Deep learning estimation of three-dimensional left atrial shape from two-chamber and four-chamber cardiac long axis views.
cardiovascular magnetic resonance
left atrial volume
machine learning
Journal
European heart journal. Cardiovascular Imaging
ISSN: 2047-2412
Titre abrégé: Eur Heart J Cardiovasc Imaging
Pays: England
ID NLM: 101573788
Informations de publication
Date de publication:
24 04 2023
24 04 2023
Historique:
received:
26
10
2022
revised:
16
12
2022
accepted:
09
01
2023
medline:
26
4
2023
pubmed:
2
2
2023
entrez:
1
2
2023
Statut:
ppublish
Résumé
Left atrial volume is commonly estimated using the bi-plane area-length method from two-chamber (2CH) and four-chamber (4CH) long axes views. However, this can be inaccurate due to a violation of geometric assumptions. We aimed to develop a deep learning neural network to infer 3D left atrial shape, volume and surface area from 2CH and 4CH views. A 3D UNet was trained and tested using 2CH and 4CH segmentations generated from 3D coronary computed tomography angiography (CCTA) segmentations (n = 1700, with 1400/100/200 cases for training/validating/testing). An independent test dataset from another institution was also evaluated, using cardiac magnetic resonance (CMR) 2CH and 4CH segmentations as input and 3D CCTA segmentations as the ground truth (n = 20). For the 200 test cases generated from CCTA, the network achieved a mean Dice score value of 93.7%, showing excellent 3D shape reconstruction from two views compared with the 3D segmentation Dice of 97.4%. The network also showed significantly lower mean absolute error values of 3.5 mL/4.9 cm2 for LA volume/surface area respectively compared to the area-length method errors of 13.0 mL/34.1 cm2 respectively (P < 0.05 for both). For the independent CMR test set, the network achieved accurate 3D shape estimation (mean Dice score value of 87.4%), and a mean absolute error values of 6.0 mL/5.7 cm2 for left atrial volume/surface area respectively, significantly less than the area-length method errors of 14.2 mL/19.3 cm2 respectively (P < 0.05 for both). Compared to the bi-plane area-length method, the network showed higher accuracy and robustness for both volume and surface area.
Identifiants
pubmed: 36725705
pii: 7022885
doi: 10.1093/ehjci/jead010
pmc: PMC10125223
doi:
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
607-615Subventions
Organisme : British Heart Foundation
ID : FS/ICRF/20/26002
Pays : United Kingdom
Organisme : British Heart Foundation
ID : FS/20/26/34952
Pays : United Kingdom
Organisme : British Heart Foundation
ID : CH/09/002/26360
Pays : United Kingdom
Organisme : Medical Research Council
ID : G0701127
Pays : United Kingdom
Organisme : British Heart Foundation
ID : CH/09/002
Pays : United Kingdom
Organisme : Chief Scientist Office
ID : CZH/4/588
Pays : United Kingdom
Informations de copyright
© The Author(s) 2023. Published by Oxford University Press on behalf of the European Society of Cardiology.
Déclaration de conflit d'intérêts
Conflict of interest: RN and KPK are employees of Siemens Healthcare Ltd. All others have no conflicts of interest to disclose.
Références
Int J Cardiovasc Imaging. 2017 Feb;33(2):209-218
pubmed: 27696111
Eur J Heart Fail. 2014 Oct;16(10):1096-103
pubmed: 25138249
Pediatr Cardiol. 2014 Apr;35(4):645-51
pubmed: 24240782
Am J Cardiol. 1999 Oct 1;84(7):829-32
pubmed: 10513783
Heart Rhythm O2. 2021 Nov 05;2(6Part A):622-632
pubmed: 34988507
Front Cardiovasc Med. 2020 Mar 05;7:25
pubmed: 32195270
Heart Rhythm. 2016 May;13(5):1037-1044
pubmed: 26775143
N Engl J Med. 2018 Sep 06;379(10):924-933
pubmed: 30145934
Cathet Cardiovasc Diagn. 1981;7(2):165-78
pubmed: 7296665
Sci Rep. 2020 Feb 12;10(1):2408
pubmed: 32051456
J Cardiovasc Electrophysiol. 2019 Sep;30(9):1416-1427
pubmed: 31111557