Deep Learning to Automatically Segment and Analyze Abdominal Aortic Aneurysm from Computed Tomography Angiography.
Abdominal Aortic Aneurysm
Convolutional Neural Network
Multi-view integration
Thrombus segmentation
Journal
Cardiovascular engineering and technology
ISSN: 1869-4098
Titre abrégé: Cardiovasc Eng Technol
Pays: United States
ID NLM: 101531846
Informations de publication
Date de publication:
08 2022
08 2022
Historique:
received:
13
07
2021
accepted:
05
11
2021
pubmed:
9
1
2022
medline:
28
9
2022
entrez:
8
1
2022
Statut:
ppublish
Résumé
Although segmentation of Abdominal Aortic Aneurysms (AAA) thrombus is a crucial step for both the planning of endovascular treatment and the monitoring of the intervention's outcome, it is still performed manually implying time consuming operations as well as operator dependency. The present paper proposes a fully automatic pipeline to segment the intraluminal thrombus in AAA from contrast-enhanced Computed Tomography Angiography (CTA) images and to subsequently analyze AAA geometry. A deep-learning-based pipeline is developed to localize and segment the thrombus from the CTA scans. The thrombus is first identified in the whole sub-sampled CTA, then multi-view U-Nets are combined together to segment the thrombus from the identified region of interest. Polygonal models are generated for the thrombus and the lumen. The lumen centerline is automatically extracted from the lumen mesh and used to compute the aneurysm and lumen diameters. The proposed multi-view integration approach returns an improvement in thrombus segmentation with respect to the single-view prediction. The thrombus segmentation model is trained over a training set of 63 CTA and a validation set of 8 CTA scans. By comparing the thrombus segmentation predicted by the model with the ground truth data, a Dice Similarity Coefficient (DSC) of 0.89 ± 0.04 is achieved. The AAA geometry analysis provided an Intraclass Correlation Coefficient (ICC) of 0.92 and a mean-absolute difference of 3.2 ± 2.4 mm, for the measurements of the total diameter of the aneurysm. Validation of both thrombus segmentation and aneurysm geometry analysis is performed over a test set of 14 CTA scans. The developed deep learning models can effectively segment the thrombus from patients affected by AAA. Moreover, the diameters automatically extracted from the AAA show high correlation with those manually measured by experts.
Identifiants
pubmed: 34997555
doi: 10.1007/s13239-021-00594-z
pii: 10.1007/s13239-021-00594-z
doi:
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
535-547Informations de copyright
© 2021. Biomedical Engineering Society.
Références
Antiga, L., M. Piccinelli, L. Botti, B. Ene-Iordache, A. Remuzzi, and D. A. Steinman. An image-based modeling framework for patient-specific computational hemodynamics. Med. Biol. Eng. Comput. 46(11):1097–1112, 2008. https://doi.org/10.1007/s11517-008-0420-1 .
doi: 10.1007/s11517-008-0420-1
pubmed: 19002516
Arash Salarian, Intraclass Correlation Coefficient (icc). https://www.mathworks.com/matlabcentral/fileexchange/22099-intraclass-correlation-coefficient-icc
Caradu, C., B. Spampinato, A. M. Vrancianu, X. Bérard, and E. Ducasse. Fully automatic volume segmentation of infrarenal abdominal aortic aneurysm computed tomography images with deep learning approaches versus physician controlled manual segmentation. J. Vasc. Surg. 74(1):246-256.e6, 2021. https://doi.org/10.1016/j.jvs.2020.11.036 .
doi: 10.1016/j.jvs.2020.11.036
pubmed: 33309556
Chaikof, E. L., et al. The Society for Vascular Surgery practice guidelines on the care of patients with an abdominal aortic aneurysm. J. Vasc. Surg. 67(1):2-77.e2, 2018. https://doi.org/10.1016/j.jvs.2017.10.044 .
doi: 10.1016/j.jvs.2017.10.044
pubmed: 29268916
Chollet, Francois et al. Keras. 2015. https://keras.io
Developers, TensorFlow. TensorFlow. Zenodo 2021. https://doi.org/10.5281/ZENODO.5043456 .
Fantazzini, A., et al. 3D automatic segmentation of aortic computed tomography angiography combining multi-view 2D Convolutional Neural Networks. Cardiovasc. Eng. Technol. 11(5):576–586, 2020. https://doi.org/10.1007/s13239-020-00481-z .
doi: 10.1007/s13239-020-00481-z
pubmed: 32783134
pmcid: 7511465
Haller, S. J., et al. Intraluminal thrombus is associated with early rupture of abdominal aortic aneurysm. J. Vasc. Surg. 67(4):1051-1058.e1, 2018. https://doi.org/10.1016/j.jvs.2017.08.069 .
doi: 10.1016/j.jvs.2017.08.069
pubmed: 29141786
Hong, H. A. and U. U. Sheikh. Automatic detection, segmentation and classification of abdominal aortic aneurysm using deep learning. In 2016 IEEE 12th International Colloquium on Signal Processing & Its Applications (CSPA), Melaka, Malaysia, Mar. 2016, pp. 242–246. https://doi.org/10.1109/CSPA.2016.7515839 .
