Research Of The Finger-based ECG Pre-Processing Algorithm

Electrocardiogram(ECG)can reflect the hearth work state and reliable internal feature information from different levels,which has been widely used in the diagnosis of the clinical cardiovascular disease.Accompanying with the emergence of ECG chips with the features of small size,lower power and easy integration recently,acquisition of the ECG signal from fingers can be realized,which appear to broaden the scope of application of the ECG.The acquisition process of the finger-based ECG generally subject to all kinds of noise,and the presence of noise will affect the further processing and analysis to the ECG signal.In the ECG Signal Analysis System,the accuracy of the R wave detection will directly affect the location of other feature waves.Before analysing the finger-based ECQto eliminate the noise in the signal without distortion and detect the R wave accurately should be given primary importance among priorities.The thesis mainly studies the finger-based ECG de-noising and R wave detection algorithm,the main ideas of this thesis are mainly as follows:1.The generation mechanism and morphological characteristics of ECG signal are studied,then the common types of the noise in ECG signal are analyzed followed by the introduction about the source and acquisition device of the finger-based ECG2.In order to weaken the integrated noise,baseline drift and frequency interference in the finger-based ECG,a series of de-noising methods based on Ensemble Empirical Mode Decomposition(EEMD)are proposed.Firstly,a de-noising algorithm is proposed based on the combination of EEMD decomposition and Mahalanobis distance for comprehensive noise;Secondly,a de-noising algorithm based on the combination of EEMD decomposition and adaptive filtering for the baseline drift is put forward;Finally,a de-noising algorithm based on the EEMD decomposition and adaptive filtering for the presence of the signal frequency interference.The feasibility and validity of the de-noising algorithm are verified by the simulated ECG signal and the real finger-based ECG in the Finger ECG Data.3.A new R wave detection algorithm of the finger-based ECG based on the EEMD decomposition and Wavelet Transform is proposed.Firstly.the high order IMFs of the EEMD decomposition of the finger-based ECG are added to get a new reconstruction signal;Secondly,on the basis of the quadratic spline wavelet,the four scales wavelet transform is applied to the new signal.Finally,the adaptive threshold decision is used to find the modulus maximum of the high frequency component in the fourth scale,which can be of great help to detect the location of R wave.The performance of the proposed R wave detection algorithm is verified by the ECG data come from the Finger ECG Data,Surface ECG Data and the MIT-BIH Arrhythmia Database.The experimental results show that algorithm has strong robustness and accuracy.This thesis finishes the pre-processing of the finger-based ECG,providing a theoretical basis and technical support for the further application for the finger-based ECG.