This study aims to identify and visualize electrocardiogram (ECG) features using an explainable deep learning–based algorithm to predict cardiac resynchronization therapy (CRT) outcome. Its performance is compared with current guideline ECG criteria and QRS_AREA. A deep learning algorithm, trained on 1.1 million ECGs from 251 473 patients, was used to compress the median beat ECG, thereby summarizing most ECG features into only 21 explainable factors (FactorECG). Pre-implantation ECGs of 1306 CRT patients from three academic centres were converted into their respective FactorECG. FactorECG predicted the combined clinical endpoint of death, left ventricular assist device, or heart transplantation [c-statistic 0.69, 95% confidence interval (CI) 0.66–0.72], significantly outperforming QRS_AREA and guideline ECG criteria [c-statistic 0.61 (95% CI 0.58–0.64) and 0.57 (95% CI 0.54–0.60), P < 0.001 for both]. The addition of 13 clinical variables was of limited added value for the FactorECG model when compared with QRS_AREA (Δ c-statistic 0.03 vs. 0.10). FactorECG identified inferolateral T-wave inversion, smaller right precordial S- and T-wave amplitude, ventricular rate, and increased PR interval and P-wave duration to be important predictors for poor outcome. An online visualization tool was created to provide interactive visualizations (https://crt.ecgx.ai). Requiring only a standard 12-lead ECG, FactorECG held superior discriminative ability for the prediction of clinical outcome when compared with guideline criteria and QRS_AREA, without requiring additional clinical variables. End-to-end automated visualization of ECG features allows for an explainable algorithm, which may facilitate rapid uptake of this personalized decision-making tool in CRT. First, an artificial intelligence algorithm (variational auto-encoder) was pretrained on 1.1 million electrocardiograms (ECGs) to learn the underlying continuous factors that generate the ECG (i.e. the FactorECG). In this process, the variational auto-encoder (VAE) learns to reconstruct ECGs as accurately as possible using only 21 continuous factors without any human input. In the training phase, the pre-procedural median beat ECGs of 1306 cardiac resynchronization therapy (CRT) patients were each converted into their FactorECG. These 21 factors were subsequently used as input in a Cox model to predict the primary composite endpoint of left ventricular assist device implantation, heart transplantation and all-cause death, and the secondary endpoint of echocardiographic response. FactorECG significantly improved outcome prediction following CRT when compared with the current guidelines and QRS_AREA. The algorithm is explainable by using the decoder to visualize the effect of the ECG factors that significantly predicted outcome on the median beat ECG morphology. Here, for example, the influence of Factor 9 (F₉) is visualized, where higher values represent a more left bundle branch block-like ECG morphology and lower values, a more right bundle branch block-like morphology. CRT, cardiac resynchronization therapy; DNN, deep neural network; ECG, electrocardiogram; HTx, heart transplantation; LVAD, left ventricular assist device.

Pacing remote from the latest electrically activated site (LEAS) in the left ventricle (LV) may diminish response to cardiac resynchronization therapy (CRT). We tested whether proximity of LV pacing site (LVPS) to LEAS, determined by non-invasive three-dimensional electrical activation mapping [electrocardiographic Imaging (ECGI)], increased likelihood of CRT response. Consecutive CRT patients underwent ECGI and chest/heart computed tomography 6–24 months of post-implant. Latest electrically activated site and the distance to LVPS (d_p) were assessed. Left ventricular end-systolic volume (LVESV) reduction of ≥15% at clinical follow-up defined response. Logistic regression probabilistically modelled non-response; variables included demographics, heart failure classification, left bundle branch block (LBBB), ischaemic heart disease (IHD), atrial fibrillation, QRS duration, baseline ejection fraction (EF) and LVESV, comorbidities, use of CRT optimization algorithm, angiotensin-converting enzyme inhibitor(ACE)/angiotensin-receptor blocker (ARB), beta-blocker, diuretics, and d_p. Of 111 studied patients [64 ± 11 years, EF 28 ± 6%, implant duration 12 ± 5 months (mean ± SD), 98% had LBBB, 38% IHD], 67% responded at 10 ± 3 months post CRT-implant. Latest electrically activated sites were outside the mid-to-basal lateral segments in 35% of the patients. d_p was 42 ± 23 mm [31 ± 14 mm for responders vs. 63 ± 24 mm non-responders (P < 0.001)]. Longer d_p and the lack of use of CRT optimization algorithm were the only independent predictors of non-response [area under the curve (AUC) 0.906]. d_p of 47 mm delineated responders and non-responders (AUC 0.931). The distance between LV pacing site and latest electrical activation is a strong independent predictor for CRT response. Non-invasive electrical evaluation to characterize intrinsic activation and guide LV lead deployment may improve CRT efficacy.

