Abstract
BACKGROUND: Cardiac resynchronisation therapy (CRT) (left bundle branch pacing or biventricular pacing) lacks a universal consensus on ECG-based criteria to ensure effective ventricular synchronisation, particularly in the emerging field of conduction system pacing. Imageless Electrocardiographic Imaging (ECGI) is a non-invasive technology that generates cardiac maps of local activation times (LAT) without requiring a CT scan, providing a quick and safe approach to assess resynchronisation during biventricular pacing or left bundle branch pacing. PURPOSE: The aim of this study is to identify ECGI-based metrics that determine appropriate ventricular resynchronisation in patients with left bundle branch block (LBBB) during CRT procedures. METHODS: Forty-four patients (69.9 ± 11.4 years; 37 male) with a CRT indication (QRS duration 167.29 ± 22.21 ms, 42 LVEF< 40%) and twenty-eight healthy controls (53.1 ± 12.8 years 23 male) were included in a multicentric observational study. Patients undergoing CRT received either left bundle branch pacing (LBBP, n=28) or biventricular pacing (BiVP, n=16). During the CRT procedures, total activation time (TAT) and ventricular electrical uncoupling (VEU) were measured using Imageless ECGI LAT maps at baseline and final phases of the procedures. VEU was defined as the difference between the mean activation times of the left ventricle and the right ventricle. Baseline maps were considered desynchronised, while the final maps represented the synchronised state. To predict ventricular synchronisation based on TAT and VEU, a random forest algorithm with 10-fold cross-validation was used, with 82% (n=36) of the CRT patients used for training and 18% (n=8) for testing. ECGI-based metrics at the final state were compared to those of the healthy control group. RESULTS: Resynchronised ventricles were associated with VEU values below 25 ms followed by TAT values under 130 ms (see Figure 1). The random forest classifier achieved an accuracy of 97.22% in predicting ventricular synchronisation in the validation set. Accuracy in test patients was 93.75%. Examples of ECGI LAT maps from LBBB patients, both pre- and post-resynchronisation, are illustrated in Figure 2. CONCLUSION: The presented ECGI-based methodology offers a promising approach for assessment of resynchronisation, both with LBBP and BiVP, in patients with LBBB. By providing real-time visualisation of ventricular resynchronisation, ECGI could serve as a valuable tool for optimising CRT outcomes. [Figure: see text] [Figure: see text]