BLAHUTA, Jiří, Petr ČERMÁK and Jakub SKÁCEL. The concept of an artificial neural network for the classification of atheromous plaques from digitized segmented histological images. Online. In ICENCO 2018 - 14th International Computer Engineering Conference: Secure Smart Societies. New York: IEEE, 2018, p. 22-25. ISBN 978-1-5386-5117-9. Available from: https://dx.doi.org/10.1109/ICENCO.2018.8636116.
Other formats:   BibTeX LaTeX RIS
Basic information
Original name The concept of an artificial neural network for the classification of atheromous plaques from digitized segmented histological images
Authors BLAHUTA, Jiří (203 Czech Republic, guarantor), Petr ČERMÁK (203 Czech Republic) and Jakub SKÁCEL (203 Czech Republic, belonging to the institution).
Edition New York, ICENCO 2018 - 14th International Computer Engineering Conference: Secure Smart Societies, p. 22-25, 4 pp. 2018.
Publisher IEEE
Other information
Original language English
Type of outcome Proceedings paper
Field of Study 20200 2.2 Electrical engineering, Electronic engineering, Information engineering
Country of publisher United States of America
Confidentiality degree is not subject to a state or trade secret
Publication form electronic version available online
WWW URL
RIV identification code RIV/47813059:19240/18:A0000316
Organization unit Faculty of Philosophy and Science in Opava
ISBN 978-1-5386-5117-9
ISSN 2475-2320
Doi http://dx.doi.org/10.1109/ICENCO.2018.8636116
Keywords in English ANN; Atheromatous plaque; Histological patterns; Image segmentation; Neural Networks
Tags ÚI
Changed by Changed by: Mgr. Kamil Matula, Ph.D., učo 7389. Changed: 26/3/2019 09:31.
Abstract
This paper is dedicted to the concept of an artificial neural network (ANN) for the classification of atheromatous plaque based on digitized histological patterns of areas segmented by means of the Region Growing algorithm. For this purpose, a multi-layered feedforward ANN with supervised learning has been used to successfully classify the segmented areas accordingly. The first phase is focused to find an optimal method for image segmentation. The Region Growing algorithm allows us to separate continuously segmented regions. For each region, appropriate features are selected, which are put into the neural network. The goal of the ANN is to classify plaque patterns into four classes according to their features. The classes represent the following types of plaque: homogeneous, heterogeneous, calcified and that with a high ratio of fat. Successful plaque classification will be a helpful tool for long-term clinical projects, e.g. investigation of plaque composition.
PrintDisplayed: 2/5/2024 21:07