ECG arrhythmia recognition using artificial neural network with S-transform based effective features

Abstract

In this paper, a potential application of Stockewell transforms (S-transform) is proposed to classify the ECG beats of the MIT-BIH database arrhythmias. Feature extraction is the important component of designing the system based on pattern recognition since even the best classifier will not perform better if the good features are not chosen properly. In this study, S-transform is used to extract the eight features which are appended with four temporal features. In this work, the performances of two approaches are compared to classify the five classes of ECG beats which is recommended by AAMI EC57 1998 standard (Association for the Advancement of Medical Instrumentation). The first approach uses temporal and S-transform based feature set, whereas the second approach uses the wavelet transform based features. These features from two approaches are independently classified using feed forward neural network (NN). Performance is evaluated on several normal and abnormal ECG signals of the MIT-BIH arrhythmia database using two techniques such as temporal and S-transform with NN classifier (TST-NN) and other wavelet transform with NN classifier (WT-NN). The experimental results demonstrate that the TST-NN technique shows better performance compared to the WT-NN technique.

Electrocardiogram stream level correlated patterns as features to classify heartbeats for arrhythmia prediction

“…The researchers assured that patient-related training data might augment the ECG classification performance. Das and Ari (2013) utilized statistical values, and temporal features are taken from the stock-well transform to classify by utilizing feed-forward NN. Das and Ari (2014a, b) utilized statistical values of stock-well transform and wavelet features along with the temporal features for categorizing AAMI recommended arrhythmia classes through a multilayer perceptron of NN.…”

Noise-tolerant modular neural network system for classifying ECG signal

“…The authors developed a three-layered NN with one hidden layer of 20 neurons that classified RR interval signals into four arrhythmia categories with an average accuracy of 99.3%. Das and Ari employed an NN approach with a pre-processing band-pass filter to reduce noise in 44 records of the database [18]. They classified five types of ECG beats with an S-transform NN approach and achieved 97.9% of average classification accuracy.…”

“…Noise corruption can generate similar morphologies to the ECG waveform, reducing the discriminating power of heartbeat patterns, and increasing the rate of false alarms for cardiac monitors [9]. Therefore, a large number of NN approaches for ECG classification have included signal preprocessing for noise reduction, using a wavelet transformer (WT) [6,[11][12][13], nonlinear cubic spline interpolation (CSI) [14], fast Fourier transformation (FFT) [15] or band-pass filters [16][17][18][19].…”