Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

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

Muhammad Imran, Jonathan R. Krebs, Vishal Balaji Sivaraman, Teng Zhang, Amarjeet Kumar, Walker R. Ueland, Michael J. Fassler, Jinlong Huang, Xiao Sun, Lisheng Wang, Pengcheng Shi, Maximilian Rokuss, Michael Baumgartner, Yannick Kirchhof, Klaus H. Maier-Hein, Fabian Isensee, Shuolin Liu, Bing Han, Bong Thanh Nguyen, Dong-jin Shin, Park Ji-Woo, Mathew Choi, Kwang-Hyun Uhm, Sung-Jea Ko, Chanwoong Lee, Jaehee Chun, Jin Sung Kim, Minghui Zhang, Hanxiao Zhang, Xin You, Yun Gu, Zhaohong Pan, Xuan Liu, Xiaokun Liang, Markus Tiefenthaler, Enrique Almar-Munoz, Matthias Schwab, Mikhail Kotyushev, Rostislav Epifanov, Marek Wodzinski, Henning Muller, Abdul Qayyum, Moona Mazher, Steven A. Niederer, Zhiwei Wang, Kaixiang Yang, Jintao Ren, Stine Sofia Korreman, Yuchong Gao, Hongye Zeng, Haoyu Zheng, Rui Zheng, Jinghua Yue, Fugen Zhou, Bo Liu, Alexander Cosman, Muxuan Liang, Chang Zhao, Gilbert R. Upchurch, Jun Ma, Yuyin Zhou, Michol A. Cooper, Wei Shao

TL;DR

The paper introduces the AortaSeg24 MICCAI Challenge and its first open dataset for multi-class segmentation of the aorta, spanning 23 clinically relevant branches and SVS/STS zones in 100 CTA volumes. It assesses 16 leading approaches, highlighting cascaded and two-stage nnU-Net-based frameworks with specialized losses (e.g., cbDice, Skeleton Recall Loss) and extensive data augmentation, evaluated via $DSC$ and $NSD$. The work provides a rigorous ranking framework (rank-then-aggregate) and analyzes score distributions and ranking stability, establishing a robust benchmark for zonal aortic measurements. By releasing the dataset, evaluation code, and top methods, the study aims to accelerate clinically actionable AI tools for automated, zonal aortic assessment and treatment planning.

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.

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

TL;DR

The paper introduces the AortaSeg24 MICCAI Challenge and its first open dataset for multi-class segmentation of the aorta, spanning 23 clinically relevant branches and SVS/STS zones in 100 CTA volumes. It assesses 16 leading approaches, highlighting cascaded and two-stage nnU-Net-based frameworks with specialized losses (e.g., cbDice, Skeleton Recall Loss) and extensive data augmentation, evaluated via and . The work provides a rigorous ranking framework (rank-then-aggregate) and analyzes score distributions and ranking stability, establishing a robust benchmark for zonal aortic measurements. By releasing the dataset, evaluation code, and top methods, the study aims to accelerate clinically actionable AI tools for automated, zonal aortic assessment and treatment planning.

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.

Paper Structure

This paper contains 29 sections, 2 equations, 8 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Existing Aortic Segmentation Methods
  4. Existing Aortic Segmentation Datasets
  5. Relevant Image Segmentation Challenges
  6. Methods
  7. Challenge Organization
  8. Challenge Dataset
  9. Image Acquisition
  10. Clinically Relevant Aortic Zones and Branches
  11. Data Annotation
  12. Evaluation Metrics and Ranking Criterion
  13. Results
  14. Challenge Participants and Submissions
  15. Approaches of the Top Five Teams
  16. ...and 14 more sections

Figures (8)

  • Figure 1: Schematic representation of different aortic branches and zones.
  • Figure 2: Comparison of various publicly available aortic segmentation datasets
  • Figure 3: Overview of the participating teams at each phase.
  • Figure 4: Box plot showing the distribution of Dice scores for 16 segmentation algorithms. The x-axis represents the team names, and the y-axis shows the Dice scores. Each dot corresponds to the Dice score for an individual subject segmented by an algorithm.
  • Figure 5: Box plot showing the distribution of NSD scores for 16 segmentation algorithms. The x-axis represents the team names, and the y-axis shows the NSD scores. Each dot corresponds to the NSD score for an individual subject segmented by an algorithm.
  • ...and 3 more figures