Consistency of the automatic algorithm in detecting complex fractionated electrograms using an electroanatomical navigation system

Background: The different settings of the automatic algorithm in the Carto system (Carto XP, Biosense Webster, Diamond Bar, CA, USA) used for detecting complex fractionated electrograms (CFEs) during atrial fibrillation (AF) may influence the identification of the fragmented electrograms. Objectives: We aimed to evaluate the impact of the different parameters on the detection of CFEs and the efficacy of the substrate modification after pulmonary vein isolation (PVI). Methods: A total of 1,159 electrograms were analyzed from 11 consecutive patients (age = 56 ± 12 years). The effect of the different algorithm factors, such as the high-voltage thresholds (0.12, 0.25, 0.5, 20 mV), detection algorithms (average complex interval [ACI] vs interval confidence level), and recording duration (2.5 seconds vs 5 seconds), on the disparities of the CFEs was investigated. Results: The proportion of the different grades of CFEs depended on the detection algorithm and recording duration. The high-voltage threshold would not affect the consistency of the CFEs irrespective of the different settings of the detection algorithm or recording duration. High-grade CFEs were most consistent with an ACI algorithm and recording duration of 5 seconds (Cronbach's alpha = 0.952). Ablation consisting of a PVI and high-grade CFE sites converted AF directly to sinus rhythm in eight of 11 patients or into atrial tachycardia in one of 11. Conclusions: The distribution and consistency of the CFE detection depended on the detection algorithm and recording duration, but not on the high-voltage threshold. Under the ACI algorithm and a recording duration of 5 seconds, high-grade CFE sites remained highest consistency. (PACE 2012; 35:980-989)