Authors: Andreas Voss Claudia Fischer Rico Schroeder Hans R Figulla Matthias Goernig
Publish Date: 2012/06/12
Volume: 50, Issue: 7, Pages: 727-736
Abstract
The objectives of this study were to introduce a new type of heartrate variability analysis improving risk stratification in patients with idiopathic dilated cardiomyopathy DCM and to provide additional information about impaired heart beat generation in these patients Beattobeat intervals BBI of 30min ECGs recorded from 91 DCM patients and 21 healthy subjects were analyzed applying the lagged segmented Poincaré plot analysis LSPPA method LSPPA includes the Poincaré plot reconstruction with lags of 1–100 rotating the cloud of points its normalized segmentation adapted to their standard deviations and finally a frequencydependent clustering The lags were combined into eight different clusters representing specific frequency bands within 0012–1153 Hz Statistical differences between low and highrisk DCM could be found within the clusters II–VIII eg cluster IV 0033–0038 Hz p = 00002 sensitivity = 857 specificity = 714 The multivariate statistics led to a sensitivity of 929 specificity of 857 and an area under the curve of 921 discriminating these patient groups We introduced the LSPPA method to investigate time correlations in BBI time series We found that LSPPA contributes considerably to risk stratification in DCM and yields the highest discriminant power in the low and very lowfrequency bands
Keywords: