ANALYSIS AND INTERPRETATION OF ECG SIGNALS

Abstract

Electrocardiogram is the graphical recording of the electrical activity of the heart and is recognized as the biological signal used for clinical diagnosis. The ECG signal is complex in nature, and even if small noise is mixed with actual signal the numerous characteristics of the signal undergoes variation. The signal voltage level is as low as 0.5 to 5mV and is susceptible to artifacts that are larger than it. The frequency components of a human ECG signal fall into the range of 0.05 to 100Hz and as far as the noise is concerned; the muscle movements, power line interface and ambient electromagnetic interference generate it. Hence filtering remains an important issue, as data corrupted with noise must either be filtered or discarded. In this report the emerging roles of the wavelet transform in ECG preprocessing and noise removal have been discussed in a step by step approach. One of the most important causes of noise is baseline wander that effect ECG signal analysis has been removed by using a new method on wavelet transform is being proposed. So baseline wander removal is very important for the classification and analysis of ECG signal. The presented research work analyses and classifies various cardiac abnormalities using three different methods. The first technique uses the whole waveform of each beat for training the network. Thus, here literally no features are extracted, only classification is done. The second technique, extracts P and QRS complex using an ANN and then Classifies the beats to their respective classes thus the ANN acts as a detector as well as a classifier. The third technique has been used for verifying the results of the first two methods. It extracts features from the data using wavelet transforms and then classifies the beats with the help of the features extracted with the same ANN used in the other two methods. The results found were quite satisfactory, although the first method performed best; all the three methods had overall accuracies above 97%. Wavelets transform based method also effectively detects the ischemia by finding out the T wave inverted or not inverted.