Highlights
- •This work describes the development and evaluation of a deep neural network “artificial intelligence” algorithm that shows high accuracy to detect pulmonary hypertension and several subtypes from 12-lead ECG alone.
- •It shows that pulmonary hypertension can even be detected with this approach up to 2 years before the clinical diagnosis was made in the standard clinical workflow.
- •This provides a new approach to detect pulmonary hypertension patients that can benefit from directed treatments in a widely-accessible non-invasive way in higher risk populations.
Abstract
Background
Pulmonary hypertension (PH) is life-threatening, and often diagnosed late in its course.
We aimed to evaluate if a deep learning approach using electrocardiogram (ECG) data
alone can detect PH and clinically important subtypes.
Research Question
Does an automated deep learning approach to ECG interpretation detect PH and its clinically
important subtypes.
Study Design and Methods
Adults with right heart catheterization (RHC) or an echocardiogram within 90 days
of an ECG at the University of California, San Francisco (2012–2019) were retrospectively
identified as PH or non-PH. A deep convolutional neural network was trained on patients’
12-lead ECG voltage data. Patients were divided into training, development, and test
sets in a ratio of 7:1:2.
Results
Overall, 5016 PH and 19,454 non-PH patients were used in the study. Mean (SD) age
at time of ECG was 62.29 (17.58) years and 49.88% were female. Mean interval between
ECG and RHC or echocardiogram was 3.66 and 2.23 days for PH and non-PH patients, respectively.
In the test dataset, the model achieved an area under the receiver operating characteristic
curve (AUC), sensitivity, and specificity, respectively of 0.89, 0.79, and 0.84 to
detect PH; 0.91, 0.83, and 0.84 to detect pre-capillary PH; 0.88, 0.81, and 0.81 to
detect PAH, and 0.80, 0.73, and 0.76 to detect Group 3 PH. We additionally applied
the trained model on ECGs from participants in the test dataset that were obtained
from up to 2 years before diagnosis of PH: AUC was ≥0.79.
Interpretation
A deep learning ECG algorithm can detect PH and PH subtypes around the time of diagnosis
and can detect PH using ECGs that were done up to 2 years before RHC/echocardiogram
diagnosis. This approach has the potential to reduce diagnostic delay in PH.
Keywords
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 accessOne-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 Journal of Cardiac FailureAlready a print subscriber? Claim online access
Already an online subscriber? Sign in
Register: Create an account
Institutional Access: Sign in to ScienceDirect
REFERENCES
- Haemodynamic definitions and updated clinical classification of pulmonary hypertension.The European respiratory journal. 2019; 531801913
- An evaluation of long-term survival from time of diagnosis in pulmonary arterial hypertension from the REVEAL Registry.Chest. 2012; 142: 448-456
- The Giessen Pulmonary Hypertension Registry: Survival in pulmonary hypertension subgroups.The Journal of Heart and Lung Transplantation. 2017; 36: 957-967
Khou V, Anderson JJ, Strange G, Corrigan C, Collins N, Celermajer DS, Dwyer N, Feenstra J, Horrigan M, Keating D, Kotlyar E, Lavender M, McWilliams TJ, Steele P, Weintraub R, Whitford H, Whyte K, Williams TJ, Wrobel JP, Keogh A and Lau EM. Diagnostic delay in pulmonary arterial hypertension: Insights from the Australian and New Zealand pulmonary hypertension registry. 2020;25:863-871.
- Time from symptoms to definitive diagnosis of idiopathic pulmonary arterial hypertension: The delay study.Pulmonary circulation. 2013; 3: 89-94
- 2015 ESC/ERS Guidelines for the diagnosis and treatment of pulmonary hypertension. The Joint Task Force for the Diagnosis and Treatment of Pulmonary Hypertension of the European Society of Cardiology (ESC) and the European Respiratory Society (ERS) Endorsed by: Association for European Paediatric and Congenital Cardiology (AEPC).International Society for Heart and Lung Transplantation (ISHLT). 2015; 46: 903-975
- Estimation of total incremental health care costs in patients with atrial fibrillation in the United States.Circ Cardiovasc Qual Outcomes. 2011; 4: 313-320
- Temporal relations of atrial fibrillation and congestive heart failure and their joint influence on mortality: the Framingham Heart Study.Circulation. 2003; 107: 2920-2925
- The patient experience of pulmonary hypertension: a large cross-sectional study of UK patients.BMC Pulmonary Medicine. 2019; 19: 67
- Delay in recognition of pulmonary arterial hypertension: factors identified from the REVEAL Registry.Chest. 2011; 140: 19-26
- Screening strategies for pulmonary arterial hypertension.European Heart Journal Supplements. 2019; 21: K9-K20
- New Strategies for the Conduct of Clinical Trials in Pediatric Pulmonary Arterial Hypertension: Outcome of a Multistakeholder Meeting With Patients, Academia, Industry, and Regulators, Held at the European Medicines Agency on Monday, June 12, 2017.Journal of the American Heart Association. 2019; 8e011306
- Validity of a 5-minute focused echocardiography with A-F mnemonic performed by non-echocardiographers in the management of patients with acute chest pain.Cardiovascular ultrasound. 2015; 13: 16
Rui P and Okeyode T. National Ambulatory Medical Care Survey: 2016 National Summary Tables. Available from: https://www.cdc.gov/nchs/data/ahcd/namcs_summary/2016_namcs_web_tables.pdf. 2016.
