Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Automatic Aorta Segmentation with Heavily Augmented, High-Resolution 3-D ResUNet: Contribution to the SEG.A Challenge

Marek Wodzinski, Henning Müller

TL;DR

The paper tackles automatic aorta segmentation from high-resolution 3-D CT angiography in the SEG.A challenge, emphasizing that extensive data preprocessing and augmentation can outperform architectural novelty in low-data regimes. It proposes a 3-D residual U-Net pipeline with heavy augmentation, fixed $400^3$ inputs, and a Dice–Focal loss objective, achieving Dice coefficients above $0.9$ with strong stability across fivefold validation and competitive surface and volumetric meshing quality. Key contributions include demonstrating the importance of data preparation, providing a robust, end-to-end pipeline, and releasing code, pretrained models, and Grand-Challenge access to support reproducibility and clinical translation. The approach achieves clinically meaningful meshes and fast inference, suggesting practical utility across centers with heterogeneous imaging protocols, while outlining avenues for improvement such as patch-based strategies and higher-resolution processing to better capture small vessel branches.

Abstract

Automatic aorta segmentation from 3-D medical volumes is an important yet difficult task. Several factors make the problem challenging, e.g. the possibility of aortic dissection or the difficulty with segmenting and annotating the small branches. This work presents a contribution by the MedGIFT team to the SEG.A challenge organized during the MICCAI 2023 conference. We propose a fully automated algorithm based on deep encoder-decoder architecture. The main assumption behind our work is that data preprocessing and augmentation are much more important than the deep architecture, especially in low data regimes. Therefore, the solution is based on a variant of traditional convolutional U-Net. The proposed solution achieved a Dice score above 0.9 for all testing cases with the highest stability among all participants. The method scored 1st, 4th, and 3rd in terms of the clinical evaluation, quantitative results, and volumetric meshing quality, respectively. We freely release the source code, pretrained model, and provide access to the algorithm on the Grand-Challenge platform.

Automatic Aorta Segmentation with Heavily Augmented, High-Resolution 3-D ResUNet: Contribution to the SEG.A Challenge

TL;DR

The paper tackles automatic aorta segmentation from high-resolution 3-D CT angiography in the SEG.A challenge, emphasizing that extensive data preprocessing and augmentation can outperform architectural novelty in low-data regimes. It proposes a 3-D residual U-Net pipeline with heavy augmentation, fixed inputs, and a Dice–Focal loss objective, achieving Dice coefficients above with strong stability across fivefold validation and competitive surface and volumetric meshing quality. Key contributions include demonstrating the importance of data preparation, providing a robust, end-to-end pipeline, and releasing code, pretrained models, and Grand-Challenge access to support reproducibility and clinical translation. The approach achieves clinically meaningful meshes and fast inference, suggesting practical utility across centers with heterogeneous imaging protocols, while outlining avenues for improvement such as patch-based strategies and higher-resolution processing to better capture small vessel branches.

Abstract

Automatic aorta segmentation from 3-D medical volumes is an important yet difficult task. Several factors make the problem challenging, e.g. the possibility of aortic dissection or the difficulty with segmenting and annotating the small branches. This work presents a contribution by the MedGIFT team to the SEG.A challenge organized during the MICCAI 2023 conference. We propose a fully automated algorithm based on deep encoder-decoder architecture. The main assumption behind our work is that data preprocessing and augmentation are much more important than the deep architecture, especially in low data regimes. Therefore, the solution is based on a variant of traditional convolutional U-Net. The proposed solution achieved a Dice score above 0.9 for all testing cases with the highest stability among all participants. The method scored 1st, 4th, and 3rd in terms of the clinical evaluation, quantitative results, and volumetric meshing quality, respectively. We freely release the source code, pretrained model, and provide access to the algorithm on the Grand-Challenge platform.
Paper Structure (21 sections, 4 figures, 4 tables)

This paper contains 21 sections, 4 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Overview
  3. Related Work
  4. Design Choices
  5. Contribution
  6. Method
  7. Preprocessing
  8. Augmentation
  9. Deep Network
  10. Objective function & Training
  11. Inference & Postprocessing
  12. Meshing
  13. Dataset
  14. Experimental Setup
  15. Source Code
  16. ...and 6 more sections

Figures (4)

  • Figure 1: Exemplary cases from each medical center. Please note the differences related to the intensity distributions (unrelated to the visualization window), field of view, and spatial resolution (e.g. Rider cases have significantly smaller voxel size).
  • Figure 2: Exemplary visualizations from the internal validation set for cases from each data source. The ground-truths are shown in green and the calculated segmentation masks in red.
  • Figure 3: Exemplary surface meshes for cases from all data sources (validation cases). Please note that the differences are mostly in fine-details related to small branches.
  • Figure 4: Exemplary surface meshes for a case with an aortic dissection (Rider dataset - validation case). Note that the dissection was correctly segmented, the only differences are related to small branches.