Evaluating omnipolar configuration for complex fractionated atrial electrograms detection: a study in animal models

Atrial fibrillation (AF) is a global health concern, often treated with the ablation of certain tissue areas. The emergence of complex fractionated atrial electrograms (CFAEs) as a biomarker, associated with factors such as fibrosis, anisotropy or heterogeneous tissue activation, initially demonstrated high efficacy [1]. However, subsequent findings have shown variable results.

This study aims to revalidate the criterion for detecting regions with CFAEs using omnipolar electrograms (oEGMs) and to compare this configuration with the traditional bipolar approach, which, due to its sensitivity to the direction of propagation, may yield low-amplitude and prolonged pulses that can be misinterpreted as CFAEs.

The data was collected from the left atrium of a female swine using a HD-Grid catheter (Figure 1). To assess the false-positive rate for both configurations, pre-ablated tissue was used. For this reason, the entire dataset of 400 recordings was filtered based on peak-to-peak amplitude (> 1.5 mV) and vector field heterogeneity [2] (VFH < 0.4). Thus, characterization was performed solely on healthy tissue.

A total of 2666 bEGMs and their respective 1333 oEGMs were analyzed. For oEGMs acquisition, the bEGM cross approach was used [3]. A cross-correlation algorithm was then employed to classify each complex into distinct groups. This algorithm groups pulses based on a maximum cross-correlation greater than 0.95. Following the criteria for defining fractionation [1], each group was classified into one of two categories: normal or CFAE.

Table 1 summarizes the results of the cross-correlation algorithm results obtained. The percentage of false positives for CFAEs identified in bEGMs (13.15%) is significantly higher than in oEGMs (3.32%), whose principle of amplitude maximization in the direction of propagation suppresses fractionation succesfully. Furthermore, the amplitude and duration ranges of the bEGMs confirms the initial hypothesis that low-amplitude and prolonged bEGMs are prone to misclassification as CFAEs.

Examples of pulses from each high cross-correlation group illustrate accurate classification. Note that CFAE pulses exhibit a greater number of baseline-deflecting peaks and display a more complex morphology.

This study reports the limitations of bipolar configuration in identifying CFAE, particularly due to its sensitivity to propagation direction when classifying entirely healthy pulses. The omnipolar approach is proposed as it significantly enhances CFAE identification by reducing false positives fourfold, and providing a robust technique for assessing cardiac substrate in detecting ablation-suitable regions in AF patients.Figure 1:

HD-Grid intracavitary catheter