A deep learning-based algorithm identifies glaucomatous discs using monoscopic fundus photographs

Sidong Liu, Stuart L. Graham, Angela Schulz, Michael Kalloniatis, Barbara Zangerl, Weidong Cai, Yang Gao, Brian Chua, Alexander Klistorner, Yuyi You

Research output: Contribution to journalArticleResearchpeer-review

Abstract

Purpose: To develop and test the performance of a deep learning-based algorithm for glaucomatous disc identification using monoscopic fundus photographs.

Design: Fundus photograph database study.

Participants: Four thousand three hundred ninety-four fundus photographs, including 3768 images from previous Sydney-based clinical studies and 626 images from publicly available online RIM-ONE and High-Resolution Fundus (HRF) databases with definitive diagnoses.

Methods: We merged all databases except the HRF database, and then partitioned the dataset into a training set (80% of all cases) and a testing set (20% of all cases). We used the HRF images as an additional testing set. We compared the performance of the artificial intelligence (AI) system against a panel of practicing ophthalmologists including glaucoma subspecialists from Australia, New Zealand, Canada, and the United Kingdom.

Main Outcome Measures: The sensitivity and specificity of the AI system in detecting glaucomatous optic discs.

Results: By using monoscopic fundus photographs, the AI system demonstrated a high accuracy rate in glaucomatous disc identification (92.7%; 95% confidence interval [CI], 91.2%–94.2%), achieving 89.3% sensitivity (95% CI, 86.8%–91.7%) and 97.1% specificity (95% CI, 96.1%–98.1%), with an area under the receiver operating characteristic curve of 0.97 (95% CI, 0.96–0.98). Using the independent online HRF database (30 images), the AI system again accomplished high accuracy, with 86.7% in both sensitivity and specificity (for ophthalmologists, 75.6% sensitivity and 77.8% specificity) and an area under the receiver operating characteristic curve of 0.89 (95% CI, 0.76–1.00).

Conclusions: This study demonstrated that a deep learning-based algorithm can identify glaucomatous discs at high accuracy level using monoscopic fundus images. Given that it is far easier to obtain monoscopic disc images than high-quality stereoscopic images, this study highlights the algorithm’s potential application in large population-based disease screening or telemedicine programs.
LanguageEnglish
Pages15-22
Number of pages8
JournalOphthalmology Glaucoma
Volume1
Issue number1
DOIs
Publication statusPublished - 15 Jul 2018

Fingerprint

Artificial Intelligence
Learning
Databases
Confidence Intervals
Sensitivity and Specificity
ROC Curve
Telemedicine
Optic Disk
New Zealand
Glaucoma
Canada
Outcome Assessment (Health Care)
Population
Ophthalmologists

Cite this

Liu, Sidong ; Graham, Stuart L. ; Schulz, Angela ; Kalloniatis, Michael ; Zangerl, Barbara ; Cai, Weidong ; Gao, Yang ; Chua, Brian ; Klistorner, Alexander ; You, Yuyi. / A deep learning-based algorithm identifies glaucomatous discs using monoscopic fundus photographs. In: Ophthalmology Glaucoma. 2018 ; Vol. 1, No. 1. pp. 15-22.
@article{5a5977cbaacc4fa8b26b4172239dcdca,
title = "A deep learning-based algorithm identifies glaucomatous discs using monoscopic fundus photographs",
abstract = "Purpose: To develop and test the performance of a deep learning-based algorithm for glaucomatous disc identification using monoscopic fundus photographs.Design: Fundus photograph database study.Participants: Four thousand three hundred ninety-four fundus photographs, including 3768 images from previous Sydney-based clinical studies and 626 images from publicly available online RIM-ONE and High-Resolution Fundus (HRF) databases with definitive diagnoses.Methods: We merged all databases except the HRF database, and then partitioned the dataset into a training set (80{\%} of all cases) and a testing set (20{\%} of all cases). We used the HRF images as an additional testing set. We compared the performance of the artificial intelligence (AI) system against a panel of practicing ophthalmologists including glaucoma subspecialists from Australia, New Zealand, Canada, and the United Kingdom.Main Outcome Measures: The sensitivity and specificity of the AI system in detecting glaucomatous optic discs.Results: By using monoscopic fundus photographs, the AI system demonstrated a high accuracy rate in glaucomatous disc identification (92.7{\%}; 95{\%} confidence interval [CI], 91.2{\%}–94.2{\%}), achieving 89.3{\%} sensitivity (95{\%} CI, 86.8{\%}–91.7{\%}) and 97.1{\%} specificity (95{\%} CI, 96.1{\%}–98.1{\%}), with an area under the receiver operating characteristic curve of 0.97 (95{\%} CI, 0.96–0.98). Using the independent online HRF database (30 images), the AI system again accomplished high accuracy, with 86.7{\%} in both sensitivity and specificity (for ophthalmologists, 75.6{\%} sensitivity and 77.8{\%} specificity) and an area under the receiver operating characteristic curve of 0.89 (95{\%} CI, 0.76–1.00).Conclusions: This study demonstrated that a deep learning-based algorithm can identify glaucomatous discs at high accuracy level using monoscopic fundus images. Given that it is far easier to obtain monoscopic disc images than high-quality stereoscopic images, this study highlights the algorithm’s potential application in large population-based disease screening or telemedicine programs.",
author = "Sidong Liu and Graham, {Stuart L.} and Angela Schulz and Michael Kalloniatis and Barbara Zangerl and Weidong Cai and Yang Gao and Brian Chua and Alexander Klistorner and Yuyi You",
year = "2018",
month = "7",
day = "15",
doi = "10.1016/j.ogla.2018.04.002",
language = "English",
volume = "1",
pages = "15--22",
journal = "Ophthalmology Glaucoma",
issn = "2589-4196",
number = "1",

}

A deep learning-based algorithm identifies glaucomatous discs using monoscopic fundus photographs. / Liu, Sidong; Graham, Stuart L.; Schulz, Angela; Kalloniatis, Michael; Zangerl, Barbara; Cai, Weidong; Gao, Yang; Chua, Brian; Klistorner, Alexander; You, Yuyi.

