Arrhythmia Classification Method Research Articles

A Novel Method of Clustering ECG Arrhythmia data using Robust Spatial Kernel Fuzzy C-Means

This paper presents an robust and efficient unsupervised method for ECG arrhythmia classification. The proposed method consists of two steps: feature selection and clustering. In proposed method, initially the input ECG data is fed into feature selection method to reduce the dimensionality. In this paper, three well known feature selection methods i.e., Principal Component Analysis, Linear Discriminant Analysis (LDA) and Regularized Locality Preserving Indexing (RLPI) are employed. Further, the reduced data is clustered using a Robust Spatial Kernel FCM (RSKFCM) method. RSKFCM is variant of FCM method which uses Gaussian kernel as distance metric and incorporates neighborhood information. To evaluate the proposed method the experimentation’s are conducted on UCI arrhythmia dataset and compared the results with traditional FCM and Kernel Fuzzy C-Means (KFCM). Extensive experimental results show that, the proposed RSKFCM method with RLPI feature selection method outperforms other methods in comparison.

An Arrhythmia Classification Method in Utilizing the Weighted KNN and the Fitness Rule

BackgroundArrhythmia is the disturbance in the heart's regular rhythmic activity. Although arrhythmia signals are generally reflected on electrocardiogram records, some arrhythmias do not continuously appear. In such cases, heart-rate variability signals may have to be obtained over long periods of time and should be inspected by experts. However, analyses by experts require a long time and may omit important information. Therefore, computer aided diagnosis systems are required to automatically detect diverse types of arrhythmias. In addition, although many efforts have been made to develop arrhythmia detection techniques, studies on algorithms that are reliable, robust, and have an excellent adaptability to diverse situations remain necessary. ObjectiveIn this paper, we integrate coefficients extracted through principal component analysis (PCA) and linear discriminant analysis (LDA) for reinforcing the original signal's representative features and use the weighted k-nearest neighbor (WKNN) applying the weighted value control for reducing the sensitivity depending on the K size of k-nearest neighbor (KNN), and reclassify the data that were not accurately classified using the fitness rule for improving the arrhythmia classification accuracy. MethodsIn the preprocessing phase, the amplitude is adjusted through normalization and baseline fluctuations, and diverse noises existing in signals are removed using wavelets. In the feature extraction and selection phase, using daubechies 2, the pre-processed electrocardiogram signals are decomposed, and d4 and a4 feature coefficients are extracted. Thereafter, the dimensions of the d4 and a4 feature coefficients are reduced using PCA and LDA, respectively, and the resultant coefficients from the reduction are combined and composed as features. In the classification phase, the test data are classified into four classes using the KNN. The goodness of fit of the classified data is tested using the fitness rules, and any beats that are not classified correctly are reclassified by applying them to the WKNN algorithm. ResultsThe results show that sensitivity = 97.57%, specificity = 99.42%, and positive predictive value = 94.41% are exhibited, as well as slightly higher sensitivity, specificity, and positive predictive values than other results. ConclusionThe classification combining the features of PCA and LDA shows better results than classification using the features of PCA and LDA each. Additionally, the classification using WKNN and fitness rules exhibits higher sensitivity, specificity, and positive predictive values than classification without using WKNN and fitness rules thereby classifying arrhythmia more efficiently.

A novel electrocardiogram arrhythmia classification method based on stacked sparse auto-encoders and softmax regression

Arrhythmia classification is crucial in electrocardiogram (ECG) based automatic cardiovascular disease diagnosis, e.g., to help prevent stroke or sudden cardiac death. However, the complex individual differences in ECG morphology make it challenging in accurately categorizing arrhythmia heartbeats. To promote robustness of the algorithm for individual differences, we propose a novel ECG arrhythmia classification method with stacked sparse auto-encoders (SSAEs) and a softmax regression (SF) model. The SSAEs is employed to hierarchically extract high-level features from huge amount of ECG data. Features are extracted automatically such that no individual difference in feature selection will bias extraction accuracy. Moreover, the input can be reconstructed completely by the features in each level of the auto-encoder. The SF is then trained to serve as a classifier for discriminating six different types of arrhythmia heartbeats. Computational experiments and comparative analyses are presented to validate the effectiveness of the theoretical models.

DSP-based arrhythmia classification using wavelet transform and probabilistic neural network

Abstract A large part of the biomedical research spectrum is dedicated to develop electrocardiogram (ECG) signal processing techniques to contribute to early diagnosis. However, it is common to find that ECG analysis methods reported are confined to off-line PC host operation. The authors present an arrhythmia classification method implemented on a Digital Signal Processing (DSP) platform intended for on-line, real-time ambulatory operation to classify eight heartbeat conditions: normal sinus rhythm (N), auricular fibrillation (AF), premature atrial contraction (PAC), left bundle branch block (LBBB), right bundle branch block (RBBB), premature ventricular contraction (PVC), sinoauricular heart block (SHB) and supraventricular tachycardia (SVT). The algorithm uses a wavelet transform process based on quadratic wavelets for identifying individual ECG waves and obtain a fiducial marker array. Classification is conducted by means of a Probabilistic Neural Network. The algorithm is tested with 17 ECG records obtained from the PhysioNet repository. The proposed classification procedure was tested initially on MATLAB and the results where compared with the equivalent analogue data fed to a DSP-based ECG data acquisition prototype through an arbitrary waveform generator. The results derived from confusion matrix tests yielded on-line classification accuracy of 92.69% (AF), 97.15% (N), 76.82% (PAC), 91.06% (LBBB), 87.5% (RBBB), 71.04% (PVC), 91.94% (SHB) and 95.45% (SVT), overall classification rate of 92.746% and 100% agreement between the MATLAB and on-line DSP implementations. The results suggest that the method and prototype presented may be suitable for being implemented on wearable sensing applications auxiliary for on-line, real-time diagnosis.