- Electrocardiograms in Low-Risk Patients Undergoing an Annual Health Examination.JAMA internal medicine. 2017; 177: 1326-1333
- National Hospital Ambulatory Medical Care Survey: 2006 emergency department summary.National Health Statistics Report. 2008; : 1-38
- The prognostic role of ECG in primary pulmonary hypertension.Chest. 2002 Feb; 121: 513-518
- Automated and Interpretable Patient ECG Profiles for Disease Detection, Tracking, and Discovery.Circulation: Cardiovascular Quality and Outcomes. 2019; 12e005289
- Deep Residual Learning for Image Recognition.in: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). 2016: 770-778
- Cardiologist-level arrhythmia detection and classification in ambulatory electrocardiograms using a deep neural network.Nature Medicine. 2019; 25: 65-69
- Rectified linear units improve restricted boltzmann machines.in: Paper presented at: Proceedings of the 27th International Conference on International Machine Learning, Haifa, Israel2010
- Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift.Proceedings of the 32nd International Conference on Machine Learning. 2015; 37: 448-456
- Adam: A method for stochastic optimization.in: 3rd International Conference on Learning Representations, ICLR. 2015: 1-15
- Why should I trust you ?” explaining the predictions of any classifier.in: KDD ’16: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. 2016: 1135-1144https://doi.org/10.1145/2939672.2939778
- The Precision-Recall Plot Is More Informative than the ROC Plot When Evaluating Binary Classifiers on Imbalanced Datasets.PLOS ONE. 2015; 10e0118432
- A global view of pulmonary hypertension.The Lancet Respiratory Medicine. 2016; 4: 306-322
- Understanding and using sensitivity, specificity and predictive values.Indian journal of ophthalmology. 2008; 56: 45-50
- Clinical and Instrumental Characteristics of Newly Diagnosed Patients with Various Forms of Pulmonary Hypertension according to the Russian National Registry.BioMed research international. 2020; 20206836973
- Survival in pulmonary hypertension in Spain: insights from the Spanish registry.The European respiratory journal. 2012; 40: 596-603
- Artificial intelligence for early prediction of pulmonary hypertension using electrocardiography.The Journal of Heart and Lung Transplantation. 2020; 39: 805-814
- Diagnostic accuracy of transthoracic echocardiography for pulmonary hypertension: a systematic review and meta-analysis.BMJ open. 2019; 9e033084
- Echocardiography reporting of pulmonary hypertension and subsequent referral to a specialty clinic.Echocardiography. 2020; 37: 8-13
- Elevated Pulmonary Pressure Noted on Echocardiogram: A Simplified Approach to Next Steps.Journal of the American Heart Association. 2021; 10e017684
- Lack of a Tricuspid Regurgitation Doppler Signal and Pulmonary Hypertension by Invasive Measurement.Journal of the American Heart Association. 2018; 7
- Treating the right patient at the right time: Access to echocardiography in Canada.Canadian Journal of Cardiology. 2006; 22: 1029-1033
- Hospital Use of Echocardiography: Insights From the Nationwide Inpatient Sample.Journal of the American College of Cardiology. 2016; 67: 502-511
- Benefits of an open access echocardiography service: a Dutch prospective cohort study.Netherlands Heart Journal. 2013; 21: 399-405
- Artificial intelligence-enhanced electrocardiography in cardiovascular disease management.Nature Reviews Cardiology. 2021;
- Performance of a Convolutional Neural Network and Explainability Technique for 12-Lead Electrocardiogram Interpretation.JAMA cardiology. 2021;
- Detection of Hypertrophic Cardiomyopathy Using a Convolutional Neural Network-Enabled Electrocardiogram.Journal of the American College of Cardiology. 2020; 75: 722-733
- Screening for cardiac contractile dysfunction using an artificial intelligence–enabled electrocardiogram.Nature Medicine. 2019; 25: 70-74
- Artificial intelligence–enabled electrocardiograms for identification of patients with low ejection fraction: a pragmatic, randomized clinical trial.Nature Medicine. 2021; 27: 815-819
- Atrial amyloidosis: an arrhythmogenic substrate for persistent atrial fibrillation.Circulation. 2002; 106: 2091-2097
- Breathing (and Coding?) a Bit Easier: Changes to International Classification of Disease Coding for Pulmonary Hypertension.Chest. 2018; 154: 207-218
- Assessing and Mitigating Bias in Medical Artificial Intelligence: The Effects of Race and Ethnicity on a Deep Learning Model for ECG Analysis.Circulation: Arrhythmia and Electrophysiology. 2020; 13e007988
Article info
Publication history
Accepted:
December 25,
2022
Received in revised form:
November 23,
2022
Received:
June 1,
2022
Publication stage
In Press Journal Pre-ProofIdentification
Copyright
© 2023 Elsevier Inc. All rights reserved.
