The aim of this paper is twofold. First, we present a thorough experimental study to show the superiority of the generalization capability of the support vector machine (SVM) approach in the automatic classification of electrocardiogram (ECG) beats. Second, we propose a novel classification system based on particle swarm optimization (PSO) to improve the generalization performance of the SVM classifier. For this purpose, we have optimized the SVM classifier design by searching for the best value of the parameters that tune its discriminant function, and upstream by looking for the best subset of features that feed the classifier. The experiments were conducted on the basis of ECG data from the Massachusetts Institute of Technology-Beth Israel Hospital (MIT-BIH) arrhythmia database to classify five kinds of abnormal waveforms and normal beats. In particular, they were organized so as to test the sensitivity of the SVM classifier and that of two reference classifiers used for comparison, i.e., the k-nearest neighbor (kNN) classifier and the radial basis function (RBF) neural network classifier, with respect to the curse of dimensionality and the number of available training beats. The obtained results clearly confirm the superiority of the SVM approach as compared to traditional classifiers, and suggest that further substantial improvements in terms of classification accuracy can be achieved by the proposed PSO-SVM classification system. On an average, over three experiments making use of a different total number of training beats (250, 500, and 750, respectively), the PSO-SVM yielded an overall accuracy of 89.72% on 40438 test beats selected from 20 patient records against 85.98%, 83.70%, and 82.34% for the SVM, the kNN, and the RBF classifiers, respectively.

Classification of Electrocardiogram Signals With Support Vector Machines and Particle Swarm Optimization

In this paper, we use cross wavelet transform (XWT) for the analysis and classification of electrocardiogram (ECG) signals. The cross-correlation between two time-domain signals gives a measure of similarity between two waveforms. The application of the continuous wavelet transform to two time series and the cross examination of the two decompositions reveal localized similarities in time and frequency. Application of the XWT to a pair of data yields wavelet cross spectrum (WCS) and wavelet coherence (WCOH). The proposed algorithm analyzes ECG data utilizing XWT and explores the resulting spectral differences. A pathologically varying pattern from the normal pattern in the QT zone of the inferior leads shows the presence of inferior myocardial infarction. A normal beat ensemble is selected as the absolute normal ECG pattern template, and the coherence between various other normal and abnormal subjects is computed. The WCS and WCOH of various ECG patterns show distinguishing characteristics over two specific regions R1 and R2, where R1 is the QRS complex area and R2 is the T-wave region. The Physikalisch-Technische Bundesanstalt diagnostic ECG database is used for evaluation of the methods. A heuristically determined mathematical formula extracts the parameter(s) from the WCS and WCOH. Empirical tests establish that the parameter(s) are relevant for classification of normal and abnormal cardiac patterns. The overall accuracy, sensitivity, and specificity after combining the three leads are obtained as 97.6%, 97.3%, and 98.8%, respectively.

Application of Cross Wavelet Transform for ECG Pattern Analysis and Classification

In this paper, a novel technique on a multiscale energy and eigenspace (MEES) approach is proposed for the detection and localization of myocardial infarction (MI) from multilead electrocardiogram (ECG). Wavelet decomposition of multilead ECG signals grossly segments the clinical components at different subbands. In MI, pathological characteristics such as hypercute T-wave, inversion of T-wave, changes in ST elevation, or pathological Q-wave are seen in ECG signals. This pathological information alters the covariance structures of multiscale multivariate matrices at different scales and the corresponding eigenvalues. The clinically relevant components can be captured by eigenvalues. In this study, multiscale wavelet energies and eigenvalues of multiscale covariance matrices are used as diagnostic features. Support vector machines (SVMs) with both linear and radial basis function (RBF) kernel and K-nearest neighbor are used as classifiers. Datasets, which include healthy control, and various types of MI, such as anterior, anteriolateral, anterioseptal, inferior, inferiolateral, and inferioposterio-lateral, from the PTB diagnostic ECG database are used for evaluation. The results show that the proposed technique can successfully detect the MI pathologies. The MEES approach also helps localize different types of MIs. For MI detection, the accuracy, the sensitivity, and the specificity values are 96%, 93%, and 99% respectively. The localization accuracy is 99.58%, using a multiclass SVM classifier with RBF kernel.

Multiscale Energy and Eigenspace Approach to Detection and Localization of Myocardial Infarction

This paper presents a useful technique for totally automatic detection of myocardial infarction from patients' ECGs. Due to the large number of heartbeats constituting an ECG and the high cost of having all the heartbeats manually labeled, supervised learning techniques have achieved limited success in ECG classification. In this paper, we first discuss the rationale for applying multiple instance learning (MIL) to automated ECG classification and then propose a new MIL strategy called latent topic MIL, by which ECGs are mapped into a topic space defined by a number of topics identified over all the unlabeled training heartbeats and support vector machine is directly applied to the ECG-level topic vectors. Our experimental results on real ECG datasets from the PTB diagnostic database demonstrate that, compared with existing MIL and supervised learning algorithms, the proposed algorithm is able to automatically detect ECGs with myocardial ischemia without labeling any heartbeats. Moreover, it improves classification quality in terms of both sensitivity and specificity.

ECG Analysis Using Multiple Instance Learning for Myocardial Infarction Detection

Signal processing techniques are an obvious choice for real-time analysis of electrocardiography (ECG) signals. However, classical signal processing techniques are unable to deal with the nonstationary nature of the ECG signal. In this context, this paper presents a new approach, i.e., discrete orthogonal stockwell transform using discrete cosine transform for efficient representation of the ECG signal in time–frequency space. These time–frequency features are further reduced in lower dimensional space using principal component analysis, representing the morphological characteristics of the ECG signal. In addition, the dynamic features (i.e., RR-interval information) are computed and concatenated to the morphological features to constitute the final feature set, which is utilized to classify the ECG signals using support vector machine (SVM). In order to improve the classification performance, particle swarm optimization technique is employed for gradually tuning the learning parameters of the SVM classifier. In this paper, ECG data exhibiting 16 classes of the most frequently occurring arrhythmic events are taken from the benchmark MIT-BIH arrhythmia database for the validation of the proposed methodology. The experimental results yielded an improved overall accuracy, sensitivity (Sp), and positive predictivity (Pp) of 98.82% in comparison with the existing approaches available in the literature.

ECG Signal Analysis Using DCT-Based DOST and PSO Optimized SVM

In this paper, a rule-based rough-set decision system for the development of a disease inference engine is described. For this purpose, an offline-data-acquisition system of paper electrocardiogram (ECG) records is developed using image-processing techniques. The ECG signals may be corrupted with six types of noise. Therefore, at first, the extracted signals are fed for noise removal. A QRS detector is also developed for the detection of R-R interval of ECG waves. After the detection of this R-R interval, the P and T waves are detected based on a syntactic approach. The isoelectric-level detection and base-line correction are also implemented for accurate computation of different attributes of P, QRS, and T waves. A knowledge base is developed from different medical books and feedbacks of reputed cardiologists regarding ECG interpretation and essential time-domain features of the ECG signal. Finally, a rule-based rough-set decision system is generated for the development of an inference engine for disease identification from these time-domain features

S. Mitra

Papers

A Rough-Set-Based Inference Engine for ECG Classification

A lossless ECG data compression technique using ASCII character encoding

An ECG signal compression technique using ASCII character encoding

ECG signal compression using ASCII character encoding and transmission via SMS

Generation of digital time database from paper ECG records and Fourier transform-based analysis for disease identification.