Robustness of deep learning segmentation of cardiac substructures in noncontrast computed tomography for breast cancer radiotherapy.
breast radiotherapy
cardiac substructures
deep learning
image segmentation
model robustness
Journal
Medical physics
ISSN: 2473-4209
Titre abrégé: Med Phys
Pays: United States
ID NLM: 0425746
Informations de publication
Date de publication:
Nov 2021
Nov 2021
Historique:
revised:
19
07
2021
received:
17
12
2021
accepted:
13
09
2021
pubmed:
22
9
2021
medline:
18
11
2021
entrez:
21
9
2021
Statut:
ppublish
Résumé
To develop and evaluate deep learning-based autosegmentation of cardiac substructures from noncontrast planning computed tomography (CT) images in patients undergoing breast cancer radiotherapy and to investigate the algorithm sensitivity to out-of-distribution data such as CT image artifacts. Nine substructures including aortic valve (AV), left anterior descending (LAD), tricuspid valve (TV), mitral valve (MV), pulmonic valve (PV), right atrium (RA), right ventricle (RV), left atrium (LA), and left ventricle (LV) were manually delineated by a radiation oncologist on noncontrast CT images of 129 patients with breast cancer; among them 90 were considered in-distribution data, also named as "clean" data. The image/label pairs of 60 subjects were used to train a 3D deep neural network while the remaining 30 were used for testing. The rest of the 39 patients were considered out-of-distribution ("outlier") data, which were used to test robustness. Random rigid transformations were used to augment the dataset during training. We investigated multiple loss functions, including Dice similarity coefficient (DSC), cross-entropy (CE), Euclidean loss as well as the variation and combinations of these, data augmentation, and network size on overall performance and sensitivity to image artifacts due to infrequent events such as the presence of implanted devices. The predicted label maps were compared to the ground-truth labels via DSC and mean and 90th percentile symmetric surface distance (90th-SSD). When modified Dice combined with cross-entropy (MD-CE) was used as the loss function, the algorithm achieved a mean DSC = 0.79 ± 0.07 for chambers and 0.39 ± 0.10 for smaller substructures (valves and LAD). The mean and 90th-SSD were 2.7 ± 1.4 and 6.5 ± 2.8 mm for chambers and 4.1 ± 1.7 and 8.6 ± 3.2 mm for smaller substructures. Models with MD-CE, Dice-CE, MD, and weighted CE loss had highest performance, and were statistically similar. Data augmentation did not affect model performances on both clean and outlier data and model robustness was susceptible to network size. For a certain type of outlier data, robustness can be improved via incorporating them into the training process. The execution time for segmenting each patient was on an average 2.1 s. A deep neural network provides a fast and accurate segmentation of large cardiac substructures in noncontrast CT images. Model robustness of two types of clinically common outlier data were investigated and potential approaches to improve them were explored. Evaluation of clinical acceptability and integration into clinical workflow are pending.
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
7172-7188Subventions
Organisme : NIH HHS
ID : S10OD025200
Pays : United States
Organisme : NIH HHS
ID : 1S10RR022984-01A1
Pays : United States
Organisme : NIH HHS
ID : 1S10OD018091-01
Pays : United States
Informations de copyright
© 2021 American Association of Physicists in Medicine.
Références
Ng AK. Review of the cardiac long-term effects of therapy for Hodgkin lymphoma. Br J Haematol. 2011;154(1):23-31. https://doi.org/10.1111/j.1365-2141.2011.08713.x
Demirci S, Nam J, Hubbs JL, Nguyen T, Marks LB. Radiation-induced cardiac toxicity after therapy for breast cancer: interaction between treatment era and follow-up duration. Int J Radiat Oncol Biol Phys. 2009;73(4):980-987. https://doi.org/10.1016/j.ijrobp.2008.11.016
Hoppe BS, Bates JE, Mendenhall NP, et al. The meaningless meaning of mean heart dose in mediastinal lymphoma in the modern radiation therapy era. Pract Radiat Oncol. 2020;10(3):e147-e154. https://doi.org/10.1016/j.prro.2019.09.015
van den Bogaard VAB, Ta BDP, van der Schaaf A, et al. Validation and modification of a prediction model for acute cardiac events in patients with breast cancer treated with radiotherapy based on three-dimensional dose distributions to cardiac substructures. J Clin Oncol. 2017;35(11):1171-1178. https://doi.org/10.1200/JCO.2016.69.8480
Thor M, Deasy JO, Hu C, et al. The role of heart-related dose-volume metrics on overall survival in the RTOG 0617 clinical trial. Int J Radiat Oncol. 2018;102(3):S96. https://doi.org/10.1016/j.ijrobp.2018.06.247
McWilliam A, Khalifa J, Vasquez Osorio E, et al. Sens methodology to investigate the impact of radiation dose to heart sub-structures on overall survival. Int J Radiat Oncol. 2020. https://doi.org/10.1016/j.ijrobp.2020.06.031
