Accurate assessment of the aorta is critical for diagnosing and managing life-threatening conditions such as aneurysms, dissections, and connective tissue disorders. Manual measurements are prone to variability and are time-consuming. We propose a modular three-stage deep-learning pipeline for the automated characterization of the thoracic aorta and aortic root on computed tomography. The first stage detects regions of interest. The second stage uses a semantic segmentation module to isolate the thoracic aorta and extract the maximal diameter. The third stage employs a multi-task network to segment the aortic root and localize key landmarks, enabling precise measurement of the maximal aortic root diameter. Our method achieves mean Dice scores of 0.94 (thoracic aorta) and 0.96 (aortic root), a mean landmark localization error of 1.69 mm, and total inference times under 50 s on consumer-grade hardware. In a study involving three expert observers (30 cases), Bland–Altman and intra-class correlation analyses demonstrate that our tool yields measurements comparable to expert annotations (ICC ). The pipeline{\textquoteright}s low computational footprint, anatomical focus on the aortic root, strong reproducibility, and high accuracy make it well suited for integration into diagnostic and pre-interventional workflows.
Giordano, L, Ceranka, J, De Sutter, S, Tanaka, K, Van Gompel, G, Lenaerts, T & Vandemeulebroucke, J 2026, A Modular Deep-Learning Pipeline for Automated Aorta Characterization on CT. in S Wu, B Shabestari & L Xing (eds), Applications of Medical Artificial Intelligence - 4th International Workshop, AMAI 2025, Held in Conjunction with MICCAI 2025, Proceedings. vol. 16206, Lecture Notes in Computer Science, vol. 16206 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 258-267, 4th International Workshop on Applications of Medical Artificial Intelligence, AMAI 2025 held in conjunction with the 28th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2025, Daejeon, Korea, Republic of, 23/09/25. https://doi.org/10.1007/978-3-032-09569-5_26, https://doi.org/10.1007/978-3-032-09569-5_26
Giordano, L., Ceranka, J., De Sutter, S., Tanaka, K., Van Gompel, G., Lenaerts, T., & Vandemeulebroucke, J. (2026). A Modular Deep-Learning Pipeline for Automated Aorta Characterization on CT. In S. Wu, B. Shabestari, & L. Xing (Eds.), Applications of Medical Artificial Intelligence - 4th International Workshop, AMAI 2025, Held in Conjunction with MICCAI 2025, Proceedings (Vol. 16206, pp. 258-267). (Lecture Notes in Computer Science; Vol. 16206 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-032-09569-5_26, https://doi.org/10.1007/978-3-032-09569-5_26
@inproceedings{ceb3c314a7ab413f9338698620535555,
title = "A Modular Deep-Learning Pipeline for Automated Aorta Characterization on CT",
abstract = "Accurate assessment of the aorta is critical for diagnosing and managing life-threatening conditions such as aneurysms, dissections, and connective tissue disorders. Manual measurements are prone to variability and are time-consuming. We propose a modular three-stage deep-learning pipeline for the automated characterization of the thoracic aorta and aortic root on computed tomography. The first stage detects regions of interest. The second stage uses a semantic segmentation module to isolate the thoracic aorta and extract the maximal diameter. The third stage employs a multi-task network to segment the aortic root and localize key landmarks, enabling precise measurement of the maximal aortic root diameter. Our method achieves mean Dice scores of 0.94 (thoracic aorta) and 0.96 (aortic root), a mean landmark localization error of 1.69 mm, and total inference times under 50 s on consumer-grade hardware. In a study involving three expert observers (30 cases), Bland–Altman and intra-class correlation analyses demonstrate that our tool yields measurements comparable to expert annotations (ICC ). The pipeline{\textquoteright}s low computational footprint, anatomical focus on the aortic root, strong reproducibility, and high accuracy make it well suited for integration into diagnostic and pre-interventional workflows.",
keywords = "Aorta, Computed tomography, Deep learning, Image processing, Inter-observer variability",
author = "Loris Giordano and Jakub Ceranka and {De Sutter}, Selene and Kaoru Tanaka and {Van Gompel}, Gert and Tom Lenaerts and Jef Vandemeulebroucke",
note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2026.; 4th International Workshop on Applications of Medical Artificial Intelligence, AMAI 2025 held in conjunction with the 28th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2025 ; Conference date: 23-09-2025 Through 23-09-2025",
year = "2026",
month = jan,
day = "2",
doi = "10.1007/978-3-032-09569-5_26",
language = "English",
isbn = "978-3-032-09568-8",
volume = "16206",
series = "Lecture Notes in Computer Science",
publisher = "Springer Science and Business Media Deutschland GmbH",
pages = "258--267",
editor = "Shandong Wu and Behrouz Shabestari and Lei Xing",
booktitle = "Applications of Medical Artificial Intelligence - 4th International Workshop, AMAI 2025, Held in Conjunction with MICCAI 2025, Proceedings",
address = "Germany",
}