Advertisement
Canadian Journal of Cardiology

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

  • Author Footnotes
    ‡ These authors contributed equally to this work.
    Chih-Min Liu
    Footnotes
    ‡ These authors contributed equally to this work.
    Affiliations
    Heart Rhythm Center, Division of Cardiology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan

    Institute of Clinical Medicine and Faculty of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
    Search for articles by this author
  • Author Footnotes
    ‡ These authors contributed equally to this work.
    Chien-Liang Liu
    Footnotes
    ‡ These authors contributed equally to this work.
    Affiliations
    Department of Industrial Engineering and Management, National Yang Ming Chiao Tung University, Hsinchu, Taiwan
    Search for articles by this author
  • Kai-Wen Hu
    Affiliations
    Department of Industrial Engineering and Management, National Yang Ming Chiao Tung University, Hsinchu, Taiwan
    Search for articles by this author
  • Vincent S. Tseng
    Affiliations
    Department of Computer Science, National Yang Ming Chiao Tung University, Hsinchu, Taiwan
    Search for articles by this author
  • Shih-Lin Chang
    Affiliations
    Heart Rhythm Center, Division of Cardiology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan

    Institute of Clinical Medicine and Faculty of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
    Search for articles by this author
  • Yenn-Jiang Lin
    Affiliations
    Heart Rhythm Center, Division of Cardiology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan

    Institute of Clinical Medicine and Faculty of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
    Search for articles by this author
  • Li-Wei Lo
    Affiliations
    Heart Rhythm Center, Division of Cardiology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan

    Institute of Clinical Medicine and Faculty of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
    Search for articles by this author
  • Fa-Po Chung
    Affiliations
    Heart Rhythm Center, Division of Cardiology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan

    Institute of Clinical Medicine and Faculty of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
    Search for articles by this author
  • Tze-Fan Chao
    Affiliations
    Heart Rhythm Center, Division of Cardiology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan

    Institute of Clinical Medicine and Faculty of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
    Search for articles by this author
  • Ta-Chuan Tuan
    Affiliations
    Heart Rhythm Center, Division of Cardiology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan

    Institute of Clinical Medicine and Faculty of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
    Search for articles by this author
  • Jo-Nan Liao
    Affiliations
    Heart Rhythm Center, Division of Cardiology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan

    Institute of Clinical Medicine and Faculty of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
    Search for articles by this author
  • Chin-Yu Lin
    Affiliations
    Heart Rhythm Center, Division of Cardiology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan

    Institute of Clinical Medicine and Faculty of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
    Search for articles by this author
  • Ting-Yung Chang
    Affiliations
    Heart Rhythm Center, Division of Cardiology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan

    Institute of Clinical Medicine and Faculty of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
    Search for articles by this author
  • Cathy Shen-Jang Fann
    Affiliations
    Institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan
    Search for articles by this author
  • Satoshi Higa
    Affiliations
    Cardiac Electrophysiology and Pacing Laboratory, Division of Cardiovascular Medicine, Makiminato Central Hospital, Okinawa, Japan
    Search for articles by this author
  • Nobumori Yagi
    Affiliations
    Division of Cardiovascular Medicine, Nakagami Hospital, Okinawa, Japan
    Search for articles by this author
  • Yu-Feng Hu
    Correspondence
    Corresponding author: Dr Yu-Feng Hu, Taipei Veterans General Hospital, 201 Sec. 2, Shih-Pai Road, Taipei, Taiwan. Tel.: +886-2-2875-7156; fax: +886-2-2873-5656.
    Affiliations
    Heart Rhythm Center, Division of Cardiology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan

    Institute of Clinical Medicine and Faculty of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan

    Institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan
    Search for articles by this author
  • Shih-Ann Chen
    Affiliations
    Heart Rhythm Center, Division of Cardiology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan

    Institute of Clinical Medicine and Faculty of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan

    Cardiovascular Center, Taichung Veterans General Hospital, Taichung, Taiwan
    Search for articles by this author
  • Author Footnotes
    ‡ These authors contributed equally to this work.
Published:August 27, 2021DOI:https://doi.org/10.1016/j.cjca.2021.08.014

      Abstract

      Background

      Brugada syndrome is a major cause of sudden cardiac death in young people and has distinctive electrocardiographic (ECG) features. We aimed to develop a deep learning–enabled ECG model for automatic screening for Brugada syndrome to identify these patients at an early point in time, thus allowing for life-saving therapy.

      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 1:1 allocation) were extracted from the hospital-based ECG database for a 2-stage analysis with a deep learning model. After trained network for identifying right bundle branch block pattern, 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 with that of board-certified practicing cardiologists. The model was further validated in an independent ECG data set collected from hospitals in Taiwan and Japan.

      Results

      The diagnoses by the deep learning model (area under the receiver operating characteristic curve [AUC] 0.96, sensitivity 88.4%, specificity 89.1%) were highly consistent with the standard diagnoses (kappa coefficient 0.78). However, the diagnoses by the cardiologists were significantly different from the standard diagnoses, with only moderate consistency (kappa coefficient 0.63). In the independent ECG cohort, the deep learning model still reached a satisfactory diagnostic performance (AUC 0.89, sensitivity 86.0%, specificity 90.0%).

