INVESTIGATION OF CORRELATION DIMENSION ESTIMATION IN HEARTBEAT TIME SERIES

Abstract

The Correlation Dimension estimation is an especially sensitive method which allows for important information to be extracted from the signal under investigation. However, due to its sensitivity, its parameters as well as details of the application process must be chosen carefully and customized to each specific problem. In this paper, we examine the application of the Correlation Dimension estimation in the case of heart rate variability signals. Specifically, we investigate issues which are not discussed in other similar studies or which are chosen differently, such as the use of the Euclidean norm and the Theiler window. We also document with detail the method of data acquisition and make a case for the importance of proper recordings the lack of which leads to controversial results.