Study On Key Techniques Of Dynamic ECG Wave Detection And Arrhythmia Classification

Dynamic ECG waveform detection and arrhythmia analysis are the key to dynamic ECG analysis system.With the development of dynamic ECG,the number of heart shot increases so significantly that the manual method has been unable to deal with a huge amount of data.In this paper,the key technology of ECG signal is studied by computer processing.And the characteristics of ECG signal are analyzed.The main efforts in this paper are as follows:Firstly,the current research condition of ECG signal technology is expounded in this paper.The characteristics of ECG waveform and the types of arrhythmia are also studied.The three major problems of ECG analysis are as follows: pretreatment problem,wave group detection problem and arrhythmia classification problem.Secondly,according to the existing ECG baseline drift,frequency interference and EMG interference of three kinds of noise,the method of wavelet analysis is used to remove the baseline drift and EMG interference.And the adaptive notch filter is used to filter out the frequency interference in this paper.Thirdly,in order to improve the accuracy of QRS detection,an adaptive threshold algorithm is used to extract the R wave crest in this paper.And it is tested in MIT-BIH arrhythmia database.Then,the difference algorithm results with other algorithms is analyzed.Finally,in order to improve the accuracy of arrhythmia classification,the method of high order statistics and principal component analysis is used to extract the features of arrhythmia in this paper.The feature vectors are extracted into the support vector machine for training and classification and realize the classification of six kinds of arrhythmia waveforms.The algorithm was validated in MIT-BIH arrhythmia database and the results are compared with the accuracy of the algorithm.