A Deep Learning-enabled Electrocardiogram Model for the Identification of a Rare Inherited Arrhythmia: Brugada Syndrome

2021 ◽  
Author(s):  
Chih-Min Liu ◽  
Chien-Liang Liu ◽  
Kai-Wen Hu ◽  
Vincent S. Tseng ◽  
Shih-Lin Chang ◽  
...  
Keyword(s):  
Deep Learning ◽  
Brugada Syndrome ◽  
Inherited Arrhythmia

A Deep Learning-enabled Electrocardiogram Model for the Identification of a Rare Inherited Arrhythmia: Brugada Syndrome (Preprint)

2020 ◽  
Author(s):  
Chih-Min Liu ◽  
Chien-Liang Liu ◽  
Kai-Wen Hu ◽  
Vincent S. Tseng ◽  
Shih-Lin Chang ◽  
...  
Keyword(s):  
Deep Learning ◽  
Brugada Syndrome ◽  
Area Under The Curve ◽  
Early Time ◽  
Learning Model ◽  
Kappa Coefficient ◽  
Mcnemar Test ◽  
Inherited Arrhythmia ◽  
Deep Learning Model

BACKGROUND Brugada syndrome is a rare inherited arrhythmia with a unique electrocardiogram (ECG) pattern (type 1 Brugada ECG pattern), which is a major cause of sudden cardiac death in young people. Automatic screening for the ECG pattern of Brugada syndrome by a deep learning model gives us the chance to identify these patients at an early time, thus allowing them to receive life-saving therapy. OBJECTIVE To develop a deep learning-enabled ECG model for diagnosing Brugada syndrome. METHODS A total of 276 ECGs with a type 1 Brugada ECG pattern (276 type 1 Brugada ECGs and another randomly retrieved 276 non-Brugada type ECGs for one to one allocation) were extracted from the hospital-based ECG database for a two-stage analysis with a deep learning model. We first trained the network to identify right bundle branch block (RBBB) pattern, and then, we transferred the first-stage learning to the second task to diagnose the type 1 Brugada ECG pattern. The diagnostic performance of the deep learning model was compared to that of board-certified practicing cardiologists. The model was also validated by the independent international data of ECGs. RESULTS The AUC (area under the curve) of the deep learning model in diagnosing the type 1 Brugada ECG pattern was 0.96 (sensitivity: 88.4%, specificity: 89.1%). The sensitivity and specificity of the cardiologists for the diagnosis of the type 1 Brugada ECG pattern were 62.7±17.8%, and 98.5±3.0%, respectively. The diagnoses by the deep learning model were highly consistent with the standard diagnoses (Kappa coefficient: 0.78, McNemar test, P = 0.86). However, the diagnoses by the cardiologists were significantly different from the standard diagnoses, with only moderate consistency (Kappa coefficient: 0.60, McNemar test, P = 2.35x10-22). For the international validation, the AUC of the deep learning model for diagnosing the type 1 Brugada ECG pattern was 0.99 (sensitivity: 85.7%, specificity: 100.0%). CONCLUSIONS The deep learning-enabled ECG model for diagnosing Brugada syndrome is a robust screening tool with better diagnostic sensitivity than that of cardiologists.

P5027Prognosis and evaluation of the risk markers of arrhythmia in a large population of Brugada syndrome patients

2019 ◽  
Vol 40 (Supplement_1) ◽  
Author(s):  
V Probst ◽  
S Anys ◽  
F Sacher ◽  
J Briand ◽  
B Guyomarch ◽  
...  
Keyword(s):  
Brugada Syndrome ◽  
Ventricular Arrhythmias ◽  
Clinical Status ◽  
Large Population ◽  
Fragmented Qrs ◽  
Increased Risk ◽  
Inherited Arrhythmia ◽  
St Elevation

Abstract Introduction Brugada syndrome (BrS) is an inherited arrhythmia syndrome with an increased risk of sudden cardiac death (SCD) despite a structurally normal heart. Many parameters have been suggested to be associated with the risk of ventricular arrhythmias, but only previous symptoms and spontaneous ECG pattern have been consistently associated with the risk of ventricular arrhythmia occurrence. Objective The aim of this study was to evaluate the association of these parameters with arrhythmic events in the largest cohort of BrS patients ever described. Methods Consecutive patients affected with BrS were recruited in a multicentric prospective registry in France (15 centers) between 1994 and 2016. Data were prospectively collected with an average follow-up of 6.5±4.7 years. ECGs were reviewed by 2 physicians blinded to clinical status. Results In this study, we enrolled a total of 1613 patients (mean age 45±15 years; 1119 males, 69%). At baseline, 462 patients (29%) were symptomatic (51 (3%) aborted SCD, 257 (16%) syncope). A spontaneous type 1 ECG pattern was present in 505 patients (31%). Implantable cardiac defibrillator was implanted in 477 patients (30%). During the follow-up, 91 patients (6%) underwent arrhythmic events (16 SCD (10%), 48 appropriate ICD therapy (3%) and 27 ventricular arrhythmias (2%). Thirty-six patients (2%) died of non-arrhythmic causes. Mean age at the first event was 44±15 years. In our cohort, event predictors were SCD (HR: 18.3; 95% CI: 11.2–29.8; p<0.0001), syncope (HR: 2.9; 95% CI: 1.8–4.9; p<0.0001), age >60 years (HR: 0.11; 95% CI: 0.032–0.377; p=0,0004), gender (HR: 2.96; 95% CI: 1.6–5.4; p=0.0005), spontaneous type 1 (HR: 2.14; 95% CI: 1.42–3.23; p=0.0003), type 1 ST elevation in peripheral ECG lead (HR: 3.6; 95% CI: 1.9–7.1; p=0,0001), fragmented QRS (HR: 3.37; 95% CI: 1.37–8.32; p=0,008), AvR sign (HR: 2.2; 95% CI: 1.4–3.8; p=0,0007), QRS >120ms in D2 lead (HR: 2.2; 95% CI: 1.4–3.6; p=0,001) and QRS >90ms in V6 (HR: 2.1; 95% CI: 1.3–3.3; p=0,001). All the others parameters including early repolarization pattern (ERP) and EPS were not predictor of events. Conclusion In the largest cohort of BrS patients ever described, we confirmed that symptoms, age, gender, spontaneous type 1, type 1 ST elevation in peripheral ECG lead, fragmented QRS, AvR sign, QRS >120ms in D2 and QRS >90ms in V6 are associated with arrhythmic events whereas ERP and EPS were not.

Deep Learning

2009 ◽  
Author(s):  
Stellan Ohlsson
Keyword(s):  
Deep Learning
Sign in / Sign up

Export Citation Format

Share Document