Automatic diagnosis of the 12-lead ECG using a deep neural network

The role of automatic electrocardiogram (ECG) analysis in clinical practice is limited by the accuracy of existing models. Deep Neural Networks (DNNs) are models composed of stacked transformations that learn tasks by examples. This technology has recently achieved striking success in a variety of task and there are great expectations on how it might improve clinical practice. Here we present a DNN model trained in a dataset with more than 2 million labeled exams analyzed by the Telehealth Network of Minas Gerais and collected under the scope of the CODE (Clinical Outcomes in Digital Electrocardiology) study. The DNN outperform cardiology resident medical doctors in recognizing 6 types of abnormalities in 12-lead ECG recordings, with F1 scores above 80% and specificity over 99%. These results indicate ECG analysis based on DNNs, previously studied in a single-lead setup, generalizes well to 12-lead exams, taking the technology closer to the standard clinical practice. The role of automatic electrocardiogram (ECG) analysis in clinical practice is limited by the accuracy of existing models. In that context, the authors present a Deep Neural Network (DNN) that recognizes different abnormalities in ECG recordings which matches or outperform cardiology and emergency resident medical doctors.

[1]  N. Estes,et al.  Computerized Interpretation of ECGs Supplement Not a Substitute , 2013 .

[2]  S. T. Buckland,et al.  An Introduction to the Bootstrap. , 1994 .

[3]  M. Kenward,et al.  An Introduction to the Bootstrap , 2007 .

[4]  Shamim Nemati,et al.  Detection of Paroxysmal Atrial Fibrillation using Attention-based Bidirectional Recurrent Neural Networks , 2018, KDD.

[5]  M. B. Alkmim,et al.  Improving patient access to specialized health care: the Telehealth Network of Minas Gerais, Brazil. , 2012, Bulletin of the World Health Organization.

[6]  Pablo Laguna,et al.  Computational techniques for ECG analysis and interpretation in light of their contribution to medical advances , 2018, Journal of The Royal Society Interface.

[7]  Geraint Rees,et al.  Clinically applicable deep learning for diagnosis and referral in retinal disease , 2018, Nature Medicine.

[8]  Andrew H. Beck,et al.  Diagnostic Assessment of Deep Learning Algorithms for Detection of Lymph Node Metastases in Women With Breast Cancer , 2017, JAMA.

[9]  T. H. Nguyen,et al.  The global, regional, and national burden of inflammatory bowel disease in 195 countries and territories, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017 , 2019, The lancet. Gastroenterology & hepatology.

[10]  Tara N. Sainath,et al.  Deep Neural Networks for Acoustic Modeling in Speech Recognition: The Shared Views of Four Research Groups , 2012, IEEE Signal Processing Magazine.

[11]  I. Hjermann,et al.  [The Minnesota code]. , 1970, Nordisk medicin.

[12]  Jeffrey M. Hausdorff,et al.  Physionet: Components of a New Research Resource for Complex Physiologic Signals". Circu-lation Vol , 2000 .

[13]  P W Macfarlane,et al.  Testing the performance of ECG computer programs: the CSE diagnostic pilot study. , 1987, Journal of electrocardiology.

[14]  W. Stead.  Clinical Implications and Challenges of Artificial Intelligence and Deep Learning. , 2018, JAMA.

[15]  Sven Behnke,et al.  Evaluation of Pooling Operations in Convolutional Architectures for Object Recognition , 2010, ICANN.

[16]  G. Veronese,et al.  Emergency physician accuracy in interpreting electrocardiograms with potential ST-segment elevation myocardial infarction: Is it enough? , 2016, Acute cardiac care.

[17]  Jian Sun,et al.  Identity Mappings in Deep Residual Networks , 2016, ECCV.

[18]  Rishikesan Kamaleswaran,et al.  A robust deep convolutional neural network for the classification of abnormal cardiac rhythm using single lead electrocardiograms of variable length , 2018, Physiological measurement.

[19]  A. F. Adams,et al.  The Survey , 2021, Dyslexia in Higher Education.

[20]  E. Antman,et al.  A Neural Network System for Detection of Atrial Fibrillation in Ambulatory Electrocardiograms , 1994, Journal of cardiovascular electrophysiology.

[21]  Masoumeh Haghpanahi,et al.  Cardiologist-level arrhythmia detection and classification in ambulatory electrocardiograms using a deep neural network , 2019, Nature Medicine.

[22]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[23]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[24]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[25]  E. Beck,et al.  Protecting the confidentiality and security of personal health information in low- and middle-income countries in the era of SDGs and Big Data , 2016, Global health action.

[26]  Mohammad Hossein Khosravi,et al.  Global, regional, and national age-sex-specific mortality for 282 causes of death in 195 countries and territories, 1980–2017: a systematic analysis for the Global Burden of Disease Study 2017 , 2018, Lancet.

[27]  Mohammad Hossein Khosravi,et al.  Global, regional, and national age-sex specific mortality for 264 causes of death, 1980–2016: a systematic analysis for the Global Burden of Disease Study 2016 , 2017, Lancet.

