Advertisement
Canadian Journal of Cardiology

A Deep-Learning Algorithm-Enhanced System Integrating Electrocardiograms and Chest X-rays for Diagnosing Aortic Dissection

Published:October 03, 2021DOI:https://doi.org/10.1016/j.cjca.2021.09.028

      Abstract

      Background

      Chest pain is the most common symptom of aortic dissection (AD), but it is often confused with other prevalent cardiopulmonary diseases. We aimed to develop deep-learning models (DLMs) with electrocardiography (ECG) and chest x-ray (CXR) features to detect AD and evaluate their performance.

      Methods

      This study included 43,473 patients in the emergency department (ED) between July 2012 and December 2019 for retrospective DLM development. A development cohort including 49,071 ED records (120 AD type A and 64 AD type B) was used to train DLMs for ECG and CXR, and 9904 independent ED records (40 AD type A and 34 AD type B) were used to validate DLM performance. Human-machine competitions of ECG and CXR were conducted. Patient characteristics and laboratory results were used to enhance the diagnostic accuracy. The DLM-enabled AD diagnostic process was prospectively evaluated in 25,885 ED visits.

      Results

      The area under the curves (AUCs) of the ECG and CXR models were 0.918 and 0.857 for detecting AD in a human-machine competition, respectively, which were better than those of the participating physicians. In the validation cohort, the AUCs of the integrated model were 0.882, 0.960, and 0.813 in all AD, AD type A, and AD type B patients, respectively, with a sensitivity of 100.0% and a specificity of 81.7% for AD type A. In patients with chest pain and D-dimer tests, the DLM could predict more precisely, achieving a positive predictive value of 62.5% in the prospective evaluation.

      Conclusions

      DLMs may serve as decision-supporting tools for identification of AD and facilitate differential diagnosis in patients with acute chest pain.

      Résumé

      Contexte

      La douleur thoracique est le symptôme le plus courant de la dissection aortique (DA), mais celle-ci est souvent confondue avec d’autres maladies cardiopulmonaires fréquentes. Notre objectif était de développer des modèles d’apprentissage profond (MAP) permettant l’analyse de données d’électrocardiographie (ECG) et de radiographie pulmonaire en vue de détecter la DA et d’évaluer leurs performances.

      Méthodologie

      Cette étude portait sur le développement rétrospectif de MAP à partir des données recueillies chez 43 473 patients admis aux urgences entre juillet 2012 et décembre 2019. Une cohorte de développement comportant 49 071 dossiers des urgences (120 cas de DA de type A et 64 cas de DA de type B) a servi à entraîner les MAP à l’analyse des données d’ECG et de radiographie pulmonaire, et 9 904 dossiers des urgences indépendants (40 cas de DA de type A et 34 cas de DA de type B) ont été utilisés pour valider les performances des MAP. L’analyse des données d’ECG et de radiographie pulmonaire a fait l’objet de compétitions entre humains et machines. Les caractéristiques des patients et les résultats des analyses de laboratoire ont été utilisés pour améliorer la précision du diagnostic. Le processus de diagnostic de la DA faisant appel aux MAP a été évalué de manière prospective sur 25 885 consultations aux urgences.

      Résultats

      La surface sous la courbe (SSC) des modèles d’analyse des données d’ECG et de radiographie pulmonaire était respectivement de 0,918 et 0,857 pour la détection de la DA dans une compétition entre humains et machine, ce qui constituait un meilleur résultat que celui obtenu par les médecins participants. Au sein de la cohorte de validation, la SSC du modèle intégré était de 0,882 dans l’ensemble des cas de DA, de 0,960 dans les cas de DA de type A et de 0,813 dans les cas de DA de type B, la sensibilité et la spécificité ayant atteint respectivement 100,0 % et 81,7 % dans les cas de DA de type A. Chez les patients présentant des douleurs thoraciques et dont le taux de D-dimères avait été mesuré, les MAP pouvaient donner des prévisions plus précises, une valeur prédictive positive de 62,5 % ayant été obtenue dans le cadre de l’évaluation prospective.

      Conclusions

      Les MAP peuvent servir d’outils d’aide à la décision dans le dépistage de la DA et faciliter le diagnostic différentiel chez les patients présentant des douleurs thoraciques aiguës.
      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

