C 2019

Modeling and extraction of retinal blood vessels from RetCam 3 based on morphological segmentation

KŘESŤANOVÁ, Alice, Jan KUBÍČEK, Juraj TIMKOVIČ, Marek PENHAKER, David OCZKA et. al.

Basic information

Original name

Modeling and extraction of retinal blood vessels from RetCam 3 based on morphological segmentation

Authors

KŘESŤANOVÁ, Alice, Jan KUBÍČEK, Juraj TIMKOVIČ, Marek PENHAKER, David OCZKA and Jan VAŇUŠ

Edition

Switzerland, Intelligent Information and Database Systems: Recent Developments, p. 255-263, 9 pp. SCI, volume 830, 2019

Publisher

Springer, Cham

Other information

Language

English

Type of outcome

Kapitola resp. kapitoly v odborné knize

Field of Study

30207 Ophthalmology

Country of publisher

Switzerland

Confidentiality degree

není předmětem státního či obchodního tajemství

Publication form

electronic version available online

References:

Organization unit

Faculty of Medicine

ISBN

978-3-030-14132-5

Keywords in English

Retina, Morphological segmentation, Detection, Blood vessels, Retinopathy of prematurity, Comparison, RetCam 3, MATLAB

Tags

International impact, Reviewed
Změněno: 24/4/2020 10:33, Mgr. Tereza Miškechová

Abstract

V originále

This paper deals with the analysis and modeling of the retinal blood vessels system. The aim of the analysis is the design and implementation of a fully automated segmentation model based on the morphological segmentation, allowing for extraction of the blood system area within the binary model, where other retinal structures are suppressed. An important feature of the model is sensitivity and robustness to declare the efficacy of segmentation in an environment with worse image parameters. For this reason, the designed model is also tested for data where the vascular system is visualized under a low contrast. Part of the analysis is the comparative testing of the designed model against selected segmentation methods based on objective criteria. The designed model was tested and verified on dataset from system RetCam 3 containing 22 images.