Statistical Methods For Analyzing ECG

Many methods have been used in the field of electrocardiogram(ECG) in modern medicine, In fact, this is a sampling procedure, ie, takes samples from population. But so far, few people have studied the shape properties using statistical methods sophisticatedly.First, we analyzed the ECG using principal component analysis, and found common features of the variations of the ECGs. We found that there are similarity and characters. Then we introduced the functional principal component analysis and extended the princi-pal component analysis, which received more and more attention from many researchers, and we used it on ECG for the first time. We found that the variation of the different waves are similar, ie, vertical and horizon shifts.Finally, we modified an existing nonlinear model of functional data which have been used on T wave, and extended it to P wave and QRS complex wave to characterize the shape of ECGs. It is much easier to reduce the dimension of the data using functional analytical which greatly improve the currently using of QT method.