Segmentation based on gabor transformation with machine learning: Modeling of retinal blood vessels system from retcam images and tortuosity extraction

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F62690094%3A18450%2F17%3A50014267" target="_blank" >RIV/62690094:18450/17:50014267 - isvavai.cz</a>

Nalezeny alternativní kódy

RIV/61989100:27240/17:10237654

Výsledek na webu

<a href="http://dx.doi.org/10.3233/978-1-61499-800-6-270" target="_blank" >http://dx.doi.org/10.3233/978-1-61499-800-6-270</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.3233/978-1-61499-800-6-270" target="_blank" >10.3233/978-1-61499-800-6-270</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Segmentation based on gabor transformation with machine learning: Modeling of retinal blood vessels system from retcam images and tortuosity extraction

Popis výsledku v původním jazyce

In a field of the clinical ophthalmology, an analysis of the retinal blood vessels is one of the major assessments in the retinal system. Retinal blood vessels system is clinically imagined either by the fundus camera or retinal probe (RetCam 3 system). The tortuosity is important parameter assessing blood vessel curvature. Unfortunately, this parameter is usually subjectively estimated in the retinal image analysis. The main aim of the analysis is an automatic segmentation with consequent extraction and modelling of the retinal blood vessels system from RetCam 3 in the form of the binary model. Segmentation algorithm utilizes the Gabor wavelet transformation (GT) giving segmentation results for individual parameters setting. Consequent retinal blood vessels classification is carried out on the base of the linear regression with gold standard. The gold standard represents a manually labelled segmentation by the ophthalmologic experts. Binary segmentation model precisely approximates blood vessels area from other structures. This model allows for the tortuosity extraction in a form of the gradient image where each blood vessel element is described by its steepness.

Název v anglickém jazyce

Segmentation based on gabor transformation with machine learning: Modeling of retinal blood vessels system from retcam images and tortuosity extraction

Popis výsledku anglicky

In a field of the clinical ophthalmology, an analysis of the retinal blood vessels is one of the major assessments in the retinal system. Retinal blood vessels system is clinically imagined either by the fundus camera or retinal probe (RetCam 3 system). The tortuosity is important parameter assessing blood vessel curvature. Unfortunately, this parameter is usually subjectively estimated in the retinal image analysis. The main aim of the analysis is an automatic segmentation with consequent extraction and modelling of the retinal blood vessels system from RetCam 3 in the form of the binary model. Segmentation algorithm utilizes the Gabor wavelet transformation (GT) giving segmentation results for individual parameters setting. Consequent retinal blood vessels classification is carried out on the base of the linear regression with gold standard. The gold standard represents a manually labelled segmentation by the ophthalmologic experts. Binary segmentation model precisely approximates blood vessels area from other structures. This model allows for the tortuosity extraction in a form of the gradient image where each blood vessel element is described by its steepness.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

<a href="/cs/project/GA17-03037S" target="_blank" >GA17-03037S: Hodnocení investic do vývoje zdravotních prostředků</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2017

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Frontiers in Artificial Intelligence and Applications

ISBN

978-1-61499-799-3

ISSN

0922-6389

e-ISSN

neuvedeno

Počet stran výsledku

14

Strana od-do

270-283

Název nakladatele

IOS Press

Místo vydání

Kitakyushu

Místo konání akce

Kitakyushu; Japan

Datum konání akce

26. 9. 2017

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—