Abstract
Segmentation of the mediastinal vasculature in computed tomography (CT) enables automated extraction of important biomarkers for cardiopulmonary disease characterization and outcome prediction. However, the limited contrast between blood and surrounding soft tissue makes manual segmentation of mediastinal structures challenging in non-contrast CT (NCCT) images, resulting in limited annotations for training deep learning models. To overcome this challenge, we propose a semi-supervised mediastinal vasculature segmentation method that utilizes knowledge distillation from unlabeled training data of contrast-enhanced dual-energy CT to achieve segmentation of the main pulmonary artery, main pulmonary veins, and aorta in NCCT. Our framework incorporates multitask learning with attention feature fusion bridges for online knowledge transfer from a related image-to-image translation task to the target segmentation task. Experimental evaluations demonstrate superior segmentation accuracy of our approach compared to fully supervised methods as well as two sequential approaches that do not leverage distillation between tasks. The proposed approach achieves a Dice similarity coefficient of 0.871 for the main pulmonary artery, 0.920 for the aorta, and 0.824 for the main pulmonary veins. By leveraging a large dataset without annotations through multitask learning and knowledge distillation, our approach improves performance in the target task of mediastinal segmentation with limited annotated training data.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Cardoso, M.J., et al.: MONAI: an open-source framework for deep learning in healthcare. arXiv preprint arXiv:2211.02701 (2022)
Chaudhary, M.F., et al.: Lung2Lung: volumetric style transfer with self-ensembling for high-resolution cross-volume computed tomography. arXiv preprint arXiv:2210.02625 (2022)
Dai, Y., Gieseke, F., Oehmcke, S., Wu, Y., Barnard, K.: Attentional feature fusion. In: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), pp. 3560–3569 (2021)
Fuld, M.K., Halaweish, A.F., Haynes, S.E., Divekar, A.A., Guo, J., Hoffman, E.A.: Pulmonary perfused blood volume with dual-energy CT as surrogate for pulmonary perfusion assessed with dynamic multidetector CT. Radiology 267(3), 747–756 (2013)
Gerard, S.E., Herrmann, J., Kaczka, D.W., Musch, G., Fernandez-Bustamante, A., Reinhardt, J.M.: Multi-resolution convolutional neural networks for fully automated segmentation of acutely injured lungs in multiple species. Med. Image Anal. 60, 101592 (2020)
Hagan, J.B.: Anaphylactoid and adverse reactions to radiocontrast agents. Immunol. Allergy Clin. 24(3), 507–519 (2004)
Hatamizadeh, A., et al.: UNETR: transformers for 3D medical image segmentation. In: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), pp. 574–584 (2022)
Heimann, T., et al.: Comparison and evaluation of methods for liver segmentation from CT datasets. IEEE Trans. Med. Imaging 28(8), 1251–1265 (2009)
Hermann, E.A., et al.: Pulmonary blood volume among older adults in the community: the MESA lung study. Circul. Cardiovas. Imaging 15(8), e014380 (2022)
Hu, T., et al.: Aorta-aware GAN for non-contrast to artery contrasted CT translation and its application to abdominal aortic aneurysm detection. Int. J. Comput. Assist. Radiol. Surg. 1–9 (2022)
Isola, P., Zhu, J.Y., Zhou, T., Efros, A.A.: Image-to-image translation with conditional adversarial networks. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 1125–1134 (2017)
Iyer, A.S., Wells, J.M., Vishin, S., Bhatt, S.P., Wille, K.M., Dransfield, M.T.: CT scan-measured pulmonary artery to aorta ratio and echocardiography for detecting pulmonary hypertension in severe COPD. Chest 145(4), 824–832 (2014)
