Bright retinal lesions detection using color fundus images containing reflective features

L. Giancardo, Edward Chaum, T. P. Karnowski, F. Meriaudeau, K. W. Tobin, Y. Li

Research output: Chapter in Book/Report/Conference proceedingConference contribution

5 Citations (Scopus)

Abstract

Recently, the research community has developed many techniques to detect and diagnose diabetic retinopathy with retinal fundus images. This is a necessary step for the implementation of a large scale screening effort in rural areas where ophthalmologists are not available. In the United States of America, the incidence of diabetes is increasing among the young population. Retina fundus images of patients younger than 20 years old present a high amount of reflectance due to the Nerve Fibre Layer (NFL). Generally, the younger the patient the more the reflectance is visible. We are not aware of algorithms able to explicitly deal with this type of artifact. This paper presents a technique to detect bright lesions in patients with a high degree of reflective NFL. First, the candidate bright lesions are detected using image equalization and histogram analysis. Then, a classifier is trained using texture descriptors (Multi-scale Local Binary Patterns) and other statistical features in order to remove the false positives in the lesion detection. Finally, the area of the lesions is used to diagnose diabetic retinopathy. Our database consists of 33 images from a telemedicine network currently under active development. When determining moderate to severe diabetic retinopathy using the bright lesions detected, the algorithm achieves a sensitivity of 100% at a specificity of 100% with a leave-one-out test.

Original languageEnglish (US)
Title of host publicationWorld Congress on Medical Physics and Biomedical Engineering
Subtitle of host publicationBiomedical Engineering for Audiology, Ophthalmology, Emergency and Dental Medicine
Pages292-295
Number of pages4
Edition11
DOIs
StatePublished - Dec 1 2009
EventWorld Congress on Medical Physics and Biomedical Engineering: Biomedical Engineering for Audiology, Ophthalmology, Emergency and Dental Medicine - Munich, Germany
Duration: Sep 7 2009Sep 12 2009

Publication series

NameIFMBE Proceedings
Number11
Volume25
ISSN (Print)1680-0737

Other

OtherWorld Congress on Medical Physics and Biomedical Engineering: Biomedical Engineering for Audiology, Ophthalmology, Emergency and Dental Medicine
CountryGermany
CityMunich
Period9/7/099/12/09

Fingerprint

Color
Telemedicine
Fibers
Medical problems
Screening
Classifiers
Textures

All Science Journal Classification (ASJC) codes

  • Bioengineering
  • Biomedical Engineering

Cite this

Giancardo, L., Chaum, E., Karnowski, T. P., Meriaudeau, F., Tobin, K. W., & Li, Y. (2009). Bright retinal lesions detection using color fundus images containing reflective features. In World Congress on Medical Physics and Biomedical Engineering: Biomedical Engineering for Audiology, Ophthalmology, Emergency and Dental Medicine (11 ed., pp. 292-295). (IFMBE Proceedings; Vol. 25, No. 11). https://doi.org/10.1007/978-3-642-03891-4-78

Bright retinal lesions detection using color fundus images containing reflective features. / Giancardo, L.; Chaum, Edward; Karnowski, T. P.; Meriaudeau, F.; Tobin, K. W.; Li, Y.

