Accuracy of Deep Learning Echocardiographic View Classification in Patients with Congenital or Structural Heart Disease: Importance of Specific Datasets
Introduction: Automated echocardiography image interpretation has the potential to transform clinical practice. However, neural networks developed in general cohorts may underperform in the setting of altered cardiac anatomy. Methods: Consecutive echocardiographic studies of patients with congenital...
Main Authors: | , , , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2022-01-01
|
Series: | Journal of Clinical Medicine |
Subjects: | |
Online Access: | https://www.mdpi.com/2077-0383/11/3/690 |
_version_ | 1797486911297683456 |
---|---|
author | Felix K. Wegner Maria L. Benesch Vidal Philipp Niehues Kevin Willy Robert M. Radke Philipp D. Garthe Lars Eckardt Helmut Baumgartner Gerhard-Paul Diller Stefan Orwat |
author_facet | Felix K. Wegner Maria L. Benesch Vidal Philipp Niehues Kevin Willy Robert M. Radke Philipp D. Garthe Lars Eckardt Helmut Baumgartner Gerhard-Paul Diller Stefan Orwat |
author_sort | Felix K. Wegner |
collection | DOAJ |
description | Introduction: Automated echocardiography image interpretation has the potential to transform clinical practice. However, neural networks developed in general cohorts may underperform in the setting of altered cardiac anatomy. Methods: Consecutive echocardiographic studies of patients with congenital or structural heart disease (C/SHD) were used to validate an existing convolutional neural network trained on 14,035 echocardiograms for automated view classification. In addition, a new convolutional neural network for view classification was trained and tested specifically in patients with C/SHD. Results: Overall, 9793 imaging files from 262 patients with C/SHD (mean age 49 years, 60% male) and 62 normal controls (mean age 45 years, 50.0% male) were included. Congenital diagnoses included among others, tetralogy of Fallot (30), Ebstein anomaly (18) and transposition of the great arteries (TGA, 48). Assessing correct view classification based on 284,250 individual frames revealed that the non-congenital model had an overall accuracy of 48.3% for correct view classification in patients with C/SHD compared to 66.7% in patients without cardiac disease. Our newly trained convolutional network for echocardiographic view detection based on over 139,910 frames and tested on 35,614 frames from C/SHD patients achieved an accuracy of 76.1% in detecting the correct echocardiographic view. Conclusions: The current study is the first to validate view classification by neural networks in C/SHD patients. While generic models have acceptable accuracy in general cardiology patients, the quality of image classification is only modest in patients with C/SHD. In contrast, our model trained in C/SHD achieved a considerably increased accuracy in this particular cohort. |
first_indexed | 2024-03-09T23:41:05Z |
format | Article |
id | doaj.art-9ac39eada9b44b12a554a8a2116a3f32 |
institution | Directory Open Access Journal |
issn | 2077-0383 |
language | English |
last_indexed | 2024-03-09T23:41:05Z |
publishDate | 2022-01-01 |
publisher | MDPI AG |
record_format | Article |
series | Journal of Clinical Medicine |
spelling | doaj.art-9ac39eada9b44b12a554a8a2116a3f322023-11-23T16:52:36ZengMDPI AGJournal of Clinical Medicine2077-03832022-01-0111369010.3390/jcm11030690Accuracy of Deep Learning Echocardiographic View Classification in Patients with Congenital or Structural Heart Disease: Importance of Specific DatasetsFelix K. Wegner0Maria L. Benesch Vidal1Philipp Niehues2Kevin Willy3Robert M. Radke4Philipp D. Garthe5Lars Eckardt6Helmut Baumgartner7Gerhard-Paul Diller8Stefan Orwat9Department of Cardiology II—Electrophysiology, University Hospital Muenster, Albert-Schweitzer-Campus 1, 48149 Muenster, GermanyDepartment of Cardiology III—Adult Congenital and Valvular Heart Disease, University Hospital Muenster, Albert-Schweitzer-Campus 1, 48149 Muenster, GermanyDepartment of Cardiology II—Electrophysiology, University Hospital Muenster, Albert-Schweitzer-Campus 1, 48149 Muenster, GermanyDepartment of Cardiology II—Electrophysiology, University Hospital Muenster, Albert-Schweitzer-Campus 1, 48149 Muenster, GermanyDepartment of Cardiology III—Adult Congenital and Valvular Heart Disease, University Hospital Muenster, Albert-Schweitzer-Campus 1, 48149 Muenster, GermanyDepartment of Cardiology III—Adult Congenital and Valvular Heart Disease, University Hospital Muenster, Albert-Schweitzer-Campus 1, 48149 Muenster, GermanyDepartment of Cardiology II—Electrophysiology, University Hospital Muenster, Albert-Schweitzer-Campus 1, 48149 Muenster, GermanyDepartment of Cardiology III—Adult Congenital and Valvular Heart Disease, University Hospital Muenster, Albert-Schweitzer-Campus 1, 48149 Muenster, GermanyDepartment of Cardiology III—Adult Congenital and Valvular Heart Disease, University Hospital Muenster, Albert-Schweitzer-Campus 1, 48149 Muenster, GermanyDepartment of Cardiology III—Adult Congenital and Valvular Heart Disease, University Hospital