In: Ophthalmology Glaucoma, Vol. 1, No. 1, 15.07.2018, p. 15-22.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - A deep learning-based algorithm identifies glaucomatous discs using monoscopic fundus photographs

AU - Liu, Sidong

AU - Graham, Stuart L.

AU - Schulz, Angela

AU - Kalloniatis, Michael

AU - Zangerl, Barbara

AU - Cai, Weidong

AU - Gao, Yang

AU - Chua, Brian

AU - Klistorner, Alexander

AU - You, Yuyi

PY - 2018/7/15

Y1 - 2018/7/15

N2 - Purpose: To develop and test the performance of a deep learning-based algorithm for glaucomatous disc identification using monoscopic fundus photographs.Design: Fundus photograph database study.Participants: Four thousand three hundred ninety-four fundus photographs, including 3768 images from previous Sydney-based clinical studies and 626 images from publicly available online RIM-ONE and High-Resolution Fundus (HRF) databases with definitive diagnoses.Methods: We merged all databases except the HRF database, and then partitioned the dataset into a training set (80% of all cases) and a testing set (20% of all cases). We used the HRF images as an additional testing set. We compared the performance of the artificial intelligence (AI) system against a panel of practicing ophthalmologists including glaucoma subspecialists from Australia, New Zealand, Canada, and the United Kingdom.Main Outcome Measures: The sensitivity and specificity of the AI system in detecting glaucomatous optic discs.Results: By using monoscopic fundus photographs, the AI system demonstrated a high accuracy rate in glaucomatous disc identification (92.7%; 95% confidence interval [CI], 91.2%–94.2%), achieving 89.3% sensitivity (95% CI, 86.8%–91.7%) and 97.1% specificity (95% CI, 96.1%–98.1%), with an area under the receiver operating characteristic curve of 0.97 (95% CI, 0.96–0.98). Using the independent online HRF database (30 images), the AI system again accomplished high accuracy, with 86.7% in both sensitivity and specificity (for ophthalmologists, 75.6% sensitivity and 77.8% specificity) and an area under the receiver operating characteristic curve of 0.89 (95% CI, 0.76–1.00).Conclusions: This study demonstrated that a deep learning-based algorithm can identify glaucomatous discs at high accuracy level using monoscopic fundus images. Given that it is far easier to obtain monoscopic disc images than high-quality stereoscopic images, this study highlights the algorithm’s potential application in large population-based disease screening or telemedicine programs.

AB - Purpose: To develop and test the performance of a deep learning-based algorithm for glaucomatous disc identification using monoscopic fundus photographs.Design: Fundus photograph database study.Participants: Four thousand three hundred ninety-four fundus photographs, including 3768 images from previous Sydney-based clinical studies and 626 images from publicly available online RIM-ONE and High-Resolution Fundus (HRF) databases with definitive diagnoses.Methods: We merged all databases except the HRF database, and then partitioned the dataset into a training set (80% of all cases) and a testing set (20% of all cases). We used the HRF images as an additional testing set. We compared the performance of the artificial intelligence (AI) system against a panel of practicing ophthalmologists including glaucoma subspecialists from Australia, New Zealand, Canada, and the United Kingdom.Main Outcome Measures: The sensitivity and specificity of the AI system in detecting glaucomatous optic discs.Results: By using monoscopic fundus photographs, the AI system demonstrated a high accuracy rate in glaucomatous disc identification (92.7%; 95% confidence interval [CI], 91.2%–94.2%), achieving 89.3% sensitivity (95% CI, 86.8%–91.7%) and 97.1% specificity (95% CI, 96.1%–98.1%), with an area under the receiver operating characteristic curve of 0.97 (95% CI, 0.96–0.98). Using the independent online HRF database (30 images), the AI system again accomplished high accuracy, with 86.7% in both sensitivity and specificity (for ophthalmologists, 75.6% sensitivity and 77.8% specificity) and an area under the receiver operating characteristic curve of 0.89 (95% CI, 0.76–1.00).Conclusions: This study demonstrated that a deep learning-based algorithm can identify glaucomatous discs at high accuracy level using monoscopic fundus images. Given that it is far easier to obtain monoscopic disc images than high-quality stereoscopic images, this study highlights the algorithm’s potential application in large population-based disease screening or telemedicine programs.

U2 - 10.1016/j.ogla.2018.04.002

DO - 10.1016/j.ogla.2018.04.002

M3 - Article

VL - 1

SP - 15

EP - 22

JO - Ophthalmology Glaucoma

T2 - Ophthalmology Glaucoma

JF - Ophthalmology Glaucoma

SN - 2589-4196

IS - 1

ER -