Optimization of a 12-Lead Electrocardiography Subset for Automated Early Left Ventricular Activation Localization Approach Based on Pace-Mapping Technology

We previously developed an automated approach based on pace mapping to localise early left ventricular (LV) activation origin. To avoid a singular system, we require pacing from at least 2 more known sites than the number of electrocardiography (ECG) leads used. Fewer leads used means fewer pacing sites required. We sought to identify an optimal minimal ECG lead set for the automated approach. We used 1715 LV endocardial pacing sites to create derivation and testing data sets. The derivation data set, consisting of 1012 known pacing sites pooled from 38 patients, was used to identify an optimal 3-lead set by means of random forest regression (RFR), and a second 3-lead set by means of exhaustive search. The performance of these sets and the calculated Frank leads was compared within the testing data set with 703 pacing sites pooled from 25 patients. The RFR yielded III, V1, and V4, whereas the exhaustive search identified leads II, V2 and V6. Comparison of these sets and the calculated Frank leads demonstrated similar performance when using 5 or more known pacing sites. Accuracy improved with additional pacing sites, achieving mean accuracy of < 5 mm, after including up to 9 pacing sites when they were focused on a suspected area of ventricular activation origin (radius < 10 mm). The RFR identified the quasi-orthogonal leads set to localise the source of LV activation, minimizing the training set of pacing sites. Localization accuracy was high with the use of these leads and was not significantly different from using leads identified by exhaustive search or empiric use of Frank leads. Nous avons développé une approche automatisée fondée sur la cartographie cardiaque pour repérer de manière précoce l’origine de l’activation du ventricule gauche (VG). Pour éviter un système unique, il faut cartographier au moins deux autres sites connus, outre les sites des dérivations d’électrocardiographie (ECG) utilisés. Un faible nombre de dérivations signifie que moins de sites de cartographie sont nécessaires. Nous avons cherché à déterminer un nombre minimal optimal de dérivations d’ECG pour l’approche automatisée. Nous avons utilisé 1 715 sites de cartographie endocardique du VG pour créer des ensembles de données de dérivation et de test. L’ensemble de données pour la dérivation, qui contient 1 012 sites de cartographie connus collectés auprès de 38 patients, a été utilisé pour cibler un ensemble optimal de trois dérivations par régression par forêt aléatoire, et un