Multi-Class Segmentation of Aortic Branches and Zones in Computed Tomography Angiography: The AortaSeg24 Challenge

Standard

Multi-Class Segmentation of Aortic Branches and Zones in Computed Tomography Angiography: The AortaSeg24 Challenge. / The AortaSeg24 challenge.

In: Medical Image Analysis, Vol. 113, 104188, 09.2026.

Research output: Contribution to journal › Article › peer-review

BibTeX

@article{4f78f50cf71146a3ae71eb1811d6ec95,

title = "Multi-Class Segmentation of Aortic Branches and Zones in Computed Tomography Angiography: The AortaSeg24 Challenge",

abstract = "Multi-class segmentation of the aorta in computed tomography angiography (CTA) scans is essential for diagnosing and planning complex endovascular treatments for patients with aortic dissections. However, existing methods reduce aortic segmentation to a binary problem, limiting their ability to measure diameters across different branches and zones. Furthermore, no open-source dataset is currently available to support the development of multi-class aortic segmentation methods. To address this gap, we organized the AortaSeg24 MICCAI Challenge, introducing the first dataset of 100 CTA volumes annotated for 23 clinically relevant aortic branches and zones. This dataset was designed to facilitate both model development and validation. The challenge attracted 121 teams worldwide, with participants leveraging state-of-the-art frameworks such as nnU-Net and exploring novel techniques, including cascaded models, data augmentation strategies, and custom loss functions. We evaluated the submitted algorithms using the Dice Similarity Coefficient (DSC) and Normalized Surface Distance (NSD), highlighting the approaches adopted by the top five performing teams. This paper presents the challenge design, dataset details, evaluation metrics, and an in-depth analysis of the top-performing algorithms. The annotated dataset, evaluation code, and implementations of the leading methods are publicly available to support further research. All resources can be accessed at https://aortaseg24.grand-challenge.org.",

keywords = "Aorta segmentation, CTA, AortaSeg24, 3D segmentation",

author = "{The AortaSeg24 challenge} and Muhammad Imran and Krebs, {Jonathan R.} and Sivaraman, {Vishal Balaji} and Teng Zhang and Amarjeet Kumar and Ueland, {Walker R.} and Fassler, {Michael J.} and Jinlong Huang and Xiao Sun and Lisheng Wang and Pengcheng Shi and Maximilian Rokuss and Michael Baumgartner and Yannick Kirchhof and Maier-Hein, {Klaus H.} and Fabian Isensee and Shuolin Liu and Bing Han and Nguyen, {Bong Thanh} and Dong-jin Shin and Park Ji-Woo and Mathew Choi and Kwang-Hyun Uhm and Sung-Jea Ko and Chanwoong Lee and Jaehee Chun and Kim, {Jin Sung} and Minghui Zhang and Hanxiao Zhang and Xin You and Yun Gu and Zhaohong Pan and Xuan Liu and Xiaokun Liang and Markus Tiefenthaler and Enrique Almar-Munoz and Matthias Schwab and Mikhail Kotyushev and Rostislav Epifanov and Marek Wodzinski and Henning Muller and Abdul Qayyum and Moona Mazher and Niederer, {Steven A.} and Zhiwei Wang and Kaixiang Yang and Jintao Ren and Korreman, {Stine Sofia} and Yuchong Gao and Hongye Zeng",

note = "This work was supported by the Department of Medicine and the Intelligent Clinical Care Center at the University of Florida College of Medicine.",

year = "2026",

month = sep,

doi = "10.1016/j.media.2026.104188",

language = "English",

volume = "113",

journal = "Medical Image Analysis",

issn = "1361-8415",

publisher = "Elsevier Science Publishing Company, Inc.",

}

RIS

TY - JOUR

T1 - Multi-Class Segmentation of Aortic Branches and Zones in Computed Tomography Angiography: The AortaSeg24 Challenge

AU - The AortaSeg24 challenge

AU - Imran, Muhammad

AU - Krebs, Jonathan R.

AU - Sivaraman, Vishal Balaji

AU - Zhang, Teng

AU - Kumar, Amarjeet

AU - Ueland, Walker R.

AU - Fassler, Michael J.

AU - Huang, Jinlong

AU - Sun, Xiao

AU - Wang, Lisheng

AU - Shi, Pengcheng

AU - Rokuss, Maximilian

AU - Baumgartner, Michael

AU - Kirchhof, Yannick

AU - Maier-Hein, Klaus H.

