A Novel Cardiac Arrhythmia Detection Method Relying on Improved DTW Method
The high-probability ictus of cardiovascular diseases as well as the rapid proliferation of personal wearable devices points to the emergence of machine aided automatic identification and diagnosis with electrocardiograph(ECG).Most of existing algorithms of the identification of normal/abnormal ECG waveforms are computationally complicated and timeconsuming,which cannot couple with the constant augmented amount of acquired data.In contrast,dynamic time warping(DTW)is beneficial in terms of ECG identification having a high accurate rate for recognizing normal ECG waveforms but with a relatively low computational complexity.However,detecting certain abnormal ECG waveforms becomes an intractable problem by means of DTW.In this paper,a pair of improved DTW schemes are proposed,followed by a differential threshold aided algorithm designed for identifying long-term ECG signals.Moreover,real-data verification shows that our improved DTW schemes outperform the original DTW method in terms of the accurate rate of identifying abnormal ECG waveforms as well as the time consumption.
ECG identification DTW method path planning cost modification differential threshold
Wenjiang Zhang Jingjing Wang Xin Zhang Kai Zhang Yong Ren
Department of Electronic Engineering,Tsinghua University,Beijing,100084,China
国际会议
重庆
英文
862-867
2017-03-25（万方平台首次上网日期，不代表论文的发表时间）