebraic method used to reduce the size of complex data sets and is used to reduce noise effects. After reducing the effect of noise and obtaining the appropriate features, the classification was made by using Boosted Trees. Accuracy, sensitivity, and specificity values were calculated to evaluate the performance of the classification