Locating Atrial Fibrillation Rotor and Focal Sources Using Iterative Navigation of Multipole Diagnostic Catheters

Figures

• 

  Figure 1. Outline of the proposed iterative catheter guidance (ICAN) and assessments to locate AF sources using a 20-pole, circular diagnostic catheter. The algorithm starts with an arbitrary catheter placement. The electrogram signals are processed and the electrogram patterns are characterized into three source detection criteria to detect whether the current location of the catheter indicates an AF source. If so, the algorithm will stop and report the location and type of the source. Otherwise, the algorithm guides the catheter center to the next location where the process is continued until a source is detected or a maximum iteration of 25 is reached. The located source and detected type is compared against the “ground truth” source trajectory and source type to quantify the ICAN performance into four categories: successful detection (correct source location regardless of the correctness of the identified type), successful detection with correct source type, successful detection with incorrect source type, false detection (incorrect source location), and detection failure (no source was found under 25 catheter placements).

• 

  Figure 2. A series of rotor and focal sources are generated. A snapshot of the transmembrane voltage map of 12 of the simulations is shown in this figure. (a) Rotor source, CL = 140 ms. (b) Focal source, CL = 150 ms. (c) Rotor source in 3D geometry, CL = 170 ms. (d) Focal source in 3D geometry, CL = 250 ms. (e) Rotor source with a far myocardial patchy scar with an area of 2.25 cm2, CL = 250 ms. (f) Rotor anchored to a small 0.76 cm2 myocardial patchy scar, CL = 260 ms. (g) Rotor source anchored to a large 2.25 cm2 myocardial patchy scar, CL = 245 ms. (h) Focal source with a far myocardial patchy scar, CL = 150 ms. (i) Focal source close to a large 2.25 cm2 myocardial patchy scar, CL = 150 ms. (j) Functional figure-of-eight with two closely placed rotor cores, CL = 180 ms. (k) Anatomical figure-of-eight with a narrow isthmus of 1.25 mm between two 2.25 cm2-myocardial patchy scars, CL = 150 ms. (l) Stable fibrillatory cluster of rotors, CL = 170 ms). The white tracings indicate the tip of the source trajectory with longer lifespans. The red dashed lines indicate the myocardial scar region. LPVs: Left Pulmonary veins and RPVs: Right Pulmonary veins.

• 

  Figure 3. (A) a, b: Simulated 20-pole circular catheter in 2D and 3D atrial geometry. c: A sample bipolar electrograms from the 2D simulation case. The red tracing indicates a low-voltage bipole. The most negative or positive deflections are detected as the local activations (small blue circles). The gray box shows one wave-propagation cycle. The black arrow in figure a shows the PWD as determined by the direction of earliest activated bipole (Bi-10 in this example) with respect to the Lasso center. (B) a, b: Simulated 64-electrode FirMap (Abbott Electrophysiology) atrial basket catheter with diameter 40 mm (the highest resolution) in 2D and 3D atrial geometry. The gray electrodes in the 3D simulation shows the electrodes on the anterior side. c: The simulated unipolar electrograms in unipoles 1 to 10 from the 2D simulation. The simulations correspond to simulation cases A (for 2D) and C (for 3D) from Fig. 2.

• 

  Figure 4. (a) An example of catheter guidance and source localization from an initial catheter location L1 to L8 and one cycle electrograms from every corresponding catheter placement. The CDR values at L6 and L7 (0.73 and 0.78, respectively), were both greater than threshold (th = 0.6), thus meeting the CDR-based criterion. The algorithm placed the rotor between the L6 and L7 locations. Here, simulation case A from Fig. 2 is used. The electrogram voltage map as the catheter is guided is constructed and shown (color map). The voltage map is constructed by interpolating the peak-to-peak voltage of the bipolar electrograms on an interpolation grid of 0.25 mm spacing and interpolation area with a radius of 15 mm (diameter of the catheter). (b) Catheter guidance paths from two different locations are shown for four simulation cases from Fig. 2. The black arrows show the iterative guidance path. The located source is shown by a black asterisk.

• 

  Figure 5. Starting from 114,921 uniformly spaced initial catheter locations across the simulated region, ICAN guided the subsequent placements of a Lasso catheter until it detected the location and type of an AF source or a maximum catheter placement of 25 was reached. (a) Percentage of successful source detection. The successful source detections with correct and incorrect source types are shown in different colors. (b) The average distance of the detected source with the “ground truth” source trajectory. The successful and false detections are shown in two different colors. (c) The average number of catheter placements until a source was detected. None of the trials exceeded 25 catheter placements. The simulation cases A–D, J, and L are as shown in Fig. 2.

• 

  Figure 6. (a) The percentages of successful source detection using ICAN for two different CL values of approximately 150 and 240 ms are shown by the bars with two different colors. The simulation cases A, B, J, and L are as shown in Fig. 2. (b) The change in the percentage of successful source detection of ICAN with a missing electrode rate of 0 to 20% for a rotor source (simulation case A in Fig. 2) and a focal source (simulation case B in Fig. 2).

• 

  Figure 7. (a) The percentage of successful source detection with the ICAN algorithm for different simulation cases and the presence of patchy myocardial scar. (b) The localized sources with a catheter starting from a total of 114,921 uniformly spaced initial positions are shown for two simulation cases. Different colors represent located sources with correct and incorrect source types. (c) The bipole electrograms of two located sources with an incorrect source type are shown. The catheter placement P1 is a rotor that is identified as a focal source and vice versa for the catheter placement P2. Simulation cases are as shown in Fig. 2. The gray box in (c) is as described in Fig. 3A.

• 

  Figure 8. (a) The average number of detected sources using the 64-electrode basket catheter and the algorithms described in “Detecting AF Sources Using Basket Catheter” for different catheter placements (see “Catheter Simulation”). The average number of true and false positives are shown in different colors. The percentage of the true positives to the detected sources is calculated as the successful source detection rate and its average over all the catheter placements is shown in a separate bar with the right-hand side scale. (b) The average distance of the detected sources to the ground truth source for the true positive and false positive detected sources are shown in bars with different colors. Simulation cases are as shown in Fig. 2.