Introduction: Recent studies have shown that the baseline QRS area is associated with the clinical response after cardiac resynchronization therapy (CRT). In this study, we investigated the association of QRS area reduction (∆QRS area) after CRT with the outcome. We hypothesize that a larger ∆QRS area is associated with a better survival and echocardiographic response.Methods and Results: Electrocardiograms (ECG) obtained before and 2–12 months after CRT from 1299 patients in a multi‐center CRT‐registry were analyzed. The QRS area was calculated from vectorcardiograms that were synthesized from 12‐lead ECGs. The primary endpoint was a combination of all‐cause mortality, heart transplantation, and left ventricular (LV) assist device implantation. The secondary endpoint was the echocardiographic response, defined as LV end‐systolic volume reduction ≥ of 15%. Patients with ∆QRS area above the optimal cut‐off value (62 µVs) had a lower risk of reaching the primary endpoint (hazard ratio: 0.43; confidence interval [CI] 0.33–0.56, p < .001), and a higher chance of echocardiographic response (odds ratio [OR] 3.3;CI 2.4–4.6, p < .0001). In multivariable analysis, ∆QRS area was independently associated with both endpoints. In patients with baseline QRS area ≥109 µVs, survival, and echocardiographic response were better when the ∆QRS area was ≥62 µVs (p < .0001). Logistic regression showed that in patients with baseline QRS area ≥109 µVs, ∆QRS area was the only significant predictor of survival (OR: 0.981; CI: 0.967–0.994, p = .006).Conclusion: ∆QRS area is an independent determinant of CRT response, especially in patients with a large baseline QRS area. Failure to achieve a large QRS area reduction with CRT is associated with a poor clinical outcome.

Introduction: We aimed to investigate the impact of the 2021 European Society of Cardiology (ESC) guideline changes in left bundle branch block (LBBB) definition on cardiac resynchronization therapy (CRT) patient selection and outcomes.Methods: The MUG (Maastricht, Utrecht, Groningen) registry, consisting of consecutive patients implanted with a CRT device between 2001 and 2015 was studied. For this study, patients with baseline sinus rhythm and QRS duration ≥ 130ms were eligible. Patients were classified according to ESC 2013 and 2021 guideline LBBB definitions and QRS duration. Endpoints were heart transplantation, LVAD implantation or mortality (HTx/LVAD/mortality) and echocardiographic response (LVESV reduction ≥15%).Results: The analyses included 1.202, typical CRT patients. The ESC 2021 definition resulted in considerably less LBBB diagnoses compared to the 2013 definition (31.6% vs. 80.9%, respectively). Applying the 2013 definition resulted in significant separation of the Kaplan–Meier curves of HTx/LVAD/mortality (p < .0001). A significantly higher echocardiographic response rate was found in the LBBB compared to the non‐LBBB group using the 2013 definition. These differences in HTx/LVAD/mortality and echocardiographic response were not found when applying the 2021 definition.Conclusion: The ESC 2021 LBBB definition leads to a considerably lower percentage of patients with baseline LBBB then the ESC 2013 definition. This does not lead to better differentiation of CRT responders, nor does this lead to a stronger association with clinical outcomes after CRT. In fact, stratification according to the 2021 definition is not associated with a difference in clinical or echocardiographic outcome, implying that the guideline changes may negatively influence CRT implantation practice with a weakened recommendation in patients that will benefit from CRT.

Background: Reduction in QRS area after cardiac resynchronization therapy (CRT) is associated with improved long-term clinical outcome. The aim of this study was to investigate whether the reduction in QRS area is associated with hemodynamic improvement by pacing different LV sites and can be used to guide LV lead placement. Methods: Patients with a class Ia/IIa CRT indication were prospectively included from three hospitals. Acute hemodynamic response was assessed as the relative change in maximum rate of rise of left ventricular (LV) pressure (%∆LVdP/dt_max). Change in QRS area (∆QRS area), in QRS duration (∆QRS duration), and %∆LVdP/dt_max were studied in relation to different LV pacing locations within a patient. Results: Data from 52 patients paced at 188 different LV pacing sites were investigated. Lateral LV pacing resulted in a larger %∆LVdP/dt_max than anterior or posterior pacing (p = 0.0007). A similar trend was found for ∆QRS area (p = 0.001) but not for ∆QRS duration (p = 0.23). Pacing from the proximal electrode pair resulted in a larger %∆LVdP/dt_max (p = 0.004), and ∆QRS area (p = 0.003) but not ∆QRS duration (p = 0.77). Within patients, correlation between ∆QRS area and %∆LVdP/dt_max was 0.76 (median, IQR 0.35; 0,89). Conclusion: Within patients, ∆QRS area is associated with %∆LVdP/dt_max at different LV pacing locations. Therefore, QRS area, which is an easily, noninvasively obtainable and objective parameter, may be useful to guide LV lead placement in CRT.

Cardiac resynchronization therapy (CRT) is a good treatment for heart failure accompanied by ventricular conduction abnormalities. Current ECG criteria in international guidelines seem to be suboptimal to select heart failure patients for CRT. The criteria QRS duration and left bundle branch block (LBBB) QRS morphology insufficiently detect left ventricular activation delay, which is required for benefit from CRT. Additionally, there are various definitions for LBBB, in which each one has a different association with CRT benefit and is prone to subjective interpretation. Recent studies have shown that the objectively measured vectorcardiographic QRS area identifies left ventricular activation delay with higher accuracy than any of the current ECG criteria. Indeed, various studies have consistently shown that a high QRS area prior to CRT predicts both echocardiographic and clinical improvement after CRT. The beneficial relation of QRS area with CRT-outcome was largely independent from QRS morphology, QRS duration, and patient characteristics known to affect CRT-outcome including ischemic etiology and sex. On top of QRS area prior to CRT, the reduction in QRS area after CRT further improves benefit. QRS area is easily obtainable from a standard 12-lead ECG though it currently requires off-line analysis. Clinical applicability will be significantly improved when QRS area is automatically determined by ECG equipment.