Poster session IV * Friday 10 December 2010, 14:00-18:00

Abstract

Background: Current modalities for intra-ventricular dyssynchrony evaluation provide global estimates of left ventricular dyssynchrony, but do not allow detailed analysis of regional differences in the timing of mechanical activation. We wished to test if advanced signal analysis on strain rate data using a cross-correlation algorithm would allow 3-dimensional mapping of mechanical activation. Methods: Time series data of 2-dimensional systolic strain rate during mechanical activation was extracted from the 3 standard views of 3 consecutive cardiac cycles using speckle tracking software. Subsequent analysis was performed on custom-build software. The data was synchronized to form 1 data set of a reconstructed 3-dimensional cycle. Strain rate data from single regions were filtered to reduce noise, and compared with an average trace using cross-correlation to provide estimates of synchrony and activation delay. These data were mapped to a bulls-eye model of the left ventricle. Results: Control individuals consistently showed only minor regional differences in mechanical activation. Patients eligible for CRT, on the other hand, had regions of severely delayed mechanical activation. Interestingly, there was a high degree of individual variation in the location of the last-activated region. 3-dimensional mapping visualized these differences in a clear and intuitive fashion (see Figure, left panel shows control individual with no dyssynchrony, right panel shows patient with severe regional dyssynchrony especially in basal segments.) Conclusion: Regional synchrony can be consistently analyzed and visualized using this new method of analysis. The early data provided here shows the potential of this method to possibly guide individualized resynchronization therapy in the future.