Evaluating Time Clustering of Nonsustained Arrythmias with the Fractal Correlation Dimension

• Activity Number: 500
• Type: Topic Contributed
• Date/Time: Wednesday, August 3, 2011 : 10:30 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Milan Seth*+
• Companies: Boston Scientific Corporation
• Keywords: Fractal ; Chaos ; Competing risks ; ICD inappropriate therapy

Abstract

The Fractal correlation dimension (FCD) has been used in a variety of disciplines to evaluate the extent to which intermittent events (for example earthquakes) occur following a chaotic process resulting in clusters of events over time. In this presentation, we describe an application of this approach to non-sustained and diverted ventricular arrythmias recorded in patients with implanted cardiac defibrilators (ICDs) and transmitting regularly to an electronic remote monitoring system. The extent of clustering as reflected by the FCD for each sampled patient (n = 1,695) was estimated using a modified Takens estimator (calculation to be described). In order to assess evidence of overall clustering, a paired non-parametric (signed rank) test was performed comparing the FCD estimates for each patient with those of matched simulated patients with an equal number of episodes randomly distributed over an equivalent follow up period. Clustering was then assessed as a predictor of the time to first inappropriate delivery of ICD shock therapy in a proportional subdistributions hazards (Fine and Gray) model adjusting for the competing risks of death and appropriate ICD therapy.