Publication detail

Retinal Image Analysis Aimed at Blood Vessel Tree Segmentation and Early Detection of Neural-Layer Deterioration

JAN, J. ODSTRČILÍK, J. GAZÁREK, J. KOLÁŘ, R.

Original Title

Retinal Image Analysis Aimed at Blood Vessel Tree Segmentation and Early Detection of Neural-Layer Deterioration

Czech Title

Analýza obrazů sítnice za účelem segmentace cévního stromu a časné detekce poškození retinální neoronové vrstvy

English Title

Retinal Image Analysis Aimed at Blood Vessel Tree Segmentation and Early Detection of Neural-Layer Deterioration

Type

journal article

Language

en

Original Abstract

An automatic method of segmenting the retinal vessel tree and estimating status of retinal neural fibre layer (NFL) from high resolution fundus camera images is presented. First, reliable blood vessel segmentation, using 2D directional matched filtering, enables to remove areas occluded by blood vessels thus leaving remaining retinal area available to the following NFL detection. The local existence of rather faint and hardly visible NFL is detected by combining several newly designed local textural features, sensitive to subtle NFL characteristics, into feature vectors submitted to a trained neural-network classifier. Obtained binary retinal maps of NFL distribution show a good agreement with both medical expert evaluations and quantitative results obtained by optical coherence tomography.

Czech abstract

Je prezentována automatická metoda segmentace retinálního cévního stromu a odhadu stavu vrstvy nervových vláken (RNFL) na sítnici z oftalmologických fundus-kamera snímků o vysokém rozlišení. Získané výsledky odhadu RNFL vykazují dobrý souhlas jak s manuálním ohodnocením snímků expertem - oftalmologem, tak zejména také s kvantitativními výsledky měření metodou optické koherenční tomografie (OCT)-

English abstract

An automatic method of segmenting the retinal vessel tree and estimating status of retinal neural fibre layer (NFL) from high resolution fundus camera images is presented. First, reliable blood vessel segmentation, using 2D directional matched filtering, enables to remove areas occluded by blood vessels thus leaving remaining retinal area available to the following NFL detection. The local existence of rather faint and hardly visible NFL is detected by combining several newly designed local textural features, sensitive to subtle NFL characteristics, into feature vectors submitted to a trained neural-network classifier. Obtained binary retinal maps of NFL distribution show a good agreement with both medical expert evaluations and quantitative results obtained by optical coherence tomography.

Keywords

retinal imaging, fundus-camera, retinal vessel tree, retinal neural fibre layer, image segmentation, 2D matched filtering, texture analysis, 2D spectra, edge maps

RIV year

2012

Released

03.09.2012

Publisher

Elsevier

Location

Amsterdam (worldwide)

Pages from

431

Pages to

441

Pages count

11

BibTex


@article{BUT92587,
  author="Jiří {Jan} and Jan {Odstrčilík} and Jiří {Gazárek} and Radim {Kolář}",
  title="Retinal Image Analysis Aimed at Blood Vessel Tree Segmentation and Early Detection of Neural-Layer Deterioration",
  annote="An automatic method of segmenting the retinal vessel tree and estimating status of retinal neural fibre layer (NFL) from high resolution fundus camera images is presented. First, reliable blood vessel segmentation, using 2D directional matched filtering, enables to remove areas occluded by blood vessels thus leaving remaining retinal area available to the following NFL detection. The local existence of rather faint and hardly visible NFL is detected by combining several newly designed local textural features, sensitive to subtle NFL characteristics, into feature vectors submitted to a trained neural-network classifier. Obtained binary retinal maps of NFL distribution show a good agreement with both medical expert evaluations and quantitative results obtained by optical coherence tomography.",
  address="Elsevier",
  chapter="92587",
  institution="Elsevier",
  number="6",
  volume="2012",
  year="2012",
  month="september",
  pages="431--441",
  publisher="Elsevier",
  type="journal article"
}