        • Bhuiya F.A.
        • Pitts S.R.
        • McCaig L.F.
        Emergency department visits for chest pain and abdominal pain: United States, 1999-2008.
        NCHS Data Brief. 2010; : 1-8
        • Amsterdam E.A.
        • Kirk J.D.
        • Bluemke D.A.
        • et al.
        Testing of low-risk patients presenting to the emergency department with chest pain: a scientific statement from the American Heart Association.
        Circulation. 2010; 122: 1756-1776
        • Evangelista A.
        • Isselbacher E.M.
        • Bossone E.
        • et al.
        Insights from the International Registry of Acute Aortic Dissection: a 20-year experience of collaborative clinical research.
        Circulation. 2018; 137: 1846-1860
        • Salmasi M.Y.
        • Al-Saadi N.
        • Hartley P.
        • et al.
        The risk of misdiagnosis in acute thoracic aortic dissection: a review of current guidelines.
        Heart. 2020; 106: 885-891
        • Harris K.M.
        • Strauss C.E.
        • Eagle K.A.
        • et al.
        Correlates of delayed recognition and treatment of acute type A aortic dissection: the International Registry of Acute Aortic Dissection (IRAD).
        Circulation. 2011; 124: 1911-1918
        • Zhan S.
        • Hong S.
        • Shan-Shan L.
        • et al.
        Misdiagnosis of aortic dissection: experience of 361 patients.
        J Clin Hypertens (Greenwich). 2012; 14: 256-260
        • Hansen M.S.
        • Nogareda G.J.
        • Hutchison S.J.
        Frequency of and inappropriate treatment of misdiagnosis of acute aortic dissection.
        Am J Cardiol. 2007; 99: 852-856
        • Erbel R.
        • Aboyans V.
        • Boileau C.
        • et al.
        2014 ESC Guidelines on the diagnosis and treatment of aortic diseases: document covering acute and chronic aortic diseases of the thoracic and abdominal aorta of the adult. The Task Force for the Diagnosis and Treatment of Aortic Diseases of the European Society of Cardiology (ESC).
        Eur Heart J. 2014; 35: 2873-2926
        • Costin N.I.
        • Korach A.
        • Loor G.
        • et al.
        Patients with type a acute aortic dissection presenting with an abnormal electrocardiogram.
        Ann Thorac Surg. 2018; 105: 92-99
        • von Kodolitsch Y.
        • Nienaber C.A.
        • Dieckmann C.
        • et al.
        Chest radiography for the diagnosis of acute aortic syndrome.
        Am J Med. 2004; 116: 73-77
        • Funakoshi H.
        • Mizobe M.
        • Homma Y.
        • Nakashima Y.
        • Takahashi J.
        • Shiga T.
        The diagnostic accuracy of the mediastinal width on supine anteroposterior chest radiographs with nontraumatic Stanford type A acute aortic dissection.
        J Gen Fam Med. 2018; 19: 45-49
        • Esteva A.
        • Robicquet A.
        • Ramsundar B.
        • et al.
        A guide to deep learning in healthcare.
        Nat Med. 2019; 25: 24-29
        • LeCun Y.
        • Bengio Y.
        • Hinton G.
        Deep learning.
        Nature. 2015; 521: 436-444
        • Attia Z.I.
        • Friedman P.A.
        • Noseworthy P.A.
        • et al.
        Age and sex estimation using artificial intelligence from standard 12-lead ECGs.
        Circ Arrhythm Electrophysiol. 2019; 12e007284
        • Attia Z.I.
        • Noseworthy P.A.
        • Lopez-Jimenez F.
        • et al.
        An artificial intelligence-enabled ECG algorithm for the identification of patients with atrial fibrillation during sinus rhythm: a retrospective analysis of outcome prediction.
        Lancet. 2019; 394: 861-867
        • Hosny A.
        • Parmar C.
        • Quackenbush J.
        • Schwartz L.H.
        • Aerts H.
        Artificial intelligence in radiology.
        Nat Rev Cancer. 2018; 18: 500-510
        • Lin C.S.
        • Lin C.
        • Fang W.H.
        • et al.
        A deep-learning algorithm (ECG12Net) for detecting hypokalemia and hyperkalemia by electrocardiography: algorithm development.
        JMIR Med Inform. 2020; 8e15931
        • Rajpurkar P.
        • Irvin J.
        • Ball R.L.
        • et al.
        Deep learning for chest radiograph diagnosis: a retrospective comparison of the CheXNeXt algorithm to practicing radiologists.
        PLoS Med. 2018; 15e1002686
      1. Huang G, Liu Z, Van Der Maaten L, Weinberger KQ. Densely connected convolutional networks. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). 2017;2261-2269.

        • Krizhevsky A.
        • Sutskever I.
        • Hinton G.E.
        ImageNet classification with deep convolutional neural networks. Proceedings of the 25th International Conference on Neural Information Processing Systems-Vol 1.
        Curran Associates Inc., Lake Tahoe, Nevada2012: 1097-1105
        • Kohl L.P.
        • Isselbacher E.M.
        • Majahalme N.
        • et al.
        Comparison of outcomes in DeBakey type AI versus AII aortic dissection.
        Am J Cardiol. 2018; 122: 689-695
        • Huo D.
        • Kou B.
        • Zhou Z.
        • Lv M.
        A machine learning model to classify aortic dissection patients in the early diagnosis phase.
        Sci Rep. 2019; 9: 2701
        • Liu L.
        • Zhang C.
        • Zhang G.
        • et al.
        A study of aortic dissection screening method based on multiple machine learning models.
        J Thorac Dis. 2020; 12: 605-614
        • Lee J.
        • Kirschner J.
        • Pawa S.
        • Wiener D.E.
        • Newman D.H.
        • Shah K.
        Computed tomography use in the adult emergency department of an academic urban hospital from 2001 to 2007.
        Ann Emerg Med. 2010; 56: 591-596
        • Broder J.
        • Warshauer D.M.
        Increasing utilization of computed tomography in the adult emergency department, 2000-2005.
        Emerg Radiol. 2006; 13: 25-30
        • Castelvecchi D.
        Can we open the black box of AI?.
        Nature. 2016; 538: 20-23
      2. Yang Z, Yang D, Dyer C, He X, Smola A, Hovy E. Hierarchical attention networks for document classification. Proceedings of the 2016 conference of the North American chapter of the Association for Computational Linguistics: Human Language Technologies: Association for Computational Linguistics. 2016;1480-1489.

        • Zhou B.
        • Khosla A.
        • Lapedriza A.
        • Oliva A.
        • Torralba A.
        Learning deep features for discriminative localization. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).
        IEEE, Las Vegas, Nevada, USA2016