A Divergence-Based Approach for the Identification of Atrial Fibrillation Focal Drivers From Multipolar Mapping: A Computational Study

Figures

• 

  FIGURE 1. Activation maps of the propagation patterns simulated on the tissue patch cellular automaton model and position of the simulated multipolar catheter. From left to right, plane wave propagation (A), focal propagation in tissues with homogeneous (B) and heterogeneous (C) conduction properties, and wave collision (D).

• 

  FIGURE 2. Simulation of atrial fibrillation in a realistic left atrial model. In gray, anatomical model of the left atrium and positions of the simulated multipolar catheter. In color, sequential snapshots of the membrane voltage at three mapping sites: at the focal driver (A), at the boundary between the focal driver and multiple wavelets region (B), and in the multiple-wavelets region (C).

• 

  FIGURE 3. Radial basis function-reconstructed activation (top) and divergence maps (bottom) corresponding to the simulated patterns in Figure 1. Analytically determined conduction velocity vector fields are displayed on the maps as normalized arrows indicating the direction of wave propagation. Divergence maps quantify pattern properties, assigning zero divergence values (green) to planar wave propagation (A), positive divergence values (red) to focal sources (B,C), and negative divergence values (blue) to collision lines (D).

• 

  FIGURE 4. Effects of missing or corrupted data on the accuracy of localization of focal drivers by divergence analysis in a tissue patch with homogeneous (pattern b) and heterogeneous conduction properties (pattern c) in the presence of partially corrupted data. Distance between exact and estimated focal driver position as a function of the progressive removal of electrograms (left) and at increasing levels of noise in activation time series (right). The horizontal dotted line indicates the threshold for accurate localization. In the right panel, black and gray lines correspond to single-beat and 10-beat position estimation, respectively. Data are median over 100 stochastic repetitions.

• 

  FIGURE 5. Application of divergence analysis during simulated atrial fibrillation in a region activated by a focal driver. (A) Schematic representation of the multipolar mapping system, indicating the central position of the bipoles, and representative synthetic electrograms corresponding to the red bipoles of the catheter. Arrows indicate the prevalent activation sequence. (B) Beat-to-beat reconstructed activation (top) and divergence maps (bottom) corresponding to the evidenced beats. The average divergence map over the analyzed epoch is displayed in the bottom right panel. The presence of the focal driver is evidenced by maximal positive divergence values (red).

• 

  FIGURE 6. Application of divergence analysis to simulated atrial fibrillation at the boundary between the focal driver and multiple wavelets region. (A) Schematic representation of the multipolar mapping system, indicating the central position of the bipoles, and representative synthetic electrograms corresponding to the red bipoles of the catheter. Arrows indicate the prevalent activation sequence. (B) Beat-to-beat reconstructed activation (top) and divergence maps (bottom) corresponding to the evidenced beats. The average divergence map over the analyzed epoch is displayed in the bottom right panel. The presence of the focal driver is evidenced by positive divergence values (red), while collision lines are indicated by negative values (blue).

• 

  FIGURE 7. Application of divergence analysis to simulated atrial fibrillation in the region dominated by multiple wavelets propagation. (A) Schematic representation of the multipolar mapping system, indicating the central position of the bipoles, and representative synthetic AF electrograms corresponding to the red bipoles of the catheter. Arrows in the electrograms indicate changes in the activation sequence. (B) Beat-to-beat reconstructed activation (top) and divergence maps (bottom), corresponding to the evidenced beats. The average divergence map over the analyzed epoch is displayed in the bottom right panel. The absence of a stable propagation pattern is evidenced by the almost-zero values of the average divergence map (green).

• 

  FIGURE 8. Proof-of-concept application of divergence analysis to multipolar electrograms acquired during left atrial mapping in an AF patient. The displayed multipolar mapping data were collected at the left atrial appendage (A), PVs region (B), and left atrial floor (C). For each mapped region, the left panels show a schematic representation of the multipolar mapping system with arrows indicating the prevalent activation sequence and the representative electrograms corresponding to the red bipoles of the catheter. The right panels show single-beat reconstructed activation and divergence maps for the evidenced beat and the average divergence map calculated over the analyzed epoch.

• 

  FIGURE 9. Reconstruction of a rotational pattern by radial basis functions during simulated atrial fibrillation in a realistic left atrial model. In panel (A), the sequential snapshots of the membrane voltage show the presence of a transient rotor, which activates the PentaRay mapping area in a counterclockwise direction. Panel (B) shows the activation map reconstructed from RBF interpolation of the activation times recorded at the PentaRay bipoles, with the normalized conduction velocity vectors indicating the direction of wave propagation. The RBF activation map tracked the activation sequence, but it was not able to reproduce the head-meet-tail part of the circuit (B).