Heart arrhythmia classification using the prediction by partial matching algorithm

This paper describes a method for heart arrhythmia classification based on the heart rate variability HRV signal and the compression algorithm PPM. The arrhythmias to be identified are: atrial fibrillation, second heart block, normal sinus rhythm, premature ventricular contraction, ventricular fibrillation and sinus bradycardia. In the learning stage the PPM algorithm builds statistical models for the extracted tachogram. In the classification stage, the tachograms are compressed by the obtained models and attributed to the class whose models results in the best compression rate. The tests were performed with 1367 segments from the MIT-BIH Arrhythmia Database and Creighton University Ventricular Tachyarrhythmia Database. The classifier was tested for several context sizes and different training/classification sets. The performance of the classifier was measured according to sensitivity, specificity and accuracy, obtaining 96.93, 99.47 and 98.73% respectively, with the context size equal to one. These results are comparable to those of the best modern classifiers.

ECG beat classification using particle swarm optimization and support vector machine

In this paper, we propose a novel ECG arrhythmia classification method using power spectral-based features and support vector machine (SVM) classifier. The method extracts electrocardiogram’s spectral and three timing interval features. Non-parametric power spectral density (PSD) estimation methods are used to extract spectral features. The proposed approach optimizes the relevant parameters of SVM classifier through an intelligent algorithm using particle swarm optimization (PSO). These parameters are: Gaussian radial basis function (GRBF) kernel parameter σ and C penalty parameter of SVM classifier. ECG records from the MIT-BIH arrhythmia database are selected as test data. It is observed that the proposed power spectral-based hybrid particle swarm optimization-support vector machine (SVMPSO) classification method offers significantly improved performance over the SVM which has constant and manually extracted parameter.

A Feature Selection-based Ensemble Method for Arrhythmia Classification

In this paper, a novel method is proposed to build an ensemble of classifiers by using a feature selection schema. The feature selection schema identifies the best feature sets that affect the arrhythmia classification. Firstly, a number of feature subsets are extracted by applying the feature selection schema to the original dataset. Then classification models are built by using the each feature subset. Finally, we combine the classification models by adopting a voting approach to form a classification ensemble. The voting approach in our method involves both classification error rate and feature selection rate to calculate the score of the each classifier in the ensemble. In our method, the feature selection rate depends on the extracting order of the feature subsets. In the experiment, we applied our method to arrhythmia dataset and generated three top disjointed feature sets. We then built three classifiers based on the top-three feature subsets and formed the classifier ensemble by using the voting approach. Our method can improve the classification accuracy in high dimensional dataset. The performance of each classifier and the performance of their ensemble were higher than the performance of the classifier that was based on whole feature space of the dataset. The classification performance was improved and a more stable classification model could be constructed with the proposed approach.

Classification of ECG arrhythmia by a modular neural network based on Mixture of Experts and Negatively Correlated Learning

In this paper, we propose a novel ECG arrhythmia classification method using the complementary features of Mixture of Experts (ME) and Negatively Correlated Learning (NCL). Negative Correlation Learning and Mixture of Experts methods utilize different error functions for simultaneous training of negatively correlated Neural Networks. The capability of a control parameter for NCL is incorporated in the error function of ME, which enables the training algorithm of ME to establish a balance in bias-variance-covariance trade-offs. ECG records from the MIT-BIH arrhythmia database are selected as test data. It is observed that the proposed classification method not only preserves the advantages and alleviates the disadvantages of its basis approaches, but also offering significantly improved performance over the original methods.

Band-limited morphometric analysis of the intracardiac signal: implications for antitachycardia devices.

Inappropriate electrical therapy and power efficiency play a major role in algorithm implementation for antitachycardia devices (ATD) that capture, store, and analyze the patient electrogram as an adjunct to rate determination. Morphologically based algorithms have been demonstrated to improve specificity, thereby decreasing occurrences of inappropriate electrical therapy. However, morphologically based algorithms are power demanding. Optimization of power efficiency can be achieved by eliminating unnecessary algorithmic computation, but must not compromise the effectiveness of algorithms, which perform direct analysis on raw signals. Significant reductions can be achieved by reduced sampling rates, which allow for increased overall ATD efficiency via concomitant decreases in computation and data storage. This investigation determined the upper and lower bounds for filter cut-off frequency beyond which detection precision by an established morphometric method for arrhythmia classification, correlation waveform analysis (CWA), was unfavorable. Four measurement statistics were used. In ten patients with inducible VT and VF, all bipolar intraventricular electrograms were classified correctly with a minimum passband of 10-50 Hz using any of the four measurement statistics. There was > or = 80% correct classification using all four measurement statistics with passbands having low frequency cutoffs < or = 15 Hz and high frequency cutoffs > or = 50 Hz. Correct classification of > or = 90% of unipolar electrograms during NSR, VT, and VF occurred using all four measurement statistics with a passband of 1-50 Hz. There was > or = 80% correct classification with passbands 1, 10, 15, or 20-500 Hz and 10-50 Hz. The classification of NSR, VT, and VF was most accurate on an intrapatient basis. Accuracy decreased using an interpatient rhythm classification. Optimum filter settings of 1-50 Hz and 10-50 Hz were determined for unipolar and bipolar electrograms, respectively. Sampling data at 120 Hz was found to be sufficient. Bipolar electrode configuration statistically outperformed unipolar data. In conclusion, morphometric analysis of bipolar and unipolar intraventricular electrograms appears to be achievable using band limited data and reduced sampling rates.