      Conclusions

      We present the first deep learning–enabled ECG model for diagnosing Brugada syndrome, which appears to be a robust screening tool with a diagnostic potential rivalling trained physicians.

      Résumé

      Contexte

      Le syndrome de Brugada est une cause majeure de mort subite cardiaque chez les jeunes et présente des caractéristiques électrocardiographiques (ECG) distinctives. Nous avons cherché à développer un modèle d'ECG basé sur l'apprentissage profond pour le dépistage automatique du syndrome de Brugada afin d'identifier ces patients à un stade précoce, ce qui permettrait de mettre en place un traitement salutaire.

      Méthodes

      Un total de 276 ECG avec un modèle d'ECG de Brugada de type 1 (276 ECG de Brugada de type 1 et 276 autres ECG, non Brugada, collectés aléatoirement, pour une allocation 1:1) ont été extraits de la base de données d'ECG de l'hôpital pour une analyse en deux étapes avec un modèle d'apprentissage profond. Après avoir entraîné le réseau à identifier le tracé du bloc de branche droit, nous avons transposé l'apprentissage de cette première étape vers une seconde tâche afin de diagnostiquer le tracé de l'ECG d'un Brugada de type 1. La performance diagnostique du modèle d'apprentissage profond a été comparée à celle de cardiologues praticiens certifiés. Le modèle a ensuite été validé dans un ensemble indépendant de données ECG recueillies dans des hôpitaux de Taïwan et du Japon.

      Résultats

      Les diagnostics établis par le modèle d'apprentissage profond (aire sous la courbe de la fonction d'efficacité du récepteur [ASC] 0,96, sensibilité 88,4 %, spécificité 89,1 %) étaient très cohérents avec les diagnostics standard (coefficient kappa 0,78). Cependant, les diagnostics des cardiologues étaient significativement différents des diagnostics standards, moyennant une cohérence modérée (coefficient kappa 0,63). Dans la cohorte d'ECG indépendants, le modèle d'apprentissage profond atteignait encore une performance diagnostique satisfaisante (ASC 0,89, sensibilité 86,0 %, spécificité 90,0 %).

      Conclusions

      Nous présentons le premier modèle d'ECG basé sur l'apprentissage profond pour le diagnostic du syndrome de Brugada, qui s'avère être un outil de dépistage robuste avec un potentiel diagnostique rivalisant avec celui des médecins formés.
      To read this article in full you will need to make a payment

      Purchase one-time access:

      Academic & Personal: 24 hour online accessCorporate R&D Professionals: 24 hour online access
      One-time access price info
      • For academic or personal research use, select 'Academic and Personal'
      • For corporate R&D use, select 'Corporate R&D Professionals'

      Subscribe:

      Subscribe to Canadian Journal of Cardiology
      Already a print subscriber? Claim online access
      Already an online subscriber? Sign in
      Institutional Access: Sign in to ScienceDirect

