2Sranking-cnn: A 2-stage ranking-CNN for diagnosis of glaucoma from fundus images using CAM-extracted ROI as an intermediate input

Tae Joon Jun, Dohyeun Kim, Hoang Minh Nguyen, Daeyoung Kim, Youngsub Eom

Research output: Contribution to conferencePaper

Abstract

Glaucoma is a disease in which the optic nerve is chronically damaged by the elevation of the intra-ocular pressure, resulting in visual field defect. Therefore, it is important to monitor and treat suspected patients before they are confirmed with glaucoma. In this paper, we propose a 2-stage ranking-CNN that classifies fundus images as normal, suspicious, and glaucoma. Furthermore, we propose a method of using the class activation map as a mask filter and combining it with the original fundus image as an intermediate input. Our results have improved the average accuracy by about 10% over the existing 3-class CNN and ranking-CNN, and especially improved the sensitivity of suspicious class by more than 20% over 3-class CNN. In addition, the extracted ROI was also found to overlap with the diagnostic criteria of the physician. The method we propose is expected to be efficiently applied to any medical data where there is a suspicious condition between normal and disease.

Original languageEnglish
Publication statusPublished - 2019 Jan 1
Event29th British Machine Vision Conference, BMVC 2018 - Newcastle, United Kingdom
Duration: 2018 Sep 32018 Sep 6

Conference

Conference29th British Machine Vision Conference, BMVC 2018
CountryUnited Kingdom
CityNewcastle
Period18/9/318/9/6

Fingerprint

Computer aided manufacturing
Masks
Optics
Chemical activation
Defects

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition

Cite this

Jun, T. J., Kim, D., Nguyen, H. M., Kim, D., & Eom, Y. (2019). 2Sranking-cnn: A 2-stage ranking-CNN for diagnosis of glaucoma from fundus images using CAM-extracted ROI as an intermediate input. Paper presented at 29th British Machine Vision Conference, BMVC 2018, Newcastle, United Kingdom.

2Sranking-cnn : A 2-stage ranking-CNN for diagnosis of glaucoma from fundus images using CAM-extracted ROI as an intermediate input. / Jun, Tae Joon; Kim, Dohyeun; Nguyen, Hoang Minh; Kim, Daeyoung; Eom, Youngsub.

2019. Paper presented at 29th British Machine Vision Conference, BMVC 2018, Newcastle, United Kingdom.

Research output: Contribution to conferencePaper

Jun, TJ, Kim, D, Nguyen, HM, Kim, D & Eom, Y 2019, '2Sranking-cnn: A 2-stage ranking-CNN for diagnosis of glaucoma from fundus images using CAM-extracted ROI as an intermediate input' Paper presented at 29th British Machine Vision Conference, BMVC 2018, Newcastle, United Kingdom, 18/9/3 - 18/9/6, .
Jun TJ, Kim D, Nguyen HM, Kim D, Eom Y. 2Sranking-cnn: A 2-stage ranking-CNN for diagnosis of glaucoma from fundus images using CAM-extracted ROI as an intermediate input. 2019. Paper presented at 29th British Machine Vision Conference, BMVC 2018, Newcastle, United Kingdom.
Jun, Tae Joon ; Kim, Dohyeun ; Nguyen, Hoang Minh ; Kim, Daeyoung ; Eom, Youngsub. / 2Sranking-cnn : A 2-stage ranking-CNN for diagnosis of glaucoma from fundus images using CAM-extracted ROI as an intermediate input. Paper presented at 29th British Machine Vision Conference, BMVC 2018, Newcastle, United Kingdom.
@conference{8d6ad1864c3c4ebfbbc3b9b3594a01b4,
title = "2Sranking-cnn: A 2-stage ranking-CNN for diagnosis of glaucoma from fundus images using CAM-extracted ROI as an intermediate input",
abstract = "Glaucoma is a disease in which the optic nerve is chronically damaged by the elevation of the intra-ocular pressure, resulting in visual field defect. Therefore, it is important to monitor and treat suspected patients before they are confirmed with glaucoma. In this paper, we propose a 2-stage ranking-CNN that classifies fundus images as normal, suspicious, and glaucoma. Furthermore, we propose a method of using the class activation map as a mask filter and combining it with the original fundus image as an intermediate input. Our results have improved the average accuracy by about 10{\%} over the existing 3-class CNN and ranking-CNN, and especially improved the sensitivity of suspicious class by more than 20{\%} over 3-class CNN. In addition, the extracted ROI was also found to overlap with the diagnostic criteria of the physician. The method we propose is expected to be efficiently applied to any medical data where there is a suspicious condition between normal and disease.",
author = "Jun, {Tae Joon} and Dohyeun Kim and Nguyen, {Hoang Minh} and Daeyoung Kim and Youngsub Eom",
year = "2019",
month = "1",
day = "1",
language = "English",
note = "29th British Machine Vision Conference, BMVC 2018 ; Conference date: 03-09-2018 Through 06-09-2018",

}

TY - CONF

T1 - 2Sranking-cnn

T2 - A 2-stage ranking-CNN for diagnosis of glaucoma from fundus images using CAM-extracted ROI as an intermediate input

AU - Jun, Tae Joon

AU - Kim, Dohyeun

AU - Nguyen, Hoang Minh

AU - Kim, Daeyoung

AU - Eom, Youngsub

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Glaucoma is a disease in which the optic nerve is chronically damaged by the elevation of the intra-ocular pressure, resulting in visual field defect. Therefore, it is important to monitor and treat suspected patients before they are confirmed with glaucoma. In this paper, we propose a 2-stage ranking-CNN that classifies fundus images as normal, suspicious, and glaucoma. Furthermore, we propose a method of using the class activation map as a mask filter and combining it with the original fundus image as an intermediate input. Our results have improved the average accuracy by about 10% over the existing 3-class CNN and ranking-CNN, and especially improved the sensitivity of suspicious class by more than 20% over 3-class CNN. In addition, the extracted ROI was also found to overlap with the diagnostic criteria of the physician. The method we propose is expected to be efficiently applied to any medical data where there is a suspicious condition between normal and disease.

AB - Glaucoma is a disease in which the optic nerve is chronically damaged by the elevation of the intra-ocular pressure, resulting in visual field defect. Therefore, it is important to monitor and treat suspected patients before they are confirmed with glaucoma. In this paper, we propose a 2-stage ranking-CNN that classifies fundus images as normal, suspicious, and glaucoma. Furthermore, we propose a method of using the class activation map as a mask filter and combining it with the original fundus image as an intermediate input. Our results have improved the average accuracy by about 10% over the existing 3-class CNN and ranking-CNN, and especially improved the sensitivity of suspicious class by more than 20% over 3-class CNN. In addition, the extracted ROI was also found to overlap with the diagnostic criteria of the physician. The method we propose is expected to be efficiently applied to any medical data where there is a suspicious condition between normal and disease.

UR - http://www.scopus.com/inward/record.url?scp=85072336030&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85072336030&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:85072336030

ER -