Automatic detection of retinal anatomy to assist diabetic retinopathy screening

Alan Duncan Fleming, Keith A Goatman, Sam Philip, J. A. Olson, Peter Frederick Sharp

Research output: Contribution to journalArticle

119 Citations (Scopus)

Abstract

Screening programmes for diabetic retinopathy are being introduced in the United Kingdom and elsewhere. These require large numbers of retinal images to be manually graded for the presence of disease. Automation of image grading would have a number of benefits. However, an important prerequisite for automation is the accurate location of the main anatomical features in the image, notably the optic disc and the fovea. The locations of these features are necessary so that lesion significance, image field of view and image clarity can be assessed. This paper describes methods for the robust location of the optic disc and fovea. The elliptical form of the major retinal blood vessels is used to obtain approximate locations, which are refined based on the circular edge of the optic disc and the local darkening at the fovea. The methods have been tested on 1056 sequential images from a retinal screening programme. Positional accuracy was better than 0.5 of a disc diameter in 98.4% of cases for optic disc location, and in 96.5% of cases for fovea location. The methods are sufficiently accurate to form an important and effective component of an automated image grading system for diabetic retinopathy screening.

Original languageEnglish
Pages (from-to)331-345
Number of pages15
JournalPhysics in Medicine and Biology
Volume52
Issue number2
DOIs
Publication statusPublished - Jan 2007

Keywords

  • automation
  • diabetic retinopathy
  • humans
  • image enhancement
  • models, statistical
  • optic disk
  • reproducibility of results
  • retina
  • retinal diseases
  • retinal vessels
  • sensitivity and specificity
  • time factors
  • model
  • color fundus images
  • visual impairment
  • blood-vessels
  • optic-nerve
  • fovea
  • disc

Cite this

Automatic detection of retinal anatomy to assist diabetic retinopathy screening. / Fleming, Alan Duncan; Goatman, Keith A; Philip, Sam; Olson, J. A.; Sharp, Peter Frederick.

In: Physics in Medicine and Biology, Vol. 52, No. 2, 01.2007, p. 331-345.

Research output: Contribution to journalArticle

Fleming, Alan Duncan ; Goatman, Keith A ; Philip, Sam ; Olson, J. A. ; Sharp, Peter Frederick. / Automatic detection of retinal anatomy to assist diabetic retinopathy screening. In: Physics in Medicine and Biology. 2007 ; Vol. 52, No. 2. pp. 331-345.
@article{e7d5f26e0792443fae99baffdb88756e,
title = "Automatic detection of retinal anatomy to assist diabetic retinopathy screening",
abstract = "Screening programmes for diabetic retinopathy are being introduced in the United Kingdom and elsewhere. These require large numbers of retinal images to be manually graded for the presence of disease. Automation of image grading would have a number of benefits. However, an important prerequisite for automation is the accurate location of the main anatomical features in the image, notably the optic disc and the fovea. The locations of these features are necessary so that lesion significance, image field of view and image clarity can be assessed. This paper describes methods for the robust location of the optic disc and fovea. The elliptical form of the major retinal blood vessels is used to obtain approximate locations, which are refined based on the circular edge of the optic disc and the local darkening at the fovea. The methods have been tested on 1056 sequential images from a retinal screening programme. Positional accuracy was better than 0.5 of a disc diameter in 98.4{\%} of cases for optic disc location, and in 96.5{\%} of cases for fovea location. The methods are sufficiently accurate to form an important and effective component of an automated image grading system for diabetic retinopathy screening.",
keywords = "automation, diabetic retinopathy, humans, image enhancement, models, statistical, optic disk, reproducibility of results, retina, retinal diseases, retinal vessels, sensitivity and specificity, time factors, model, color fundus images, visual impairment, blood-vessels, optic-nerve, fovea, disc",
author = "Fleming, {Alan Duncan} and Goatman, {Keith A} and Sam Philip and Olson, {J. A.} and Sharp, {Peter Frederick}",
year = "2007",
month = "1",
doi = "10.1088/0031-9155/52/2/002",
language = "English",
volume = "52",
pages = "331--345",
journal = "Physics in Medicine and Biology",
issn = "0031-9155",
publisher = "IOP Publishing Ltd.",
number = "2",

}

TY - JOUR

T1 - Automatic detection of retinal anatomy to assist diabetic retinopathy screening

AU - Fleming, Alan Duncan

AU - Goatman, Keith A

AU - Philip, Sam

AU - Olson, J. A.

AU - Sharp, Peter Frederick

PY - 2007/1

Y1 - 2007/1

N2 - Screening programmes for diabetic retinopathy are being introduced in the United Kingdom and elsewhere. These require large numbers of retinal images to be manually graded for the presence of disease. Automation of image grading would have a number of benefits. However, an important prerequisite for automation is the accurate location of the main anatomical features in the image, notably the optic disc and the fovea. The locations of these features are necessary so that lesion significance, image field of view and image clarity can be assessed. This paper describes methods for the robust location of the optic disc and fovea. The elliptical form of the major retinal blood vessels is used to obtain approximate locations, which are refined based on the circular edge of the optic disc and the local darkening at the fovea. The methods have been tested on 1056 sequential images from a retinal screening programme. Positional accuracy was better than 0.5 of a disc diameter in 98.4% of cases for optic disc location, and in 96.5% of cases for fovea location. The methods are sufficiently accurate to form an important and effective component of an automated image grading system for diabetic retinopathy screening.

AB - Screening programmes for diabetic retinopathy are being introduced in the United Kingdom and elsewhere. These require large numbers of retinal images to be manually graded for the presence of disease. Automation of image grading would have a number of benefits. However, an important prerequisite for automation is the accurate location of the main anatomical features in the image, notably the optic disc and the fovea. The locations of these features are necessary so that lesion significance, image field of view and image clarity can be assessed. This paper describes methods for the robust location of the optic disc and fovea. The elliptical form of the major retinal blood vessels is used to obtain approximate locations, which are refined based on the circular edge of the optic disc and the local darkening at the fovea. The methods have been tested on 1056 sequential images from a retinal screening programme. Positional accuracy was better than 0.5 of a disc diameter in 98.4% of cases for optic disc location, and in 96.5% of cases for fovea location. The methods are sufficiently accurate to form an important and effective component of an automated image grading system for diabetic retinopathy screening.

KW - automation

KW - diabetic retinopathy

KW - humans

KW - image enhancement

KW - models, statistical

KW - optic disk

KW - reproducibility of results

KW - retina

KW - retinal diseases

KW - retinal vessels

KW - sensitivity and specificity

KW - time factors

KW - model

KW - color fundus images

KW - visual impairment

KW - blood-vessels

KW - optic-nerve

KW - fovea

KW - disc

U2 - 10.1088/0031-9155/52/2/002

DO - 10.1088/0031-9155/52/2/002

M3 - Article

VL - 52

SP - 331

EP - 345

JO - Physics in Medicine and Biology

JF - Physics in Medicine and Biology

SN - 0031-9155

IS - 2

ER -