Heartbeat classification system based on neural networks and dimensionality reduction

Abstract Introduction This paper presents a complete approach for the automatic classification of heartbeats to assist experts in the diagnosis of typical arrhythmias, such as right bundle branch block, left bundle branch block, premature ventricular beats, premature atrial beats and paced beats. Methods A pre-processing step was performed on the electrocardiograms (ECG) for baseline removal. Next, a QRS complex detection algorithm was implemented to detect the heartbeats, which contain the primary information that is employed in the classification approach. Next, ECG segmentation was performed, by which a set of features based on the RR interval and the beat waveform morphology were extracted from the ECG signal. The size of the feature vector was reduced by principal component analysis. Finally, the reduced feature vector was employed as the input to an artificial neural network. Results Our approach was tested on the Massachusetts Institute of Technology arrhythmia database. The classification performance on a test set of 18 ECG records of 30 min each achieved an accuracy of 96.97%, a sensitivity of 95.05%, a specificity of 90.88%, a positive predictive value of 95.11%, and a negative predictive value of 92.7%. Conclusion The proposed approach achieved high accuracy for classifying ECG heartbeats and could be used to assist cardiologists in telecardiology services. The main contribution of our classification strategy is in the feature selection step, which reduced classification complexity without major changes in the performance.

ECG-based detection of left ventricle hypertrophy

Introduction Left ventricle hypertrophy (LVH) is an important risk factor for cardiovascular morbidity and mortality. It is characterized by a thickening of the walls of the left ventricle. The transthoracic echocardiogram is a very accurate method for LVH detection. However, the electrocardiogram (ECG) offers an alternative method in diagnosing LVH, besides being less expensive and easier to obtain. In this context, this study proposes an ECG based approach for left ventricle hypertrophy (LVH) classification. Methods According to the literature, several indexes have so far been proposed that suggest specific changes in cardiac structure, however, generally speaking there is no consensus about the best criteria. This way, instead of considering only one LVH criterion, a score derived from electrocardiographic traces was employed which explores the complementarity of the best criteria through a fusion strategy. The best criteria are those which discriminate normal and LVH ECGs. Results The experiments were performed in the Monica database with a group of fifty men. Half of the individuals had LVH diagnosed by calculating the left ventricular mass index measured by transthoracic echocardiography. The score fusion proposed achieved a sensitivity of 78.3% and specificity of 91.3%, outperforming all isolated LVH criteria. Discussion Unlike the other methods, our score must be estimated within a computer because of its high complexity. Even with this limitation it is much less expensive than using the echocardiography.