Extraction and Quantification Clusters of Three-Dimensional Lorenz Plots.

Lorenz plot (LP) method gives a global view of long-time electrocardiogram signals, is an efficient simple visualization tool to analyze heart rate variability and cardiac arrhythmias. 50 records of RR-interval time series with frequent premature complexes exported from 24-hour Holter. Constructed three-dimensional LP (3DLP) with three successive RR intervals as X, Y and Z axis in Cartesian coordinate system for each record, then stereographic projection it along the space diagonal. The radii of dots distinguished eccentric clusters (ECs) points from centric clusters' with the accuracy of 94 ± 6.0%. The eccentric scatter-dots were separated and identified by the frequency distribution characteristic on azimuth, with the accuracy of 93 ± 13.3% ECs. The APF < -2.8° of CPN EC supported the ventricular extrasystoles diagnosis, with excellent sensitivity (1.0) and specificity (0.92). The transformed coordinates (polar radius and angle) of 3DLP could extract and quantify clusters to diagnose arrhythmia, and might provide additional prognosis information.