Human Identification by ECG Signals through Neural Network

Abstract

A cardiogram (ECG) may be a bioelectrical signal that records the hearts electrical activity versus time. it's a vital diagnostic tool for assessing heart functions. The interpretation of graphical record signal is AN application of pattern recognition. The techniques utilized in this pattern recognition comprise: signal pre-processing, QRS detection, feature extraction and neural network for signal classification. totally different graphical record feature inputs were utilized in the experiments to check and notice a fascinating options input for graphical record classification. Among totally different structures, it had been found that a 3 layer network structure with twenty five inputs, five neurons within the output layer and five neurons in its hidden layers possessed the most effective performance with highest recognition rate of ninety one.8% for 5 viscus conditions