Lin, T.Y., Goyal, P., Girshick, R., He, K., Dollár, P.: Focal loss for dense object detection. In: Proceedings of the IEEE International Conference on Computer Vision (ICCV), pp. 2980–2988 (2017)
Liu, J., et al.: DyeFreeNet: deep virtual contrast CT synthesis. In: Burgos, N., Svoboda, D., Wolterink, J.M., Zhao, C. (eds.) SASHIMI 2020. LNCS, vol. 12417, pp. 80–89. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-59520-3_9
Liu, Y., et al.: An incentive-based program coupled with sildenafil provides enhanced success of smoking cessation associated with an accelerated loss of CT assessed smoking-associated lung density (inflammation) and improved DLCO. In: D76. COPD: Clinical Studies, pp. A7556–A7556. American Thoracic Society (2020)
Liu, Z., et al.: Swin transformer: hierarchical vision transformer using shifted windows. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), pp. 10012–10022 (2021)
Milletari, F., Navab, N., Ahmadi, S.A.: V-Net: fully convolutional neural networks for volumetric medical image segmentation. In: 2016 Fourth International Conference on 3D Vision (3DV), pp. 565–571. IEEE (2016)
Pang, H., et al.: NCCT-CECT image synthesizers and their application to pulmonary vessel segmentation. Comput. Methods Prog. Biomed. 231, 107389 (2023)
Ristea, N.C., et al.: CyTran: a cycle-consistent transformer with multi-level consistency for non-contrast to contrast CT translation. Neurocomputing 538, 126211 (2023)
Salehi, S.S.M., Erdogmus, D., Gholipour, A.: Tversky loss function for image segmentation using 3D fully convolutional deep networks. In: Wang, Q., Shi, Y., Suk, H.-I., Suzuki, K. (eds.) MLMI 2017. LNCS, vol. 10541, pp. 379–387. Springer, Cham (2017). https://doi.org/10.1007/978-3-319-67389-9_44
Wang, H.J., et al.: Automated 3D segmentation of the aorta and pulmonary artery on non-contrast-enhanced chest computed tomography images in lung cancer patients. Diagnostics 12(4), 967 (2022)
Wells, J.M., et al.: Pulmonary arterial enlargement and acute exacerbations of COPD. N. Engl. J. Med. 367(10), 913–921 (2012)
Zhang, R., Isola, P., Efros, A.A., Shechtman, E., Wang, O.: The unreasonable effectiveness of deep features as a perceptual metric. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 586–595 (2018)
Acknowledgements
The authors thank the other investigators, the staff, and the participants of the Multi-Ethnic Study of Atherosclerosis (MESA) for their valuable contributions. A full list of participating MESA investigators and institutions can be found www.mesa-nhlbi.org. This work was supported by NIH/NHLBI R01-HL077612, R01-HL093081, R01-HL121270, and R01-HL142028. MESA and the MESA SHARe project are conducted and supported by the National Heart, Lung, and Blood Institute (NHLBI) in collaboration with MESA investigators. Support for MESA is provided by contracts HHSN268201500003I, N01-HC-95159-69, UL1-TR-000040, UL1-TR-001079, UL1-TR-001420, UL1-TR-001881, and DK063491.
Author information
Authors and Affiliations
Corresponding authors
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2024 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Chaudhary, M.F.A., Hosseini, S.S., Barr, R.G., Reinhardt, J.M., Hoffman, E.A., Gerard, S.E. (2024). Bridging the Task Barriers: Online Knowledge Distillation Across Tasks for Semi-supervised Mediastinal Segmentation in CT. In: Cao, X., Xu, X., Rekik, I., Cui, Z., Ouyang, X. (eds) Machine Learning in Medical Imaging. MLMI 2023. Lecture Notes in Computer Science, vol 14348. Springer, Cham. https://doi.org/10.1007/978-3-031-45673-2_31
Download citation
DOI: https://doi.org/10.1007/978-3-031-45673-2_31
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-45672-5
Online ISBN: 978-3-031-45673-2
eBook Packages: Computer ScienceComputer Science (R0)
