AVT: Multicenter aortic vessel tree CTA dataset collection with ground truth segmentation masks
Related Experiment Videos
Lukas Radl1,2, Yuan Jin1,2,3, Antonio Pepe1,2
1Graz University of Technology (TU Graz), Graz, Styria, Austria.
Insights
This study introduces a multicenter database of 56 aortic vessel trees from computed tomography angiography (CTA) scans. This dataset aids in studying aortic geometry and developing advanced segmentation algorithms.
Area of Science:
- Medical Imaging
- Cardiovascular Research
- Computational Anatomy
Background:
- Aortic vessel tree segmentation is crucial for diagnosing cardiovascular diseases.
- Existing datasets lack diversity in geometry and geographic origin.
- Variability in aortic shape necessitates comprehensive data for robust modeling.
Purpose of the Study:
- To create a diverse, multicenter database of human aortic vessel trees.
- To provide semi-automatically generated segmentation masks (ground truth) for each scan.
- To facilitate the study of aortic geometry and shape variability.
Main Methods:
- Collected 56 computed tomography angiography (CTA) scans covering the entire aorta and its major branches.
- Acquired data from three different collections and multiple hospitals, ensuring varied resolutions.
- Generated semi-automatic segmentation masks for the aortic vessel tree in each scan.
Main Results:
- Established a comprehensive aortic vessel tree database with 56 CTA scans.
- Included detailed segmentation masks, enabling precise anatomical analysis.
- Data exhibits significant geometric and shape variations due to diverse origins and resolutions.
Conclusions:
- The database serves as a valuable resource for studying human aortic geometry.
- It enables the development of robust statistical models for aortic shape.
- The collection is suitable for evaluating and training automatic aortic vessel tree segmentation algorithms, including deep learning approaches.
Abstract:
In this article, we present a multicenter aortic vessel tree database collection, containing 56 aortas and their branches. The datasets have been acquired with computed tomography angiography (CTA) scans and each scan covers the ascending aorta, the aortic arch and its branches into the head/neck area, the thoracic aorta, the abdominal aorta and the lower abdominal aorta with the iliac arteries branching into the legs. For each scan, the collection provides a semi-automatically generated segmentation mask of the aortic vessel tree (ground truth). The scans come from three different collections and various hospitals, having various resolutions, which enables studying the geometry/shape variabilities of human aortas and its branches from different geographic locations. Furthermore, creating a robust statistical model of the shape of human aortic vessel trees, which can be used for various tasks such as the development of fully-automatic segmentation algorithms for new, unseen aortic vessel tree cases, e.g. by training deep learning-based approaches. Hence, the collection can serve as an evaluation set for automatic aortic vessel tree segmentation algorithms.