[28]  H. Wellens,et al.  Computer-Interpreted Electrocardiograms: Benefits and Limitations. , 2017, Journal of the American College of Cardiology.

[29]  Takaya Saito,et al.  The Precision-Recall Plot Is More Informative than the ROC Plot When Evaluating Binary Classifiers on Imbalanced Datasets , 2015, PloS one.

[30]  Meng Wu,et al.  ENCASE: An ENsemble ClASsifiEr for ECG classification using expert features and deep neural networks , 2017, 2017 Computing in Cardiology (CinC).

[31]  Saeed Babaeizadeh,et al.  Densely connected convolutional networks and signal quality analysis to detect atrial fibrillation using short single-lead ECG recordings , 2017, 2017 Computing in Cardiology (CinC).

[32]  Tara N. Sainath,et al.  The shared views of four research groups ) , 2012 .

[33]  Shweta H. Jambukia,et al.  Classification of ECG signals using machine learning techniques: A survey , 2015, 2015 International Conference on Advances in Computer Engineering and Applications.

[34]  J. L. Willems,et al.  The diagnostic performance of computer programs for the interpretation of electrocardiograms. , 1992, The New England journal of medicine.

[35]  A. Shah,et al.  Errors in the computerized electrocardiogram interpretation of cardiac rhythm. , 2007, Journal of electrocardiology.

[36]  Barbara C. A. Marino,et al.  Implementing myocardial infarction systems of care in low/middle-income countries , 2018, Heart.

[37]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[38]  S. Luo,et al.  A review of electrocardiogram filtering. , 2010, Journal of electrocardiology.

[39]  Q. Mcnemar.  Note on the sampling error of the difference between correlated proportions or percentages , 1947, Psychometrika.

[40]  Mohammed J. Zaki,et al.  Lazy Associative Classification , 2006, Sixth International Conference on Data Mining (ICDM'06).

[41]  Qiao Li,et al.  AF classification from a short single lead ECG recording: The PhysioNet/computing in cardiology challenge 2017 , 2017, 2017 Computing in Cardiology (CinC).

[42]  C. Naylor,et al.  On the Prospects for a (Deep) Learning Health Care System , 2018, JAMA.

[43]  U. Rajendra Acharya,et al.  Application of deep convolutional neural network for automated detection of myocardial infarction using ECG signals , 2017, Inf. Sci..

[44]  Keiichi Fukuda,et al.  Artificial intelligence to predict needs for urgent revascularization from 12-leads electrocardiography in emergency patients , 2019, PloS one.

[45]  E. W. Hancock,et al.  AHA/ACCF/HRS recommendations for the standardization and interpretation of the electrocardiogram: part IV: the ST segment, T and U waves, and the QT interval: a scientific statement from the American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology; the , 2009, Journal of the American College of Cardiology.

[46]  Sergey Ioffe,et al.  Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , 2015, ICML.

[47]  Jacob Cohen.  A Coefficient of Agreement for Nominal Scales , 1960 .

[48]  E. W. Hancock,et al.  AHA/ACCF/HRS recommendations for the standardization and interpretation of the electrocardiogram: part III: intraventricular conduction disturbances: a scientific statement from the American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology; the American , 2009, Journal of the American College of Cardiology.

[49]  G. Lip,et al.  Accuracy of diagnosing atrial fibrillation on electrocardiogram by primary care practitioners and interpretative diagnostic software: analysis of data from screening for atrial fibrillation in the elderly (SAFE) trial , 2007, BMJ : British Medical Journal.

[50]  Naif Alajlan,et al.  Deep learning approach for active classification of electrocardiogram signals , 2016, Inf. Sci..

[51]  P W Macfarlane,et al.  Automated serial ECG comparison based on the Minnesota code. , 1996, Journal of electrocardiology.

[52]  Ram Bilas Pachori,et al.  A Novel Approach for Detection of Myocardial Infarction From ECG Signals of Multiple Electrodes , 2019, IEEE Sensors Journal.

[53]  J. Rapin,et al.  A deep neural network learning algorithm outperforms a conventional algorithm for emergency department electrocardiogram interpretation. , 2019, Journal of electrocardiology.

[54]  P. Macfarlane,et al.  The university of glasgow (Uni-G) ECG analysis program , 2005, Computers in Cardiology, 2005.

[55]  Geoffrey E. Hinton.  Deep Learning-A Technology With the Potential to Transform Health Care. , 2018, JAMA.

[56]  A. Marty.  Minnesota Code Manual of Electrocardiographic Findings , 1983 .

[57]  Jian Sun,et al.  Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[58]  R. Sassi,et al.  PDF-ECG in clinical practice: A model for long-term preservation of digital 12-lead ECG data. , 2017, Journal of electrocardiology.

[59]  Dana H. Brooks,et al.  Detecting ischemic stress to the myocardium using laplacian eigenmaps and changes to conduction velocity , 2017, 2017 Computing in Cardiology (CinC).

[60]  P W Macfarlane,et al.  Methodology of ECG Interpretation in the Glasgow Program , 1990, Methods of Information in Medicine.