Kaladji, A., A. Lucas, G. Kervio, P. Haigron, and A. Cardon. Sizing for endovascular aneurysm repair: clinical evaluation of a new automated three-dimensional software. Ann. Vasc. Surg. 24(7):912–920, 2010. https://doi.org/10.1016/j.avsg.2010.03.018 .
doi: 10.1016/j.avsg.2010.03.018
pubmed: 20831992
pmcid: 3001481
Lareyre, F., C. Adam, M. Carrier, C. Dommerc, C. Mialhe, and J. Raffort. A fully automated pipeline for mining abdominal aortic aneurysm using image segmentation. Sci. Rep. 9(1):13750, 2019. https://doi.org/10.1038/s41598-019-50251-8 .
doi: 10.1038/s41598-019-50251-8
pubmed: 31551507
pmcid: 6760111
López-Linares, K., et al. Fully automatic detection and segmentation of abdominal aortic thrombus in post-operative CTA images using Deep Convolutional Neural Networks. Med. Image Anal. 46:202–214, 2018. https://doi.org/10.1016/j.media.2018.03.010 .
doi: 10.1016/j.media.2018.03.010
pubmed: 29609054
López-Linares, K., I. García, A. García-Familiar, I. Macía, and M. A. G. Ballester. 3D convolutional neural network for abdominal aortic aneurysm segmentation. 2019. https://doi.org/10.1016/j.media.2018.03.010 .
doi: 10.1016/j.media.2018.03.010
Lorensen, W. E., and H. E. Cline. Marching cubes: A high resolution 3D surface construction algorithm. ACM SIGGRAPH Comput. Graph. 21(4):163–169, 1987. https://doi.org/10.1145/37402.37422 .
doi: 10.1145/37402.37422
Maiora, J., B. Ayerdi, and M. Graña. Random forest active learning for AAA thrombus segmentation in computed tomography angiography images. Neurocomputing. 126:71–77, 2014. https://doi.org/10.1016/j.neucom.2013.01.051 .
doi: 10.1016/j.neucom.2013.01.051
Maurer, C. R., R. Qi, and V. Raghavan. A linear time algorithm for computing exact Euclidean distance transforms of binary images in arbitrary dimensions. IEEE Trans. Pattern Anal. Mach. Intell. 25(2):265–270, 2003. https://doi.org/10.1109/TPAMI.2003.1177156 .
doi: 10.1109/TPAMI.2003.1177156
McCormick, M., X. Liu, J. Jomier, C. Marion, and L. Ibanez. ITK: enabling reproducible research and open science. Front. Neuroinform. 2014. https://doi.org/10.3389/fninf.2014.00013 .
doi: 10.3389/fninf.2014.00013
pubmed: 24600387
pmcid: 3929840
Óleary, S. A., E. G. Kavanagh, P. A. Grace, T. M. McGloughlin, and B. J. Doyle. The biaxial mechanical behaviour of abdominal aortic aneurysm intraluminal thrombus: classification of morphology and the determination of layer and region specific properties. J. Biomech. 47(6):1430–1437, 2014. https://doi.org/10.1016/j.jbiomech.2014.01.041 .
doi: 10.1016/j.jbiomech.2014.01.041
Ronneberger, O., P. Fischer, and T. Brox, ‘U-Net: Convolutional Networks for Biomedical Image Segmentation’, ArXiv150504597 Cs, May 2015, doi: https://doi.org/10.1007/978-3-319-24574-4_28 .
Schroeder, W., K. Martin, and B. Lorensen. The visualization toolkit: an object-oriented approach to 3D graphics; visualize data in 3D—medical, engineering or scientific; build your own applications with C++, Tcl, Java or Python; includes source code for VTK (supports Unix, Windows and Mac), 4th ed. Clifton Park, NY: Kitware Inc, 2006.
Shen, D., G. Wu, and H.-I. Suk. Deep learning in medical image analysis. Annu. Rev. Biomed. Eng. 19(1):221–248, 2017. https://doi.org/10.1146/annurev-bioeng-071516-044442 .
doi: 10.1146/annurev-bioeng-071516-044442
pubmed: 28301734
pmcid: 5479722
Singh, K., et al. Intra- and interobserver variability in the measurements of abdominal aortic and common iliac artery diameter with computed tomography. The Tromsø study. Eur. J. Vasc. Endovasc. Surg. 25(5):399–407, 2003. https://doi.org/10.1053/ejvs.2002.1856 .
doi: 10.1053/ejvs.2002.1856
pubmed: 12713777
Wolf, I., et al. The medical imaging interaction toolkit. Med. Image Anal. 9(6):594–604, 2005. https://doi.org/10.1016/j.media.2005.04.005 .
doi: 10.1016/j.media.2005.04.005
pubmed: 15896995
Zhang, Y., Q. Liao, and J. Zhang. Exploring efficient volumetric medical image segmentation using 2.5D method: an empirical study. ArXiv201006163 Cs Eess, 2020. http://arxiv.org/abs/2010.06163