Novel Algorithmic Methods in Mapping of Atrial and Ventricular TachycardiaCLINICAL PERSPECTIVE
Background—Conventional methods to assess local activation time (LAT) detect the peak of the bipolar electrogram (B-LATPeak) or the maximal negative slope of the unipolar electrogram (U-LATSlope). We evaluated 3 novel methods to assess LAT: onset (B-LATOnset) and center of mass (B-LATCoM) of bipolar electrogram, and maximal negative slope of unipolar electrogram within a predefined bipolar window (U-LATSlope-hybrid).
Methods and Results—In 1753 atrial tachycardia and 1426 ventricular tachycardia recordings, the performance of the methods in detecting LAT was evaluated pair-wise (eg, B-LATPeak versus B-LATOnset). For each comparison, histogram analysis of the differences in LAT values was performed. Variation in differences (P95-P5) in low quality (LQ) was compared with high-quality electrograms. In a separate data set (12 atrial tachycardia and 10 ventricular tachycardia), we evaluated for each method the accuracy in algorithmic activation mapping. Both in atrial tachycardia and ventricular tachycardia, the variation in difference between the conventional and novel methods was larger in LQ electrograms. In contrast, variation in difference between the novel methods was comparable in LQ and high-quality electrograms. Except for LATSlope-hybrid, all methods showed decreased mapping accuracy with increasing percentage of LQ electrograms. U-LATSlope-hybrid accurately mapped activation in 16 out of 22 maps (versus B-LATCoM, 14; B-LATPeak, 14; B-LATOnset, 13; U-LATSlope, 4).
Conclusions—In LQ atrial and ventricular electrograms, the novel LAT methods (B-LATOnset, B-LATCoM, and U-LATSlope-hybrid) show less variation than the conventional methods. The U-LATSlope-hybrid, a hybrid method that accurately detects the maximal negative unipolar slope, is associated with the highest accuracy in algorithmic mapping of atrial tachycardia/ventricular tachycardia.
- Received July 10, 2013.
- Accepted April 6, 2014.
- © 2014 American Heart Association, Inc.