Human identification using time normalized QT signal and the QRS complex of the ECG

In this paper the possibility of using the ECG signal features for Biometrics identification is investigated. A test set of 550 lead I ECG traces recorded from 22 healthy individuals measured at different times are used to validate the system. The proposed system extracts special parts of the ECG signal starting from the QRS complex to the end of the T wave. Two different approaches are used to compensate for the change in the signal duration with the change in Heart Rate, By the first approach time domain normalization according to Framingham correction formula[1] is used. By the second approach the QT signal to a fixed length is applied. Finally the ability to use only the QRS complex for identification without any time domain manipulation has been investigated. Selected DCT coefficients obtained from the normalized signals were introduced to a Neural Network based classifier. Three methods are proposed which give identification rates ranging from 97.727 % to 99.09%.