      References

        • Brugada J.
        • Campuzano O.
        • Arbelo E.
        • Sarquella-Brugada G.
        • Brugada R.
        Present status of Brugada syndrome: JACC state-of-the-art review.
        J Am Coll Cardiol. 2018; 72: 1046-1059
        • Al-Khatib S.M.
        • Stevenson W.G.
        • Ackerman M.J.
        • et al.
        2017 AHA/ACC/HRS guideline for management of patients with ventricular arrhythmias and the prevention of sudden cardiac death: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines and the Heart Rhythm Society.
        J Am Coll Cardiol. 2018; 72: e91-e220
        • Priori S.G.
        • Blomstrom-Lundqvist C.
        • Mazzanti A.
        • et al.
        2015 ESC guidelines for the management of patients with ventricular arrhythmias and the prevention of sudden cardiac death: the Task Force for the Management of Patients with Ventricular Arrhythmias and the Prevention of Sudden Cardiac Death of the European Society of Cardiology (ESC). Endorsed by: Association for European Paediatric and Congenital Cardiology (AEPC).
        Eur Heart J. 2015; 36: 2793-2867
        • Priori S.G.
        • Wilde A.A.
        • Horie M.
        • et al.
        HRS/EHRA/APHRS expert consensus statement on the diagnosis and management of patients with inherited primary arrhythmia syndromes: document endorsed by HRS, EHRA, and APHRS in May 2013 and by ACCF, AHA, PACES, and AEPC in June 2013.
        Heart Rhythm. 2013; 10: 1932-1963
        • Sarquella-Brugada G.
        • Campuzano O.
        • Arbelo E.
        • Brugada J.
        • Brugada R.
        Brugada syndrome: clinical and genetic findings.
        Genet Med. 2016; 18: 3-12
        • Quan X.Q.
        • Li S.
        • Liu R.
        • Zheng K.
        • Wu X.F.
        • Tang Q.
        A meta-analytic review of prevalence for Brugada ECG patterns and the risk for death.
        Medicine. 2016; 95: e5643
        • Postema P.G.
        About Brugada syndrome and its prevalence.
        Europace. 2012; 14: 925-928
        • Tan H.L.
        • Hofman N.
        • van Langen I.M.
        • van der Wal A.C.
        • Wilde A.A.
        Sudden unexplained death: heritability and diagnostic yield of cardiological and genetic examination in surviving relatives.
        Circulation. 2005; 112: 207-213
        • Sieira J.
        • Dendramis G.
        • Brugada P.
        Pathogenesis and management of Brugada syndrome.
        Nat Rev Cardiol. 2016; 13: 744-756
        • Sroubek J.
        • Probst V.
        • Mazzanti A.
        • et al.
        Programmed ventricular stimulation for risk stratification in the Brugada syndrome: a pooled analysis.
        Circulation. 2016; 133: 622-630
        • Probst V.
        • Veltmann C.
        • Eckardt L.
        • et al.
        Long-term prognosis of patients diagnosed with Brugada syndrome: results from the FINGER Brugada Syndrome Registry.
        Circulation. 2010; 121: 635-643
        • Sieira J.
        • Ciconte G.
        • Conte G.
        • et al.
        Asymptomatic Brugada syndrome: clinical characterization and long-term prognosis.
        Circ Arrhythm Electrophysiol. 2015; 8: 1144-1150
        • Gallahue F.E.
        • Uzgiris R.
        • Burke R.
        • Abrahams W.
        Brugada syndrome presenting as an “acute myocardial infarction.”.
        J Emerg Med. 2009; 37: 15-20
        • Chutani S.
        • Imran N.
        • Grubb B.
        • Kanjwal Y.
        Propafenone-induced Brugada-like ECG changes mistaken as acute myocardial infarction.
        Emerg Med J. 2008; 25: 117-118
        • Hannun A.Y.
        • Rajpurkar P.
        • Haghpanahi M.
        • et al.
        Cardiologist-level arrhythmia detection and classification in ambulatory electrocardiograms using a deep neural network.
        Nat Med. 2019; 25: 65-69
        • Bizopoulos P.
        • Koutsouris D.
        Deep learning in cardiology.
        IEEE Rev Biomed Engin. 2019; 12: 168-193
        • Lovis C.
        Unlocking the power of artificial intelligence and big data in medicine.
        J Med Internet Res. 2019; 21e16607
        • Bussink B.E.
        • Holst A.G.
        • Jespersen L.
        • et al.
        Right bundle branch block: prevalence, risk factors, and outcome in the general population: results from the Copenhagen City Heart Study.
        Eur Heart J. 2013; 34: 138-146
        • McHugh M.L.
        Interrater reliability: the kappa statistic.
        Biochem Med (Zagreb). 2012; 22: 276-282
        • Chollet F.
        Deep Learning with Python.
        Manning Publications, 2017
        • Antzelevitch C.
        • Brugada P.
        • Borggrefe M.
        • et al.
        Brugada syndrome: report of the second consensus conference: endorsed by the Heart Rhythm Society and the European Heart Rhythm Association.
        Circulation. 2005; 111: 659-670
        • Schwartz P.J.
        • Stramba-Badiale M.
        • Crotti L.
        • et al.
        Prevalence of the congenital long-QT syndrome.
        Circulation. 2009; 120: 1761-1767
        • Hayashi K.
        • Fujino N.
        • Uchiyama K.
        • et al.
        Long QT syndrome and associated gene mutation carriers in Japanese children: results from ECG screening examinations.
        Clin Sci (Lond). 2009; 117: 415-424
        • Murakoshi N.
        • Aonuma K.
        Epidemiology of arrhythmias and sudden cardiac death in Asia.
        Circ J. 2013; 77: 2419-2431
        • Peritz D.C.
        • Chung E.H.
        Criteria for evaluating rSr′ patterns due to high precordial ECG lead placement accurately confirm absence of a Brugada ECG pattern.
        J Electrocardiol. 2016; 49: 182-186
        • Brugada P.
        • Brugada J.
        Right bundle branch block, persistent ST segment elevation and sudden cardiac death: a distinct clinical and electrocardiographic syndrome. A multicenter report.
        J Am Coll Cardiol. 1992; 20: 1391-1396
        • Pan S.J.
        • Yang Q.
        A survey on transfer learning.
        IEEE Trans Knowl Data Eng. 2010; 22: 1345-1359
        • Murugesan B.
        • Ravichandran V.
        • Ram K.
        • et al.
        ECGNet: deep network for arrhythmia classification.
        2018 IEEE Int Symp Med Meas Appl (MeMeA) (2018). 2018; : 1-6
        • van Steenkiste G.
        • van Loon G.
        • Crevecoeur G.
        Transfer learning in ECG classification from human to horse using a novel parallel neural network architecture.
        Sci Rep. 2020; 10: 186

      Linked Article