AU - Isensee, Fabian

AU - Liu, Shuolin

AU - Han, Bing

AU - Nguyen, Bong Thanh

AU - Shin, Dong-jin

AU - Ji-Woo, Park

AU - Choi, Mathew

AU - Uhm, Kwang-Hyun

AU - Ko, Sung-Jea

AU - Lee, Chanwoong

AU - Chun, Jaehee

AU - Kim, Jin Sung

AU - Zhang, Minghui

AU - Zhang, Hanxiao

AU - You, Xin

AU - Gu, Yun

AU - Pan, Zhaohong

AU - Liu, Xuan

AU - Liang, Xiaokun

AU - Tiefenthaler, Markus

AU - Almar-Munoz, Enrique

AU - Schwab, Matthias

AU - Kotyushev, Mikhail

AU - Epifanov, Rostislav

AU - Wodzinski, Marek

AU - Muller, Henning

AU - Qayyum, Abdul

AU - Mazher, Moona

AU - Niederer, Steven A.

AU - Wang, Zhiwei

AU - Yang, Kaixiang

AU - Ren, Jintao

AU - Korreman, Stine Sofia

AU - Gao, Yuchong

AU - Zeng, Hongye

N1 - This work was supported by the Department of Medicine and the Intelligent Clinical Care Center at the University of Florida College of Medicine.

PY - 2026/9

Y1 - 2026/9

N2 - Multi-class segmentation of the aorta in computed tomography angiography (CTA) scans is essential for diagnosing and planning complex endovascular treatments for patients with aortic dissections. However, existing methods reduce aortic segmentation to a binary problem, limiting their ability to measure diameters across different branches and zones. Furthermore, no open-source dataset is currently available to support the development of multi-class aortic segmentation methods. To address this gap, we organized the AortaSeg24 MICCAI Challenge, introducing the first dataset of 100 CTA volumes annotated for 23 clinically relevant aortic branches and zones. This dataset was designed to facilitate both model development and validation. The challenge attracted 121 teams worldwide, with participants leveraging state-of-the-art frameworks such as nnU-Net and exploring novel techniques, including cascaded models, data augmentation strategies, and custom loss functions. We evaluated the submitted algorithms using the Dice Similarity Coefficient (DSC) and Normalized Surface Distance (NSD), highlighting the approaches adopted by the top five performing teams. This paper presents the challenge design, dataset details, evaluation metrics, and an in-depth analysis of the top-performing algorithms. The annotated dataset, evaluation code, and implementations of the leading methods are publicly available to support further research. All resources can be accessed at https://aortaseg24.grand-challenge.org.

AB - Multi-class segmentation of the aorta in computed tomography angiography (CTA) scans is essential for diagnosing and planning complex endovascular treatments for patients with aortic dissections. However, existing methods reduce aortic segmentation to a binary problem, limiting their ability to measure diameters across different branches and zones. Furthermore, no open-source dataset is currently available to support the development of multi-class aortic segmentation methods. To address this gap, we organized the AortaSeg24 MICCAI Challenge, introducing the first dataset of 100 CTA volumes annotated for 23 clinically relevant aortic branches and zones. This dataset was designed to facilitate both model development and validation. The challenge attracted 121 teams worldwide, with participants leveraging state-of-the-art frameworks such as nnU-Net and exploring novel techniques, including cascaded models, data augmentation strategies, and custom loss functions. We evaluated the submitted algorithms using the Dice Similarity Coefficient (DSC) and Normalized Surface Distance (NSD), highlighting the approaches adopted by the top five performing teams. This paper presents the challenge design, dataset details, evaluation metrics, and an in-depth analysis of the top-performing algorithms. The annotated dataset, evaluation code, and implementations of the leading methods are publicly available to support further research. All resources can be accessed at https://aortaseg24.grand-challenge.org.

KW - Aorta segmentation

KW - CTA

KW - AortaSeg24

KW - 3D segmentation

UR - https://www.scopus.com/pages/publications/105043956388

UR - https://www.mendeley.com/catalogue/9f3c8317-2398-3c4e-8f52-fb3ba31f209f/

UR - http://arxiv.org/abs/2502.05330

U2 - 10.1016/j.media.2026.104188

DO - 10.1016/j.media.2026.104188

M3 - Article

C2 - 42413466

VL - 113

JO - Medical Image Analysis

JF - Medical Image Analysis

SN - 1361-8415

M1 - 104188

ER -

ID: 81007112