English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

In the clinical ophthalmology, the retinal blood vessels processing represent a significant issue regarding the clinical diagnosis. A level of the blood vessels curvature may serve as a reliable indicator of the pathological process. For curvature estimation, a precise model of the retinal blood vessels is necessary. In this paper, we propose a method based on the sensitive edge detector utilizing the fuzzy rules and morphological techniques. The fuzzy edge detector is able to even detect edges while suppressing the high frequency image noise in the non-contrast environment where the image spatial characteristics are weak. Consequent morphological operations serve for adjustment of the segmentation procedure to obtain the smooth model which effectively separates the retinal blood vessels from the retinal background. In the final step, we obtain the binary mathematical model of the retinal blood vessels. We have verified the proposed method against the gold standard images. We have applied the proposed solution on the low-contrast retinal data from the RetCam 3 which is standard for Retinopathy of prematurity. Mostly, when using the RetCam 3, the retinal data has lower contrast therefore, the segmentation procedure is supposed to be robust, even in the noisy environment. 1 RETINAL IMAGE PROCESSING Retinal image processing is one of the crucial tasks for the ophthalmologic practice. In the retinal area we can recognize several structures being clinically evaluated. The optical disc (optical nerve) is clinically perceived as a central point of the retinal structure. It is also a starting point where the beginning of the retinal blood vessels is observable. Optical disc is utilized as a reference point when the retinal lesions are present to track their dynamical progress over the time because the optical disc geometrical features should be stable over the time. Regarding a lot of clinical reasons, the retinal blood vessels are perceived as the most important clinical structure for the retinal diagnosis. On the base of the blood vessels curvature (clinically called tortuosity) we can at least estimate whether the retina exhibits signs of the pathological process. Generally it supposed that no retinal blood vessel is narrow, but it is curved in the whole length. Nevertheless, we recognize an extreme curving which can be observed as oscillating waves is some parts of the blood vessels (Fig. 1). Judging by a current state of the ophthalmological practice, the software tools allowing for automatic modeling and quantification of the retinal blood vessels would significantly contribute to the precise diagnosis. Currently, there is a lack of such software instruments which would generate the retinal mathematical models. It may be caused by several reasons. There are more clinical imaging alternatives for the retinal visualization, especially fundus cameras and retinal probes. These devices give images in different resolution, and other spatial features. Regarding segmentation procedure effectivity, it is supposed the worse image features images have, the worse segmentation accuracy we achieve. We primarily process the data from the retinal probes (RetCam 3) which have a worse image features, therefore the segmentation procedure should have higher sensitivity against the image noise, and simultaneously should be robust even in the low-contrast environment. When Kubicek, J., Timkovic, J., Penhaker, M., Oczka, D., Krestanova, A., Augustynek, M. and Cerny, M. Retinal Blood Vessels Modeling based on Fuzzy Sobel Edge Detection and Morphological Segmentation. DOI: 10.5220/0007237501210126 In Proceedings of the 12th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2019), pages 121-126 ISBN: 978-989-758-353-7 Copyright c © 2019 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved 121 comparing with the Fundus camera data processing, for the RetCam 3 data the segmentation procedure is supposed to be robust even in the noisy environment. (Tan 2016; Meng 2015; Wankhede 2015) Figure 1: Comparison of a retina with the physiological blood vessels (left) – vessels are direct, and blood vessels having the pathological tortuosity signs in a form of the oscillations (right).