Advanced search
2 files | 2.54 MB Add to list

Using the polar transform for efficient deep learning-based aorta segmentation in CTA images

Author
Organization
Abstract
Medical image segmentation often requires segmenting multiple elliptical objects on a single image. This includes, among other tasks, segmenting vessels such as the aorta in axial CTA slices. In this paper, we present a general approach to improving the semantic segmentation performance of neural networks in these tasks and validate our approach on the task of aorta segmentation. We use a cascade of two neural networks, where one performs a rough segmentation based on the U-Net architecture and the other performs the final segmentation on polar image transformations of the input. Connected component analysis of the rough segmentation is used to construct the polar transformations, and predictions on multiple transformations of the same image are fused using hysteresis thresholding. We show that this method improves aorta segmentation performance without requiring complex neural network architectures. In addition, we show that our approach improves robustness and pixel-level recall while achieving segmentation performance in line with the state of the art.
Keywords
Image and Video Processing (eess.IV), Computer Vision and Pattern Recognition (cs.CV), FOS: Electrical engineering, electronic engineering, information engineering, FOS: Computer and information sciences, Convolutional neural network, medical image processing, medical image segmentation, semantic segmentation

Downloads

  • 2206.10294.pdf
    • full text (Author's original)
    • |
    • open access
    • |
    • PDF
    • |
    • 653.62 KB
  • (...).pdf
    • full text (Published version)
    • |
    • UGent only
    • |
    • PDF
    • |
    • 1.88 MB

Citation

Please use this url to cite or link to this publication:

MLA
Benčević, Marin, et al. “Using the Polar Transform for Efficient Deep Learning-Based Aorta Segmentation in CTA Images.” 2022 International Symposium ELMAR, Proceedings, edited by M Mustra et al., IEEE, 2022, pp. 191–94, doi:10.1109/ELMAR55880.2022.9899786.
APA
Benčević, M., Habijan, M., Galić, I., & Babin, D. (2022). Using the polar transform for efficient deep learning-based aorta segmentation in CTA images. In M. Mustra, B. Zovko-Cihlar, & J. Vukovic (Eds.), 2022 International Symposium ELMAR, proceedings (pp. 191–194). https://doi.org/10.1109/ELMAR55880.2022.9899786
Chicago author-date
Benčević, Marin, Marija Habijan, Irena Galić, and Danilo Babin. 2022. “Using the Polar Transform for Efficient Deep Learning-Based Aorta Segmentation in CTA Images.” In 2022 International Symposium ELMAR, Proceedings, edited by M Mustra, B Zovko-Cihlar, and J Vukovic, 191–94. IEEE. https://doi.org/10.1109/ELMAR55880.2022.9899786.
Chicago author-date (all authors)
Benčević, Marin, Marija Habijan, Irena Galić, and Danilo Babin. 2022. “Using the Polar Transform for Efficient Deep Learning-Based Aorta Segmentation in CTA Images.” In 2022 International Symposium ELMAR, Proceedings, ed by. M Mustra, B Zovko-Cihlar, and J Vukovic, 191–194. IEEE. doi:10.1109/ELMAR55880.2022.9899786.
Vancouver
1.
Benčević M, Habijan M, Galić I, Babin D. Using the polar transform for efficient deep learning-based aorta segmentation in CTA images. In: Mustra M, Zovko-Cihlar B, Vukovic J, editors. 2022 International Symposium ELMAR, proceedings. IEEE; 2022. p. 191–4.
IEEE
[1]
M. Benčević, M. Habijan, I. Galić, and D. Babin, “Using the polar transform for efficient deep learning-based aorta segmentation in CTA images,” in 2022 International Symposium ELMAR, proceedings, Zadar, Croatia, 2022, pp. 191–194.
@inproceedings{8760359,
  abstract     = {{Medical image segmentation often requires segmenting multiple elliptical objects on a single image. This includes, among other tasks, segmenting vessels such as the aorta in axial CTA slices. In this paper, we present a general approach to improving the semantic segmentation performance of neural networks in these tasks and validate our approach on the task of aorta segmentation. We use a cascade of two neural networks, where one performs a rough segmentation based on the U-Net architecture and the other performs the final segmentation on polar image transformations of the input. Connected component analysis of the rough segmentation is used to construct the polar transformations, and predictions on multiple transformations of the same image are fused using hysteresis thresholding. We show that this method improves aorta segmentation performance without requiring complex neural network architectures. In addition, we show that our approach improves robustness and pixel-level recall while achieving segmentation performance in line with the state of the art.}},
  author       = {{Benčević, Marin and Habijan, Marija and Galić, Irena and Babin, Danilo}},
  booktitle    = {{2022 International Symposium ELMAR, proceedings}},
  editor       = {{Mustra, M and Zovko-Cihlar, B and Vukovic, J}},
  isbn         = {{9781665470032}},
  issn         = {{1334-2630}},
  keywords     = {{Image and Video Processing (eess.IV),Computer Vision and Pattern Recognition (cs.CV),FOS: Electrical engineering,electronic engineering,information engineering,FOS: Computer and information sciences,Convolutional neural network,medical image processing,medical image segmentation,semantic segmentation}},
  language     = {{eng}},
  location     = {{Zadar, Croatia}},
  pages        = {{191--194}},
  publisher    = {{IEEE}},
  title        = {{Using the polar transform for efficient deep learning-based aorta segmentation in CTA images}},
  url          = {{http://doi.org/10.1109/ELMAR55880.2022.9899786}},
  year         = {{2022}},
}

Altmetric
View in Altmetric
Web of Science
Times cited: