Automatic segmentation of multiple cardiovascular structures from cardiac computed tomography angiography images using deep learning.
Journal
PloS one
ISSN: 1932-6203
Titre abrégé: PLoS One
Pays: United States
ID NLM: 101285081
Informations de publication
Date de publication:
2020
2020
Historique:
received:
26
11
2019
accepted:
23
03
2020
entrez:
7
5
2020
pubmed:
7
5
2020
medline:
31
7
2020
Statut:
epublish
Résumé
To develop, demonstrate and evaluate an automated deep learning method for multiple cardiovascular structure segmentation. Segmentation of cardiovascular images is resource-intensive. We design an automated deep learning method for the segmentation of multiple structures from Coronary Computed Tomography Angiography (CCTA) images. Images from a multicenter registry of patients that underwent clinically-indicated CCTA were used. The proximal ascending and descending aorta (PAA, DA), superior and inferior vena cavae (SVC, IVC), pulmonary artery (PA), coronary sinus (CS), right ventricular wall (RVW) and left atrial wall (LAW) were annotated as ground truth. The U-net-derived deep learning model was trained, validated and tested in a 70:20:10 split. The dataset comprised 206 patients, with 5.130 billion pixels. Mean age was 59.9 ± 9.4 yrs., and was 42.7% female. An overall median Dice score of 0.820 (0.782, 0.843) was achieved. Median Dice scores for PAA, DA, SVC, IVC, PA, CS, RVW and LAW were 0.969 (0.979, 0.988), 0.953 (0.955, 0.983), 0.937 (0.934, 0.965), 0.903 (0.897, 0.948), 0.775 (0.724, 0.925), 0.720 (0.642, 0.809), 0.685 (0.631, 0.761) and 0.625 (0.596, 0.749) respectively. Apart from the CS, there were no significant differences in performance between sexes or age groups. An automated deep learning model demonstrated segmentation of multiple cardiovascular structures from CCTA images with reasonable overall accuracy when evaluated on a pixel level.
Sections du résumé
OBJECTIVES
To develop, demonstrate and evaluate an automated deep learning method for multiple cardiovascular structure segmentation.
BACKGROUND
Segmentation of cardiovascular images is resource-intensive. We design an automated deep learning method for the segmentation of multiple structures from Coronary Computed Tomography Angiography (CCTA) images.
METHODS
Images from a multicenter registry of patients that underwent clinically-indicated CCTA were used. The proximal ascending and descending aorta (PAA, DA), superior and inferior vena cavae (SVC, IVC), pulmonary artery (PA), coronary sinus (CS), right ventricular wall (RVW) and left atrial wall (LAW) were annotated as ground truth. The U-net-derived deep learning model was trained, validated and tested in a 70:20:10 split.
RESULTS
The dataset comprised 206 patients, with 5.130 billion pixels. Mean age was 59.9 ± 9.4 yrs., and was 42.7% female. An overall median Dice score of 0.820 (0.782, 0.843) was achieved. Median Dice scores for PAA, DA, SVC, IVC, PA, CS, RVW and LAW were 0.969 (0.979, 0.988), 0.953 (0.955, 0.983), 0.937 (0.934, 0.965), 0.903 (0.897, 0.948), 0.775 (0.724, 0.925), 0.720 (0.642, 0.809), 0.685 (0.631, 0.761) and 0.625 (0.596, 0.749) respectively. Apart from the CS, there were no significant differences in performance between sexes or age groups.
CONCLUSIONS
An automated deep learning model demonstrated segmentation of multiple cardiovascular structures from CCTA images with reasonable overall accuracy when evaluated on a pixel level.
Identifiants
pubmed: 32374784
doi: 10.1371/journal.pone.0232573
pii: PONE-D-19-32919
pmc: PMC7202628
doi:
Banques de données
Dryad
['10.5061/dryad.9s4mw6mc9']
Types de publication
Journal Article
Research Support, Non-U.S. Gov't
Langues
eng
Sous-ensembles de citation
IM
Pagination
e0232573Déclaration de conflit d'intérêts
James K. Min received funding from the Dalio Foundation. Dr. Pontone receives institutional research grant support and is a speaker for Heartflow, Medtronic, GE Healthcare, Bracco Diagnostics, and Bayer Life Sciences. Dr. Jonathon Leipsic is a consultant and has reported stock options with Circle CVI and HeartFlow. Gabriel Maliakal previously worked at Weill Cornell Medicine but is now an employee at Cleerly Health. The work done in this study was during Gabriel Maliakal’s full-time employment at Weill Cornell Medicine. Dr. Min previously worked at Weill Cornell Medicine but is now an employee and has an equity interest in Cleerly Health. The work done in this study was during Dr.Min’s full-time employment at Weill Cornell Medicine. Dr. Shaw has an equity interest in Cleerly Health. This does not alter our adherence to PLOS ONE policies on sharing data and materials. There are no patents, products in development or marketed products to declare.
Références
JAMA. 2018 Sep 18;320(11):1101-1102
pubmed: 30178065
Eur Heart J Cardiovasc Imaging. 2016 Sep;17(9):1009-17
pubmed: 26758412
Acad Radiol. 2009 Aug;16(8):981-7
pubmed: 19394871
Radiographics. 2015 Nov-Dec;35(7):1873-92
pubmed: 26452112
Comput Med Imaging Graph. 2018 Jun;66:90-99
pubmed: 29573583
Radiographics. 2012 Jul-Aug;32(4):991-1008
pubmed: 22786990
Annu Int Conf IEEE Eng Med Biol Soc. 2012;2012:2343-6
pubmed: 23366394
Int J Comput Assist Radiol Surg. 2017 Sep;12(9):1481-1499
pubmed: 28421319
Med Image Anal. 2011 Dec;15(6):863-76
pubmed: 21737337
Eur Heart J. 2019 Jun 21;40(24):1975-1986
pubmed: 30060039
Med Phys. 2015 Jul;42(7):3822-33
pubmed: 26133584
Circ Cardiovasc Imaging. 2018 Jul;11(7):e007562
pubmed: 30012825
AJR Am J Roentgenol. 2017 Apr;208(4):739-749
pubmed: 28026210
Am Heart J. 2016 Dec;182:72-79
pubmed: 27914502
Eur Radiol. 2019 Sep;29(9):4613-4623
pubmed: 30673817
J Cardiovasc Comput Tomogr. 2016 Nov - Dec;10(6):435-449
pubmed: 27780758
Int J Comput Assist Radiol Surg. 2014 Mar;9(2):211-9
pubmed: 23877280
JACC Cardiovasc Imaging. 2020 May;13(5):1163-1171
pubmed: 31607673
Deep Learn Med Image Anal Multimodal Learn Clin Decis Support (2017). 2017;2017:240-248
pubmed: 34104926