| 神经网络自适应共振理论应用于心电图QT间期自动检测的研究 |
| |
| 引用本文： | 陈粟蔚,宁新宝,王炜,欧阳楷,邹睿.神经网络自适应共振理论应用于心电图QT间期自动检测的研究[J].南京大学学报(自然科学版),2000,36(4). |
| |
| 作者姓名： | 陈粟蔚 宁新宝 王炜 欧阳楷 邹睿 |
| |
| 作者单位： | 南京大学电子科学与工程系!江苏南京210093(陈粟蔚，宁新宝)，南京大学物理系!江苏南京210093(王炜)，首都医科大学生物医学工程系!北京100054(欧阳楷，邹睿) |
| |
| 基金项目： | 江苏省青年人才基金的资助项目! (BQ940 0 4) |
| |
| 摘 要： | 介绍一种利用改进的ART(自适应共振理论 )人工神经网络检测心电图QT间期的方法 .用经过训练的ART网络对心电波形进行分类和识别 ,较准确的测出T波终点位置 ,从而提高QT间期检测的正确率 .该算法在一定程度上模拟了心电专家手工测量QT间期的过程 ,用于检测包括MIT心律失常数据库 (MIT/BIHArrhythmiaDatabase)在内的共近千个心跳波形 ,并以两位心电手工测量结果作为标准数据进行对比 .对比结果显示 ,我们的算法对单路心电波形自动检测的准确度与通常手工测量具有可比性 ,充分体现了人工神经网络的优越性 ,在要求自动检测大量数据的场合具有显著的应用价值 .
|
| 关 键 词： | 心电图 QT间期 人工神经网络 自适应共振理论(ART) |
THE AUTOMATIC MEASUREMENT OF QT INTERVAL IN ECG USING A NEURAL NETWORK BASED ON ADAPTIVE RESONANCE THEORY |
| |
| CHEN Su wei ,NING Xin bao ,WANG Wei ,OU Yang kai ,ZOU Rui.THE AUTOMATIC MEASUREMENT OF QT INTERVAL IN ECG USING A NEURAL NETWORK BASED ON ADAPTIVE RESONANCE THEORY[J].Journal of Nanjing University: Nat Sci Ed,2000,36(4). |
| |
| Authors: | CHEN Su wei NING Xin bao WANG Wei OU Yang kai ZOU Rui |
| |
| Institution: | CHEN Su wei 1,NING Xin bao 1,WANG Wei 2,OU Yang kai 3,ZOU Rui 3 |
| |
| Abstract: | A new algorithm for automatic QT interval measurement in ordinary and Holter electrocardiogram (ECG) is proposed. We do classification and recognition of ECG signal with a revised artificial neural network architcture, based on Adaptive Resonance Theory(ART). Our algorithm emulates manual measure in detecting T wave ends. We measured the QT interval of about 1?000 beats of ECG signals, more than half from the MIT BIH Arrhythmia Database, and compared the outcomes with that of two experts manual measure. Result shows that our new algorithm have achieved comparable accuracy with manual measure with much lower processing expenses. It outlines attractive potentialities of applying artificial neural networks to automatic analysis of cardiac signals. |
| |
| Keywords: | ECG QT interval artificial neural networks adaptive resonance theory(ART) |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
