Depending on the amount of data to process, file generation may take longer.

If it takes too long to generate, you can limit the data by, for example, reducing the range of years.

Chapter

Download BibTeX

Title

Segmentation of Retina Vessels in 2D OCT-Reconstructed Fundus Images with 3D UNet

Authors

[ 1 ] Instytut Automatyki i Robotyki, Wydział Automatyki, Robotyki i Elektrotechniki, Politechnika Poznańska | [ P ] employee

Scientific discipline (Law 2.0)

[2.2] Automation, electronics, electrical engineering and space technologies

Year of publication

2023

Chapter type

chapter in monograph / paper

Publication language

english

Keywords
EN
  • optical coherence tomography (OCT)
  • retina vessel segmentation
  • convolutional neural networks
  • UNet
Abstract

EN This paper presents an analysis of the effectiveness of blood vessel segmentation in human retinal images. The segmentation process was carried out for three-dimensional optical coherence tomography (OCT) scans using the author’s publicly available CAVRI-C database. A 3D version of the U-Net artificial neural network was used. Various dimensions of the input blocks were investigated to determine their influence on the segmentation process. The effectiveness of the proposed solutions was evaluated using various metrics and the results obtained were compared with the 2D solution, also based on neural networks. The use of the proposed solutions allowed us to improve the segmentation even by 15% (in the case of the F1-score).

Pages (from - to)

60 - 65

URL

https://ieeexplore.ieee.org/document/10274427

Book

SPA 2023 Signal Processing : Algorithms, Architectures, Arrangements, and Applications : Conference Proceedings, Poznan, 20th-22nd September 2023

Presented on

SPA 2023 26th IEEE Signal Processing - Algorithms, Architectures, Arrangements, and Applications, 20-22.09.2023, Poznań, Polska

Ministry points / chapter

20

This website uses cookies to remember the authenticated session of the user. For more information, read about Cookies and Privacy Policy.