Muenster, Albert-Schweitzer-Campus 1, 48149 Muenster, GermanyIntroduction: Automated echocardiography image interpretation has the potential to transform clinical practice. However, neural networks developed in general cohorts may underperform in the setting of altered cardiac anatomy. Methods: Consecutive echocardiographic studies of patients with congenital or structural heart disease (C/SHD) were used to validate an existing convolutional neural network trained on 14,035 echocardiograms for automated view classification. In addition, a new convolutional neural network for view classification was trained and tested specifically in patients with C/SHD. Results: Overall, 9793 imaging files from 262 patients with C/SHD (mean age 49 years, 60% male) and 62 normal controls (mean age 45 years, 50.0% male) were included. Congenital diagnoses included among others, tetralogy of Fallot (30), Ebstein anomaly (18) and transposition of the great arteries (TGA, 48). Assessing correct view classification based on 284,250 individual frames revealed that the non-congenital model had an overall accuracy of 48.3% for correct view classification in patients with C/SHD compared to 66.7% in patients without cardiac disease. Our newly trained convolutional network for echocardiographic view detection based on over 139,910 frames and tested on 35,614 frames from C/SHD patients achieved an accuracy of 76.1% in detecting the correct echocardiographic view. Conclusions: The current study is the first to validate view classification by neural networks in C/SHD patients. While generic models have acceptable accuracy in general cardiology patients, the quality of image classification is only modest in patients with C/SHD. In contrast, our model trained in C/SHD achieved a considerably increased accuracy in this particular cohort.https://www.mdpi.com/2077-0383/11/3/690deep learningneural networkstructural heart diseasecongenital heart diseaseechocardiography |
spellingShingle | Felix K. Wegner Maria L. Benesch Vidal Philipp Niehues Kevin Willy Robert M. Radke Philipp D. Garthe Lars Eckardt Helmut Baumgartner Gerhard-Paul Diller Stefan Orwat Accuracy of Deep Learning Echocardiographic View Classification in Patients with Congenital or Structural Heart Disease: Importance of Specific Datasets Journal of Clinical Medicine deep learning neural network structural heart disease congenital heart disease echocardiography |
title | Accuracy of Deep Learning Echocardiographic View Classification in Patients with Congenital or Structural Heart Disease: Importance of Specific Datasets |
title_full | Accuracy of Deep Learning Echocardiographic View Classification in Patients with Congenital or Structural Heart Disease: Importance of Specific Datasets |
title_fullStr | Accuracy of Deep Learning Echocardiographic View Classification in Patients with Congenital or Structural Heart Disease: Importance of Specific Datasets |
title_full_unstemmed | Accuracy of Deep Learning Echocardiographic View Classification in Patients with Congenital or Structural Heart Disease: Importance of Specific Datasets |
title_short | Accuracy of Deep Learning Echocardiographic View Classification in Patients with Congenital or Structural Heart Disease: Importance of Specific Datasets |
title_sort | accuracy of deep learning echocardiographic view classification in patients with congenital or structural heart disease importance of specific datasets |
topic | deep learning neural network structural heart disease congenital heart disease echocardiography |
url | https://www.mdpi.com/2077-0383/11/3/690 |
work_keys_str_mv | AT felixkwegner accuracyofdeeplearningechocardiographicviewclassificationinpatientswithcongenitalorstructuralheartdiseaseimportanceofspecificdatasets AT marialbeneschvidal accuracyofdeeplearningechocardiographicviewclassificationinpatientswithcongenitalorstructuralheartdiseaseimportanceofspecificdatasets AT philippniehues accuracyofdeeplearningechocardiographicviewclassificationinpatientswithcongenitalorstructuralheartdiseaseimportanceofspecificdatasets AT kevinwilly accuracyofdeeplearningechocardiographicviewclassificationinpatientswithcongenitalorstructuralheartdiseaseimportanceofspecificdatasets AT robertmradke accuracyofdeeplearningechocardiographicviewclassificationinpatientswithcongenitalorstructuralheartdiseaseimportanceofspecificdatasets AT philippdgarthe accuracyofdeeplearningechocardiographicviewclassificationinpatientswithcongenitalorstructuralheartdiseaseimportanceofspecificdatasets AT larseckardt accuracyofdeeplearningechocardiographicviewclassificationinpatientswithcongenitalorstructuralheartdiseaseimportanceofspecificdatasets AT helmutbaumgartner accuracyofdeeplearningechocardiographicviewclassificationinpatientswithcongenitalorstructuralheartdiseaseimportanceofspecificdatasets AT gerhardpauldiller accuracyofdeeplearningechocardiographicviewclassificationinpatientswithcongenitalorstructuralheartdiseaseimportanceofspecificdatasets AT stefanorwat accuracyofdeeplearningechocardiographicviewclassificationinpatientswithcongenitalorstructuralheartdiseaseimportanceofspecificdatasets |