Dynamical measures to detect fine information controlled by digital filters in electrocardiograms

We study some dynamical measures applied to the study of electrocardiograms with two particular purposes. The first is the traditional need to determine the robustness of the measures with respect to limitations of the ECG such as a small number of data points, short sampling time and low sampling frequency. The second and more important purpose is to narrow the localization of the information captured by each measure in the particular case of ECGs. This may be an important contribution in order to be able to relate the information captured by a particular measure and aspects of the processes involved in the cardiac activity. With these purposes we have used a particular data base of electrocardiograms, ECG, constructed to study information that is not detectable by the expert visual analysis of the morphology of the ECG, nor by standard measures. This data base has been constructed by using carefully designed and controlled digital filters applied in small but critical ranges of the power spectrum of the ECG.