Feature extraction from ECG for classification by artificial neural networks

The ability of properly trained artificial neural networks to correctly classify patterns makes them particularly suitable for the interpretation of ECG (electrocardiography) signals. Attention was given to three classes of ECGs, namely, normal and two cardiac myopathies, and anterior and inferior infarctions. Suitable features were extracted from the digitized bipolar limb lead ECG signals, and results are presented to show that a multilayer perceptron can correctly discriminate between the three chosen classes.<>