Background: The rate of sudden cardiac death (SCD) in Brugada syndrome (BrS) is ≈1%/y. Noninvasive electrocardiographic imaging is a noninvasive mapping system that has a role in assessing BrS depolarization and repolarization abnormalities. This study aimed to analyze electrocardiographic imaging parameters during ajmaline test (AJT).Methods and Results: All consecutive epicardial maps of the right ventricle outflow tract (RVOT‐EPI) in BrS with CardioInsight were retrospectively analyzed. (1) RVOT‐EPI activation time (RVOT‐AT); (2) RVOT‐EPI recovery time, and (3) RVOT‐EPI activation‐recovery interval (RVOT‐ARI) were calculated. ∆RVOT‐AT, ∆RVOT‐EPI recovery time, and ∆RVOT‐ARI were defined as the difference in parameters before and after AJT. SCD‐BrS patients were defined as individuals presenting a history of aborted SCD. Thirty‐nine patients with BrS were retrospectively analyzed and 12 patients (30.8%) were SCD‐BrS. After AJT, an increase in both RVOT‐AT [105.9 milliseconds versus 65.8 milliseconds,P<0.001] and RVOT‐EPI recovery time [403.4 milliseconds versus 365.7 milliseconds,P<0.001] was observed. No changes occurred in RVOT‐ARI [297.5 milliseconds versus 299.9 milliseconds,P=0.7]. Before AJT no differences were observed between SCD‐BrS and non SCD‐BrS in RVOT‐AT, RVOT‐EPI recovery time, and RVOT‐ARI (P=0.9,P=0.91,P=0.86, respectively). Following AJT, SCD‐BrS patients showed higher RVOT‐AT, higher ∆RVOT‐AT, lower RVOT‐ARI, and lower ∆RVOT‐ARI (P<0.001,P<0.001,P=0.007,P=0.002, respectively). At the univariate logistic regression, predictors of SCD‐BrS were the following: RVOT‐AT after AJT (specificity: 0.74, sensitivity 1.00, area under the curve 0.92); ∆RVOT‐AT (specificity: 0.74, sensitivity 0.92, area under the curve 0.86); RVOT‐ARI after AJT (specificity 0.96, sensitivity 0.58, area under the curve 0.79), and ∆RVOT‐ARI (specificity 0.85, sensitivity 0.67, area under the curve 0.76).Conclusions: Noninvasive electrocardiographic imaging can be useful in evaluating the results of AJT in BrS.

Elevated filling pressure of the left ventricle (LV) defines diastolic dysfunction. The gold standard for diagnosis is represented by the measurement of LV end-diastolic pressure (LVEDP) during cardiac catheterization, but it has the disadvantage of being an invasive procedure. This study aimed to investigate the correlation between LVEDP and cardiac serum biomarkers such as natriuretic peptides (mid-regional pro-atrial natriuretic peptide [MR-proANP], B-type natriuretic peptide [BNP], and N-terminal prohormone BNP [NT-proBNP]), soluble ST2 (sST2), galectin-3 and mid-regional pro-adrenomedullin (MR-proAMD). Consecutive patients hospitalized in a tertiary center and undergoing left cardiac catheterization were included in the study. Diastolic dysfunction was considered present if the end-expiratory LVEDP was ≥ 15 mmHg. Cardiac biomarkers were determined from pre-procedural peripheral venous blood samples. A total of 110 patients were included, of whom 76 (69.0%) were males, with a median age of 65 (55–71) years. Median LVEDP was 13.5 (8–19) mmHg and diastolic dysfunction was present in 50 (45.4%) of the patients. LVEDP correlated with BNP (p < 0.0001, r = 0.39 [0.20–0.53]), NT-proBNP (p < 0.0001, r = 0.40 [0.22–0.55]), MR-proANP (p = 0.001, r = 0.30 [0.11–0.46]), sST2 (p < 0.0001, r = 0.47 [0.30–0.60]), but not with MR-proAMD (p = 0.77) or galectin-3 (p = 0.76). In the final stepwise multivariable binary logistic regression model, diastolic dysfunction was predicted by NT-proBNP, mitral average E/e’, sST2, atrial fibrillation, and left atrium reservoir strain. BNP, NT-proBNP, MR-proANP, and sST2 had predictive value for diastolic dysfunction. In contrast, galectin-3 and MR-proAMD were not associated with increased filling pressures. Furthermore, NT-proBNP and sST2 significantly improved diastolic dysfunction prediction in the final multivariable model.

Background: Brugada syndrome (BrS) has been associated with sudden cardiac death in otherwise healthy subjects, and drug‐induced BrS accounts for 55% to 70% of all patients with BrS. This study aims to develop a deep convolutional neural network and evaluate its performance in recognizing and predicting BrS diagnosis. Methods and Results: Consecutive patients who underwent ajmaline testing for BrS following a standardized protocol were included. ECG tracings from baseline and during ajmaline were transformed using wavelet analysis and a deep convolutional neural network was separately trained to (1) recognize and (2) predict BrS type I pattern. The resultant networks are referred to as BrS‐Net. A total of 1188 patients were included, of which 361 (30.3%) patients developed BrS type I pattern during ajmaline infusion. When trained and evaluated on ECG tracings during ajmaline, BrS‐Net recognized a BrS type I pattern with an AUC‐ROC of 0.945 (0.921–0.969) and an AUC‐PR of 0.892 (0.815–0.939). When trained and evaluated on ECG tracings at baseline, BrS‐Net predicted a BrS type I pattern during ajmaline with an AUC‐ROC of 0.805 (0.845–0.736) and an AUC‐PR of 0.605 (0.460–0.664). Conclusions: BrS‐Net, a deep convolutional neural network, can identify BrS type I pattern with high performance. BrS‐Net can predict from baseline ECG the development of a BrS type I pattern after ajmaline with good performance in an unselected population.