Thor M, Yorke ED, Jackson A, Deasy JO, Rimner A, Hotca-Cho AE. Successful implementation of Atlas segmentation for cardiac substructures. Int J Radiat Oncol. 2017;99(2):E728. https://doi.org/10.1016/j.ijrobp.2017.06.2354
Zhou R, Liao Z, Pan T, et al. Cardiac atlas development and validation for automatic segmentation of cardiac substructures. Radiother Oncol. 2017;122(1):66-71. https://doi.org/10.1016/j.radonc.2016.11.016
Morris ED, Ghanem AI, Pantelic MV, Walker EM, Han X, Glide-Hurst CK. Cardiac substructure segmentation and dosimetry using a novel hybrid magnetic resonance and computed tomography cardiac Atlas. Int J Radiat Oncol Biol Phys. 2019;103(4):985-993. https://doi.org/10.1016/j.ijrobp.2018.11.025
Kaderka R, Gillespie EF, Mundt RC, et al. Geometric and dosimetric evaluation of atlas based auto-segmentation of cardiac structures in breast cancer patients. Radiother Oncol. 2019;131:215-220. https://doi.org/10.1016/j.radonc.2018.07.013
Morris ED, Ghanem AI, Dong M, Pantelic MV, Walker EM, Glide-Hurst CK. Cardiac substructure segmentation with deep learning for improved cardiac sparing. Med Phys. 2019. https://doi.org/10.1002/mp.13940
Haq R, Hotca A, Apte A, Rimner A, Deasy JO, Thor M. Cardio-pulmonary substructure segmentation of radiotherapy computed tomography images using convolutional neural networks for clinical outcomes analysis. Phys Imag Radiat Oncol. 2020;14:61-66. https://doi.org/10.1016/j.phro.2020.05.009
Harms J, Lei Y, Tian S, et al. Automatic delineation of cardiac substructures using a region-based fully convolutional network. Med Phys. 2021. https://doi.org/10.1002/mp.14810
Karani N, Chaitanya K, Baumgartner C, Konukoglu E. A lifelong learning approach to brain MR segmentation across scanners and protocols. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol 11070 LNCS. Springer Verlag; 2018:476-484. https://doi.org/10.1007/978-3-030-00928-1_54
Udupa JK, McLaughlin DJ, Wu X, et al. Image quality and segmentation. In: Webster RJ, Fei B, eds. Medical Imaging 2018: Image-Guided Procedures, Robotic Interventions, and Modeling. SPIE-Intl Soc Optical Eng; 2018:85. https://doi.org/10.1117/12.2293622
Full PM, Isensee F, Jäger PF, Maier-Hein K. Studying robustness of semantic segmentation under domain shift in cardiac MRI. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 12592 LNCS; 2020:238-249. Accessed April 6, 2021. http://arxiv.org/abs/2011.07592
He K, Zhang X, Ren S, Sun J. Deep residual learning for image recognition. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition. IEEE Computer Society; 2016:770-778. https://doi.org/10.1109/CVPR.2016.90
Ronneberger O, Fischer P, Brox T. U-net: convolutional networks for biomedical image segmentation. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 9351. Springer Verlag; 2015: 234-241. https://doi.org/10.1007/978-3-319-24574-4_28
Sudre CH, Li W, Vercauteren T, Ourselin S, Cardoso MJ. Generalised Dice overlap as a deep learning loss function for highly unbalanced segmentations. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 10553. Springer; 2017: 240-248. https://doi.org/10.1007/978-3-319-67558-9_28
Ben naceur M, Akil M, Saouli R, Kachouri R. Fully automatic brain tumor segmentation with deep learning-based selective attention using overlapping patches and multi-class weighted cross-entropy. Med Image Anal. 2020;63:101692. https://doi.org/10.1016/j.media.2020.101692
Milletari F, Navab N, Ahmadi SA. V-Net: fully convolutional neural networks for volumetric medical image segmentation. Proceedings-2016 4th International Conference on 3D Vision, 3DV 2016. Institute of Electrical and Electronics Engineers Inc.; 2016:565-571. https://doi.org/10.1109/3DV.2016.79
Musgrave K, Belongie S, Lim S-N. A metric learning reality check. 2020. Accessed July 15, 2020. http://arxiv.org/abs/2003.08505
Glorot X, Bengio Y. Understanding the difficulty of training deep feedforward neural networks. J Mach Learn Res. 2010;9:249-256.
Feng M, Moran JM, Koelling T, et al. Development and validation of a heart atlas to study cardiac exposure to radiation following treatment for breast cancer. Int J Radiat Oncol Biol Phys. 2011;79(1):10-18. https://doi.org/10.1016/j.ijrobp.2009.10.058
Wennstig AK, Garmo H, Hållström P, et al. Inter-observer variation in delineating the coronary arteries as organs at risk. Radiother Oncol. 2017;122(1):72-78. https://doi.org/10.1016/j.radonc.2016.11.007
Duane F, Aznar MC, Bartlett F, et al. A cardiac contouring atlas for radiotherapy. Radiother Oncol. 2017;122(3):416-422. https://doi.org/10.1016/j.radonc.2017.01.008
Villeirs GM, Van Vaerenbergh K, Vakaet L, et al. Strahlentherapie und onkologie interobserver delineation variation using CT versus combined CT + MRI in intensity-modulated radiotherapy for prostate cancer. Schlüsselwörter: prostataneoplasien · Strahlentherapie · Computertomographie (CT), vergleichende Studien · Kernspin-tomographie (MRT), vergleichende Studien · Diagnostische Radiologie · Beobachterleistung · Konformale Strahlen therapie. Strahlenther Onkol. 2005;181(7):424-454. https://doi.org/10.1007/s00066-005-1383-x