摘要
目的提出Meyer多分辨率逼近的新方法,消除心电信号基线漂移。方法带基线漂移的心电信号减去其在V9上的正交投影就基本消除了基线漂移。通过多分辨率逼近的数值分析,讨论了小波基的选择,正交投影的选择和数据长度的选择。对目前此类算法因缺少定量化分析方法而难以进行效果的比较,提出算法精度的定量化分析方法,构建算法效果比较的共同基础。结果应用算法于MIT-BIH数据库中具有严重基线漂移的心电数据,采用Meyer小波,带基线漂移的心电信号在V9上的正交投影和基线漂移的差异为(0.3±3.6)μV。结论应用Meyer小波在层次9上消除30 s左右ECG的基线漂移是合适的。
Objective To propose a new method of Meyer multi-resolution approximation for eliminating ECG baseline drift.Methods By subtracting the orthogonal projection on V9 from ECG with baseline drift,the baseline drift was removed.Under the numerical analysis of multi-resolution approximation,the selection of wavelets,the orthogonal projection and the data length were discussed.At present,it was difficult to compare the eliminating results of baseline drift by different methods.The quantitative analysis was introduced.Common basis of results comparison was set up.Results Applying the method to the serious baseline drift ECG data from MIT-BIH database,the difference between the orthogonal projection of ECG with baseline drift on V9 and baseline drift was（0.3±3.6）μV.Conclusion About 30 s of ECG baseline drift can be removed using Meyer wavelet at level 9.
出处
《航天医学与医学工程》
CAS
CSCD
北大核心
2012年第1期46-49,共4页
Space Medicine & Medical Engineering
基金
国家自然科学基金(60872057)
浙江省自然科学基金(R1090244
Y1101237
Y1110944)
浙江省公益技术应用研究项目(2011C23132)
关键词
心电信号
多分辨率逼近
基线漂移
定量化分析
electrocardiogram
multi-resolution approximation
baseline drift
quantitative analysis
