Contextual detection of diabetic pathology in wide-field retinal angiograms

Colin R. Buchanan; Emanuele Trucco

Contextual detection of diabetic pathology in wide-field retinal angiograms

Colin R. Buchanan
, Emanuele Trucco

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

We report a novel algorithm to locate vascular leakage and ischemia in retinal angiographic image sequences leveraging contextual knowledge of co-occurring pathologies. The key contributions are the use of spatio-temporal features exploiting the evolution of intensity levels over the sequence and contextual knowledge to detect ischemia. The specific nature of these diseased regions is determined using an AdaBoost learning algorithm. Training was performed with a varied set of 16 ground-truth image sequences, and testing on unseen Images. The images used were acquired with an Optos ultra-wide-field scanning laser ophthalmoscope. Evaluation against manual annotations demonstrates successful location of 93% of leakage regions and 70% of ischemic regions.

Original language	English
Title of host publication	2008 30TH ANNUAL INTERNATIONAL CONFERENCE OF THE IEEE ENGINEERING IN MEDICINE AND BIOLOGY SOCIETY, VOLS 1-8
Place of Publication	NEW YORK
Publisher	IEEE Computer Society
Pages	5437-5440
Number of pages	4
ISBN (Print)	978-1-4244-1814-5
Publication status	Published - 2008
Event	30th Annual International Conference of the IEEE Engineering in Medicine and Biology Society - Vancouver, Canada Duration: 20 Aug 2008 → 25 Aug 2008

Conference

Conference	30th Annual International Conference of the IEEE Engineering in Medicine and Biology Society
Country/Territory	Canada
City	Vancouver
Period	20/08/08 → 25/08/08

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being

Keywords

FLUORESCEIN ANGIOGRAMS
OCULAR FUNDUS
LEAKAGE
QUANTIFICATION
IMAGES

Cite this

@inproceedings{cb2359354b5f499f939b3c8bc6f6d0e2,

title = "Contextual detection of diabetic pathology in wide-field retinal angiograms",

abstract = "We report a novel algorithm to locate vascular leakage and ischemia in retinal angiographic image sequences leveraging contextual knowledge of co-occurring pathologies. The key contributions are the use of spatio-temporal features exploiting the evolution of intensity levels over the sequence and contextual knowledge to detect ischemia. The specific nature of these diseased regions is determined using an AdaBoost learning algorithm. Training was performed with a varied set of 16 ground-truth image sequences, and testing on unseen Images. The images used were acquired with an Optos ultra-wide-field scanning laser ophthalmoscope. Evaluation against manual annotations demonstrates successful location of 93\% of leakage regions and 70\% of ischemic regions.",

keywords = "FLUORESCEIN ANGIOGRAMS, OCULAR FUNDUS, LEAKAGE, QUANTIFICATION, IMAGES",

author = "Buchanan, \{Colin R.\} and Emanuele Trucco",

note = "Article number 4650444; 30th Annual International Conference of the IEEE Engineering in Medicine and Biology Society ; Conference date: 20-08-2008 Through 25-08-2008",

year = "2008",

language = "English",

isbn = "978-1-4244-1814-5",

pages = "5437--5440",

booktitle = "2008 30TH ANNUAL INTERNATIONAL CONFERENCE OF THE IEEE ENGINEERING IN MEDICINE AND BIOLOGY SOCIETY, VOLS 1-8",

publisher = "IEEE Computer Society",

}

Buchanan, CR & Trucco, E 2008, Contextual detection of diabetic pathology in wide-field retinal angiograms. in 2008 30TH ANNUAL INTERNATIONAL CONFERENCE OF THE IEEE ENGINEERING IN MEDICINE AND BIOLOGY SOCIETY, VOLS 1-8. IEEE Computer Society, NEW YORK, pp. 5437-5440, 30th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, Vancouver, Canada, 20/08/08.

TY - GEN

T1 - Contextual detection of diabetic pathology in wide-field retinal angiograms

AU - Buchanan, Colin R.

AU - Trucco, Emanuele

N1 - Article number 4650444

PY - 2008

Y1 - 2008

N2 - We report a novel algorithm to locate vascular leakage and ischemia in retinal angiographic image sequences leveraging contextual knowledge of co-occurring pathologies. The key contributions are the use of spatio-temporal features exploiting the evolution of intensity levels over the sequence and contextual knowledge to detect ischemia. The specific nature of these diseased regions is determined using an AdaBoost learning algorithm. Training was performed with a varied set of 16 ground-truth image sequences, and testing on unseen Images. The images used were acquired with an Optos ultra-wide-field scanning laser ophthalmoscope. Evaluation against manual annotations demonstrates successful location of 93% of leakage regions and 70% of ischemic regions.

AB - We report a novel algorithm to locate vascular leakage and ischemia in retinal angiographic image sequences leveraging contextual knowledge of co-occurring pathologies. The key contributions are the use of spatio-temporal features exploiting the evolution of intensity levels over the sequence and contextual knowledge to detect ischemia. The specific nature of these diseased regions is determined using an AdaBoost learning algorithm. Training was performed with a varied set of 16 ground-truth image sequences, and testing on unseen Images. The images used were acquired with an Optos ultra-wide-field scanning laser ophthalmoscope. Evaluation against manual annotations demonstrates successful location of 93% of leakage regions and 70% of ischemic regions.

KW - FLUORESCEIN ANGIOGRAMS

KW - OCULAR FUNDUS

KW - LEAKAGE

KW - QUANTIFICATION

KW - IMAGES

M3 - Conference contribution

SN - 978-1-4244-1814-5

SP - 5437

EP - 5440

BT - 2008 30TH ANNUAL INTERNATIONAL CONFERENCE OF THE IEEE ENGINEERING IN MEDICINE AND BIOLOGY SOCIETY, VOLS 1-8

PB - IEEE Computer Society

CY - NEW YORK

T2 - 30th Annual International Conference of the IEEE Engineering in Medicine and Biology Society

Y2 - 20 August 2008 through 25 August 2008

ER -

Contextual detection of diabetic pathology in wide-field retinal angiograms

Abstract

Conference

UN SDGs

Keywords

Fingerprint

Cite this