Detection of new vessels on the optic disc using retinal photographs

Keith A. Goatman, Alan D. Fleming, Sam Philip, Graeme J. Williams, John A. Olson, Peter F. Sharp

Research output: Contribution to journalArticle

73 Citations (Scopus)

Abstract

Proliferative diabetic retinopathy is a rare condition likely to lead to severe visual impairment. It is characterized by the development of abnormal new retinal vessels. We describe a method for automatically detecting new vessels on the optic disc using retinal photography. Vessel-like candidate segments are first detected using a method based on watershed lines and ridge strength measurement. Fifteen feature parameters, associated with shape, position, orientation, brightness, contrast and line density are calculated for each candidate segment. Based on these features, each segment is categorized as normal or abnormal using a support vector machine (SVM) classifier. The system was trained and tested by cross-validation using 38 images with new vessels and 71 normal images from two diabetic retinal screening centers and one hospital eye clinic. The discrimination performance of the fifteen features was tested against a clinical reference standard. Fourteen features were found to be effective and used in the final test. The area under the receiver operator characteristic curve was 0.911 for detecting images with new vessels on the disc. This accuracy may be sufficient for it to play a useful clinical role in an automated retinopathy analysis system.
Original languageEnglish
Pages (from-to)972-979
Number of pages8
JournalIEEE Transactions on Medical Imaging
Volume30
Issue number4
Early online date13 Dec 2010
DOIs
Publication statusPublished - Apr 2011

Fingerprint

Optic Disk
Photography
Watersheds
Support vector machines
Luminance
Optics
Screening
Classifiers
Retinal Vessels
Vision Disorders
Diabetic Retinopathy
Support Vector Machine

Keywords

  • diabetes
  • image segmentation
  • optical imaging
  • pixel
  • retina
  • retinopathy
  • support vector machines

Cite this

Detection of new vessels on the optic disc using retinal photographs. / Goatman, Keith A.; Fleming, Alan D.; Philip, Sam; Williams, Graeme J.; Olson, John A.; Sharp, Peter F.

In: IEEE Transactions on Medical Imaging, Vol. 30, No. 4, 04.2011, p. 972-979.

Research output: Contribution to journalArticle

Goatman, Keith A. ; Fleming, Alan D. ; Philip, Sam ; Williams, Graeme J. ; Olson, John A. ; Sharp, Peter F. / Detection of new vessels on the optic disc using retinal photographs. In: IEEE Transactions on Medical Imaging. 2011 ; Vol. 30, No. 4. pp. 972-979.
@article{c462a62f307e4b339d6082fe63e1f555,
title = "Detection of new vessels on the optic disc using retinal photographs",
abstract = "Proliferative diabetic retinopathy is a rare condition likely to lead to severe visual impairment. It is characterized by the development of abnormal new retinal vessels. We describe a method for automatically detecting new vessels on the optic disc using retinal photography. Vessel-like candidate segments are first detected using a method based on watershed lines and ridge strength measurement. Fifteen feature parameters, associated with shape, position, orientation, brightness, contrast and line density are calculated for each candidate segment. Based on these features, each segment is categorized as normal or abnormal using a support vector machine (SVM) classifier. The system was trained and tested by cross-validation using 38 images with new vessels and 71 normal images from two diabetic retinal screening centers and one hospital eye clinic. The discrimination performance of the fifteen features was tested against a clinical reference standard. Fourteen features were found to be effective and used in the final test. The area under the receiver operator characteristic curve was 0.911 for detecting images with new vessels on the disc. This accuracy may be sufficient for it to play a useful clinical role in an automated retinopathy analysis system.",
keywords = "diabetes, image segmentation, optical imaging, pixel, retina, retinopathy, support vector machines",
author = "Goatman, {Keith A.} and Fleming, {Alan D.} and Sam Philip and Williams, {Graeme J.} and Olson, {John A.} and Sharp, {Peter F.}",
year = "2011",
month = "4",
doi = "10.1109/TMI.2010.2099236",
language = "English",
volume = "30",
pages = "972--979",
journal = "IEEE Transactions on Medical Imaging",
issn = "0278-0062",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "4",

}

TY - JOUR

T1 - Detection of new vessels on the optic disc using retinal photographs

AU - Goatman, Keith A.

AU - Fleming, Alan D.

AU - Philip, Sam

AU - Williams, Graeme J.

AU - Olson, John A.

AU - Sharp, Peter F.

PY - 2011/4

Y1 - 2011/4

N2 - Proliferative diabetic retinopathy is a rare condition likely to lead to severe visual impairment. It is characterized by the development of abnormal new retinal vessels. We describe a method for automatically detecting new vessels on the optic disc using retinal photography. Vessel-like candidate segments are first detected using a method based on watershed lines and ridge strength measurement. Fifteen feature parameters, associated with shape, position, orientation, brightness, contrast and line density are calculated for each candidate segment. Based on these features, each segment is categorized as normal or abnormal using a support vector machine (SVM) classifier. The system was trained and tested by cross-validation using 38 images with new vessels and 71 normal images from two diabetic retinal screening centers and one hospital eye clinic. The discrimination performance of the fifteen features was tested against a clinical reference standard. Fourteen features were found to be effective and used in the final test. The area under the receiver operator characteristic curve was 0.911 for detecting images with new vessels on the disc. This accuracy may be sufficient for it to play a useful clinical role in an automated retinopathy analysis system.

AB - Proliferative diabetic retinopathy is a rare condition likely to lead to severe visual impairment. It is characterized by the development of abnormal new retinal vessels. We describe a method for automatically detecting new vessels on the optic disc using retinal photography. Vessel-like candidate segments are first detected using a method based on watershed lines and ridge strength measurement. Fifteen feature parameters, associated with shape, position, orientation, brightness, contrast and line density are calculated for each candidate segment. Based on these features, each segment is categorized as normal or abnormal using a support vector machine (SVM) classifier. The system was trained and tested by cross-validation using 38 images with new vessels and 71 normal images from two diabetic retinal screening centers and one hospital eye clinic. The discrimination performance of the fifteen features was tested against a clinical reference standard. Fourteen features were found to be effective and used in the final test. The area under the receiver operator characteristic curve was 0.911 for detecting images with new vessels on the disc. This accuracy may be sufficient for it to play a useful clinical role in an automated retinopathy analysis system.

KW - diabetes

KW - image segmentation

KW - optical imaging

KW - pixel

KW - retina

KW - retinopathy

KW - support vector machines

U2 - 10.1109/TMI.2010.2099236

DO - 10.1109/TMI.2010.2099236

M3 - Article

VL - 30

SP - 972

EP - 979

JO - IEEE Transactions on Medical Imaging

JF - IEEE Transactions on Medical Imaging

SN - 0278-0062

IS - 4

ER -