A Patchwork Inverse Method in Combination with the Activation Time Gradient to Detect Regions of Slow Conduction in Sinus Rhythm

Oumayma Bouhamama1, Mark Potse2, Remi Dubois3, Lisl Weynans4, Laura Bear5
1Université de Bordeaux, IMB, UMR5251, 2Université de Bordeaux, 3IHU LIRYC, 4Bordeaux University, 5IHU-LIRYC


Background: Noninvasive electrocardiographic imaging (ECGI) provides real-time panoramic images of epicardial electrical activity from potential measurements on the torso surface. Several numerical methods are commonly used for ECGI, including the finite-element method (FEM), the boundary-element method (BEM) and the method of fundamental solutions (MFS). However, doubts remain about the ability of these methods to detect the presence of regions of slow conduction in structurally abnormal hearts.

Objective: The purpose of this study was to develop a method to assess the ability of ECGI to find regions of slow conduction in sinus rhythm.

Method: Slow conduction was inferred from large activation time gradients. To determine these gradients we computed the activation time difference between each node and all other nodes within 10 mm, and tested five different methods (maximum, minimum, average, nearest neighbour and random) to define a unique difference value at each node. To estimate the activation times themselves, we tested the current ECGI methods (FEM, MFS) and a novel numerical method called the Patchwork Method (PM), which locally chooses the optimal ECGI method.

Results: Two methods of computing activation time gradients (maximum and average) accurately located 6 of the 7 zones of tissue damage. Furthermore, the PM method succeeded in locating 5 of the 6 slow conduction zones recorded whereas classical ECGI methods did not locate any damaged zone.

Conclusion: In this study, the novel ECGI approach presented is an efficient tool to help overcome some of the limitations of classical numerical methods in sinus rhythm in structurally abnormal hearts, showing its ability in detecting regions of slow conduction by using the average or maximum activation time gradient.