Geometric measures of Poincare plots for the detection of small sympathovagal shifts.

The Poincare plot has been a nonlinear tool for the analysis of heart rate variability. However, due to a lack of tools to quantitatively characterize the plots, the Poincare plot has been used visually only. One of the problems with a Poincare plot of a large number of R-R intervals has been the outliers. These make the plots much larger than they should be while contributing relatively little from a density point of view. By concentrating on the central R-R pairs the outliers can be ignored. A contour that includes 90% of the R-R pairs is used. The mean R-R values, the histogram and the 90% contour are used to characterize the Poincare plots. Holter recordings from 4 hours of night sleep during normal and mild hyperglycaemic episodes will be compared using the new set of tools.