Abstract
Background
To accurately diagnose and control heart failure (HF), it is important to carry out
a simple assessment of elevated pulmonary arterial wedge pressure (PAWP). The aim
of this study was to develop and validate an objective method for detecting elevated
PAWP by applying deep learning (DL) to a chest x-ray (CXR).
Methods
We enrolled 1013 consecutive patients with a right-heart catheter between October
2009 and February 2020. We developed a convolutional neural network to identify patients
with elevated PAWP (> 18 mm Hg) as the actual value of PAWP to be used in the dataset
for training. In the prospective validation dataset used to detect elevated PAWP,
the area under the receiver operating characteristic curve (AUC) was calculated using
the DL model that evaluated the CXR.
Results
In the prospective validation dataset, the AUC of the DL model with CXR was not significantly
different from the AUC produced by brain natriuretic peptide (BNP) and the echocardiographic
left-ventricular diastolic dysfunction (DD) algorithm (DL model: 0.77 vs BNP: 0.77
vs DD algorithm: 0.70; respectively; P = NS for all comparisons); it was, however, significantly higher than the AUC of
the cardiothoracic ratio (DL model vs cardiothoracic ratio [CTR]: 0.66, P = 0.044). The model based on 3 parameters (BNP, DD algorithm, and CTR) was improved
by adding the DL model (AUC: from 0.80 to 0.86; P = 0.041).
Conclusions
Applying the DL model based on a CXR (a classical, universal, and low-cost test) is
useful for screening for elevated PAWP.
Résumé
Contexte
Pour diagnostiquer et contrôler avec précision l'insuffisance cardiaque (IC), il est
important de procéder à un contrôle simple d'une pression artérielle pulmonaire d'occlusion
(PAPO) élevée. Le but de cette étude était de développer et de valider une méthode
objective pour détecter une pression artérielle pulmonaire élevée en appliquant l'apprentissage
profond (AP) à une radiographie du thorax (RT).
Méthodes
Entre octobre 2009 et février 2020, nous avons recruté 1 013 patients rencontrés consécutivement
avec cathétérisme du cœur droit. Nous avons développé un réseau neuronal convolutif
pour identifier les patients présentant une PAPO élevée (> 18 mmHg) établie en tant
que valeur réelle de la PAPO à utiliser dans le jeu de données dédié à l'apprentissage.
Dans le jeu de données à but de validation prospective utilisé pour détecter une PAPO
élevée, l'aire sous la courbe (ASC) de la fonction d’efficacité du récepteur a été
calculée à l'aide du modèle d'AP qui a évalué la RT.
Résultats
Dans le jeu de données de validation prospective, l'ASC du modèle d'AP avec RT n'était
pas significativement différente de l'ASC produite par le peptide natriurétique de
type B (BNP) et l'algorithme échocardiographique de la dysfonction diastolique (DD)
du ventricule gauche (modèle d'AP : 0,77 vs BNP : 0,77 vs algorithme DD : 0,70; respectivement;
P = NS pour toutes les comparaisons); elle était toutefois significativement plus élevée
que l'ASC du rapport cardiothoracique (modèle d'AP vs rapport cardiothoracique [RCT]
: 0,66, P = 0,044). Le modèle basé sur 3 paramètres (BNP, algorithme DD et RCT) a été amélioré
en ajoutant le modèle d'AP (ASC : de 0,80 à 0,86; P = 0,041).
Conclusions
L'application du modèle d'AP basé sur une RT (un test classique, universel et peu
coûteux) est utile pour le dépistage d'une PAPO élevée.
