Artificial Bees Colony Optimized Neural Network Model for ECG Signals Classification

摘要

The ECG signal is a representation of bioelectrical activity of the heart's pumping action. The doctor regularly uses a temporal recording of ECG and waveforms characteristics to study and diagnose the overall heart functioning. In some heart diseases, the correct diagnosis in an early time is essential for the patient survival. This need leads to the necessity to automate normal beat signals discrimination from abnormal beat signals. In our study, we have chosen the Multilayer Perceptron (MLP) as a classifier for this type of signals into two categories: normal (N) and pathological (V). To train this network, we used the database "MIT BIH arrhythmia database." This training is improved using a novel swarm optimization algorithm called Artificial Bees Colony (ABC) inspired from the foraging intelligence of honey bees. The (ABC) has the advantage of using fewer control parameters compared to other swarm optimization Algorithms. We propose several algorithms to filter, detect R peaks and extract the features of cardiac cycles to get it ready to be classified.