World Congress on Medical Physics and Biomedical Engineering: Biomedical Engineering for Audiology, Ophthalmology, Emergency and Dental Medicine. 11. ed. 2009. p. 292-295 (IFMBE Proceedings; Vol. 25, No. 11).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Giancardo, L, Chaum, E, Karnowski, TP, Meriaudeau, F, Tobin, KW & Li, Y 2009, Bright retinal lesions detection using color fundus images containing reflective features. in World Congress on Medical Physics and Biomedical Engineering: Biomedical Engineering for Audiology, Ophthalmology, Emergency and Dental Medicine. 11 edn, IFMBE Proceedings, no. 11, vol. 25, pp. 292-295, World Congress on Medical Physics and Biomedical Engineering: Biomedical Engineering for Audiology, Ophthalmology, Emergency and Dental Medicine, Munich, Germany, 9/7/09. https://doi.org/10.1007/978-3-642-03891-4-78
Giancardo L, Chaum E, Karnowski TP, Meriaudeau F, Tobin KW, Li Y. Bright retinal lesions detection using color fundus images containing reflective features. In World Congress on Medical Physics and Biomedical Engineering: Biomedical Engineering for Audiology, Ophthalmology, Emergency and Dental Medicine. 11 ed. 2009. p. 292-295. (IFMBE Proceedings; 11). https://doi.org/10.1007/978-3-642-03891-4-78
Giancardo, L. ; Chaum, Edward ; Karnowski, T. P. ; Meriaudeau, F. ; Tobin, K. W. ; Li, Y. / Bright retinal lesions detection using color fundus images containing reflective features. World Congress on Medical Physics and Biomedical Engineering: Biomedical Engineering for Audiology, Ophthalmology, Emergency and Dental Medicine. 11. ed. 2009. pp. 292-295 (IFMBE Proceedings; 11).
@inproceedings{6d47cdf0e85942de8db1feb8b9b62ae8,
title = "Bright retinal lesions detection using color fundus images containing reflective features",
abstract = "Recently, the research community has developed many techniques to detect and diagnose diabetic retinopathy with retinal fundus images. This is a necessary step for the implementation of a large scale screening effort in rural areas where ophthalmologists are not available. In the United States of America, the incidence of diabetes is increasing among the young population. Retina fundus images of patients younger than 20 years old present a high amount of reflectance due to the Nerve Fibre Layer (NFL). Generally, the younger the patient the more the reflectance is visible. We are not aware of algorithms able to explicitly deal with this type of artifact. This paper presents a technique to detect bright lesions in patients with a high degree of reflective NFL. First, the candidate bright lesions are detected using image equalization and histogram analysis. Then, a classifier is trained using texture descriptors (Multi-scale Local Binary Patterns) and other statistical features in order to remove the false positives in the lesion detection. Finally, the area of the lesions is used to diagnose diabetic retinopathy. Our database consists of 33 images from a telemedicine network currently under active development. When determining moderate to severe diabetic retinopathy using the bright lesions detected, the algorithm achieves a sensitivity of 100{\%} at a specificity of 100{\%} with a leave-one-out test.",
author = "L. Giancardo and Edward Chaum and Karnowski, {T. P.} and F. Meriaudeau and Tobin, {K. W.} and Y. Li",
year = "2009",
month = "12",
day = "1",
doi = "10.1007/978-3-642-03891-4-78",
language = "English (US)",
isbn = "9783642038907",
series = "IFMBE Proceedings",
number = "11",
pages = "292--295",
booktitle = "World Congress on Medical Physics and Biomedical Engineering",
edition = "11",

}

TY - GEN

T1 - Bright retinal lesions detection using color fundus images containing reflective features

AU - Giancardo, L.

AU - Chaum, Edward

AU - Karnowski, T. P.

AU - Meriaudeau, F.

AU - Tobin, K. W.

AU - Li, Y.

PY - 2009/12/1

Y1 - 2009/12/1

N2 - Recently, the research community has developed many techniques to detect and diagnose diabetic retinopathy with retinal fundus images. This is a necessary step for the implementation of a large scale screening effort in rural areas where ophthalmologists are not available. In the United States of America, the incidence of diabetes is increasing among the young population. Retina fundus images of patients younger than 20 years old present a high amount of reflectance due to the Nerve Fibre Layer (NFL). Generally, the younger the patient the more the reflectance is visible. We are not aware of algorithms able to explicitly deal with this type of artifact. This paper presents a technique to detect bright lesions in patients with a high degree of reflective NFL. First, the candidate bright lesions are detected using image equalization and histogram analysis. Then, a classifier is trained using texture descriptors (Multi-scale Local Binary Patterns) and other statistical features in order to remove the false positives in the lesion detection. Finally, the area of the lesions is used to diagnose diabetic retinopathy. Our database consists of 33 images from a telemedicine network currently under active development. When determining moderate to severe diabetic retinopathy using the bright lesions detected, the algorithm achieves a sensitivity of 100% at a specificity of 100% with a leave-one-out test.

AB - Recently, the research community has developed many techniques to detect and diagnose diabetic retinopathy with retinal fundus images. This is a necessary step for the implementation of a large scale screening effort in rural areas where ophthalmologists are not available. In the United States of America, the incidence of diabetes is increasing among the young population. Retina fundus images of patients younger than 20 years old present a high amount of reflectance due to the Nerve Fibre Layer (NFL). Generally, the younger the patient the more the reflectance is visible. We are not aware of algorithms able to explicitly deal with this type of artifact. This paper presents a technique to detect bright lesions in patients with a high degree of reflective NFL. First, the candidate bright lesions are detected using image equalization and histogram analysis. Then, a classifier is trained using texture descriptors (Multi-scale Local Binary Patterns) and other statistical features in order to remove the false positives in the lesion detection. Finally, the area of the lesions is used to diagnose diabetic retinopathy. Our database consists of 33 images from a telemedicine network currently under active development. When determining moderate to severe diabetic retinopathy using the bright lesions detected, the algorithm achieves a sensitivity of 100% at a specificity of 100% with a leave-one-out test.

UR - http://www.scopus.com/inward/record.url?scp=77950138059&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77950138059&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-03891-4-78

DO - 10.1007/978-3-642-03891-4-78

M3 - Conference contribution

SN - 9783642038907

T3 - IFMBE Proceedings

SP - 292

EP - 295

BT - World Congress on Medical Physics and Biomedical Engineering

ER -