ANALYSIS OF ECG SIGNAL DETECTION FOR CARDIAC ARRHYTHMIAS

Abstract

Cardiac iarrhythmia is one iof the imost common diseases suffering around ithe world. It is a disorder of the heart in which beats are too slow or too fast. To analyze the characteristics of the heart, iElectrocardiogram i(ECG) is used. ECG is used to irecord the electrical iactivity iof the heart. However, a normal ECG signal has many noises such as baseline iwander, ipowerline interference, andi motion iartifacts. There arei lots of arrhythmia detection techniques such as wavelet transforms, ifiltering techniques, iartificial ineural networks, genetic algorithms, syntactic methods, iHilbert transform, iMarkov models, etc. In ithis paper, iDiscrete iWavelet iTransform is iused to denoise these noises to detecti the QRS icomplex and R peak. Mutual information (MI) is used to classify the arrhythmias. MATLAB software is used to implement for MIT/BIH Arrhythmia database.