|
KMID : 1144120140040040388
|
|
Biomedical Engineering Letters
2014 Volume.4 No. 4 p.388 ~ p.395
|
|
|
Heartbeat classification using decision level fusion
|
|
Zhang Zhancheng
Luo Xiaoqing
|
|
|
Abstract
|
|
|
Purpose: Automatic heartbeat classification is an important technique to assist doctors to identify ectopic heartbeats in long-term ECG recording. In this paper, we employed a multi-lead fused classification schema to improve the performance of heartbeat classification.
Methods: In this paper, we introduce a multi-lead fused classification schema, in which a multi-class heartbeat classification task is decomposed into a serials of one-versus-one (OvO) support vector machine (SVM) binary classifiers, then the corresponding OvO binary classifiers of all leads are fused based on the decision score of each binary classifier, the final label is predicted by voting the fused OvO classifiers. The ECG features adopted include inter-beat and intra-beat intervals, amplitude morphology, area morphology, morphological distance and wavelet coefficients. The electrocardiograms (ECG) from the MIT-BIH arrhythmia database (MIT-BIH-AR) are used to evaluate the proposed fusion method. Following the recommendation of the Advancement of Medical Instrumentation (AAMI), all the heartbeat samples of MIT-BIH-AR are grouped into four classes, namely, normal or bundle branch block (N), supraventricular ectopic (S), ventricular ectopic (V) and fusion of ventricular and normal (F). The division of training and testing data complies with the inter-patient schema.
Results: Experimental results show that the average classification accuracy of the proposed feature selection method is 87.88%, the sensitivities for the classes N, S, V and F are 88.63%, 74.23%, 88.06% and 73.45% respectively, and the corresponding positive predictive values are 98.54%, 59.76%, 82.33% and 6.96% respectively.
Conclusions: The proposed method demonstrates better performance than the existing fusion methods.
|
|
|
KEYWORD
|
|
|
Heartbeat classification, Support vector machine, Decision fusion, Multi-lead
|
|
|
FullTexts / Linksout information
|
|
|
|
|
|
Listed journal information
|
|
|
|
|