基于自适应学习的心律失常心拍分类方法

Adaptive learning-based method for classification of arrhythmic heartbeats

心律失常是因心脏疾病引起的心电活动中的异常症状,早期心室收缩(PVC)是由异位心跳引起的常见心律失常形式。通过心电图(ECG)信号检测PVC对于预测可能的心力衰竭具有重要意义。本文提出一种面向PVC心拍分类的心电信号分类算法,重点研究基于自适应学习的PVC异常心拍分类特征提取模型,通过计算心拍关联后验概率,结合领域专家标注信息训练分类器,提高整体分类效果。实验采用MIT-BIH心律失常数据库的ECG数据,研究结果表明所提方法针对非线性流形结构数据,能够有效提升小样本心拍自适应分类器的准确性。

Arrhythmia is a common electrocardiogram(ECG)abnormality in heart diseases,among which premature ventricular contraction(PVC)is a widespread arrhythmia caused by ectopic heartbeat.The detection of PVC by ECG signals is significant for predicting possible heart failure.Herein an ECG signal classification algorithm for PVC heartbeat classification is proposed.The feature extraction model of PVC abnormal heartbeat classification based on adaptive learning is mainly studied.The posterior probability of heartbeat correlation is calculated and then combined with information classifier to improve the classification performance.The ECG data adopted in this study are from MIT-BIH arrhythmia database.The research results show that the proposed algorithm can effectively improve the accuracy of adaptive classifier for small-sample heart beats of nonlinear manifold data.