A SIMPLE SOFTWARE ROUTINE FOR THE REPRODUCIBLE PROCESSING OF THE ELECTROCARDIOGRAM

Ideally, an objective assessment of the ECG should firstly take a large number of data points in the original waveform to define a reference point (in order to give good noise rejection). Secondly, it should recognize a turning point so that the definition of a reference does not vary with changes in signal strength or noise configuration. With these aims in mind, the current article assumes a linear model for the morphology of the ECG with a linear regression analysis providing a best fit between the model and the ECG. The software programs are sufficiently general in that they have been applied to individual and enhanced electrocardiograms of both an adult and fetal origin.