Abstract

Heart diseases are amongst the most common causes of death in the industrialized world. Since the cardiological system is very complex and hard to capture in its entirety, researchers are looking for indicators of its health. A promising one is the heart rate variability (HRV), i.e., the variation of the time intervals between two heartbeats. It reflects physiological processes, which influence the rhythm of the heart. An approach by researchers is a visualization tool, the Poincaré plot, to analyse HRV. Numerous data models exist in order to automatically quantify Poincaré plots. To extract as much information as possible from Poinca-ré plots, it has to be filtered from artefacts and outliers before applying the data models. The goal of this work is to test the influence of two dif-ferent filtering methods on the Poincaré plot quantifica-tion methods.
A test case was constructed were a database with healthy heart rates and one with pathological heart rates were filtered with the two methods. Thereafter two Poincaré plot measures were evaluated using the filtered data sets. Afterwards the differences between these data sets were statistically examined. It can be concluded that the fully automated filtering via clustering shows no large drawbacks compared to the traditional method of ECG annotation based filtering for HRV-analysis via Poincaré plots.