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 Canadian Journal of CardiologyAlready a print subscriber? Claim online access
Already an online subscriber? Sign in
Register: Create an account
Institutional Access: Sign in to ScienceDirect
References
- ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure 2008: the Task Force for the diagnosis and treatment of acute and chronic heart failure 2008 of the European Society of Cardiology. Developed in collaboration with the Heart Failure Association of the ESC (HFA) and endorsed by the European Society of Intensive Care Medicine (ESICM).Eur Heart J. 2008; 29: 2388-2442
- Economic burden of heart failure in the elderly.Pharmacoeconomics. 2008; 26: 447-462
- A simple derived prediction score for the identification of an elevated pulmonary artery wedge pressure using precatheterization clinical data in patients referred to a pulmonary hypertension center.Chest. 2016; 149: 1261-1268
- ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure 2012: the Task Force for the diagnosis and treatment of acute and chronic heart failure 2012 of the European Society of Cardiology. Developed in collaboration with the Heart Failure Association (HFA) of the ESC.Eur Heart J. 2012; 33: 1787-1847
- Echocardiographic evaluation of left ventricular filling pressure in patients with heart failure with preserved ejection fraction: usefulness of inferior vena cava measurements and 2016 EACVI/ASE recommendations.J Card Fail. 2020; 26: 507-514
- 2016 ESC guidelines for the diagnosis and treatment of acute and chronic heart failure: the Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC). Developed with the special contribution of the Heart Failure Association (HFA) of the ESC.Eur J Heart Fail. 2016; 18: 891-975
- The role of the chest X-ray in the evaluation of chronic severe heart failure: things are not always as they appear.Clin Cardiol. 1988; 11: 486-488
- Diagnosing and grading heart failure with tomographic perfusion lung scintigraphy: validation with right heart catheterization.ESC Heart Fail. 2018; 5: 902-910
- Recommendations for the evaluation of left ventricular diastolic function by echocardiography: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging.J Am Soc Echocardiogr. 2016; 29: 277-314
- Recommendations for the evaluation of left ventricular diastolic function by echocardiography: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging.Eur Heart J Cardiovasc Imaging. 2016; 17: 1321-1360
- Application of deep learning for fast detection of COVID-19 in x-rays using nCOVnet.Chaos Solitons Fractals. 2020; 138: 109944
- Implementation of a deep learning-based computer-aided detection system for the interpretation of chest radiographs in patients suspected for COVID-19.Korean J Radiol. 2020; 21: 1150-1160
- Deep learning.Nature. 2015; 521: 436-444
- Prediction of pulmonary to systemic flow ratio in patients with congenital heart disease using deep learning-based analysis of chest radiographs.JAMA Cardiol. 2020; 5: 449-457
- Prognostic importance of elevated jugular venous pressure and a third heart sound in patients with heart failure.N Engl J Med. 2001; 345: 574-581
- The use of diuretics in heart failure with congestion: a position statement from the Heart Failure Association of the European Society of Cardiology.Eur J Heart Fail. 2019; 21: 137-155
- Recommendations for cardiac chamber quantification by echocardiography in adults: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging.Eur Heart J Cardiovasc Imaging. 2015; 16: 233-270
- Adam: a method for stochastic optimization.arXiv. 2014; 1412: 6980
- Optuna: a next-generation hyperparameter optimization framework.Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. 2019; : 2623-2631
- Grad-cam: Visual explanations from deep networks via gradient-based localization.Proceedings of the IEEE International Conference on Computer Vision. 2017; : 618-626
- Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach.Biometrics. 1988; 44: 837-845
- The limited reliability of physical signs for estimating hemodynamics in chronic heart failure.JAMA. 1989; 261: 884-888
- 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).Eur Heart J. 2016; 37: 67-119
- Deep learning to predict elevated pulmonary artery pressure in patients with suspected pulmonary hypertension using standard chest x ray.Sci Rep. 2020; 10: 19311
- 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines.J Am Coll Cardiol. 2013; 62: e147-239
- Echo-Doppler estimation of left ventricular filling pressure: results of the multicentre EACVI Euro-Filling study.Eur Heart J Cardiovasc Imaging. 2017; 18: 961-968
- Estimating left ventricular filling pressure by echocardiography.J Am Coll Cardiol. 2017; 69: 1937-1948
- Novel parameter to predict left ventricular filling pressure: product of log B-type natriuretic peptide and ratio of mitral inflow early and late diastolic filling velocities.J Card Fail. 2018; 24: 29-32
- 2017 comprehensive update of the canadian cardiovascular society guidelines for the management of heart failure.Can J Cardiol. 2017; 33: 1342-1433
Article info
Publication history
Published online: February 17, 2021
Accepted:
February 11,
2021
Received:
November 4,
2020
Footnotes
See editorial by Adams and Haddad, pages 1153–1155 of this issue.
See page 1205 for disclosure information.
Identification
Copyright
© 2021 Canadian Cardiovascular Society. Published by Elsevier Inc. All rights reserved.
ScienceDirect
Access this article on ScienceDirectLinked Article
- Artificial Intelligence to Diagnose Heart Failure Based on Chest X-Rays and Potential Clinical ImplicationsCanadian Journal of CardiologyVol. 37Issue 8
- PreviewHeart failure (HF) is the most rapidly increasing cardiovascular disease globally and is associated with high rates of morbidity and mortality and high burden on health systems.1 HF is a clinical syndrome defined hemodynamically by the inability of the heart to pump blood at a rate commensurate with metabolic demands or the ability of the heart to meet metabolic demands only with high filling pressures.2 There remains no single diagnostic test for the diagnosis of HF, and history, physical examination, laboratory markers, and imaging are integrated to establish a diagnosis and guide clinical management.
- Full-Text
- Preview
