ePrints@IIScePrints@IISc Home | About | Browse | Latest Additions | Advanced Search | Contact | Help

HIGH-PERFORMANCE OPTIC DISC SEGMENTATION USING CONVOLUTIONAL NEURAL NETWORKS

Mohan, Dhruv and Kumar, JR Harish and Seelamantula, Chandra Sekhar (2018) HIGH-PERFORMANCE OPTIC DISC SEGMENTATION USING CONVOLUTIONAL NEURAL NETWORKS. In: 25th IEEE International Conference on Image Processing (ICIP), OCT 07-10, 2018, Athens, GREECE, pp. 4038-4042.

[img] PDF
Iee_Int_Con_Ima_Pro_4038_2018.pdf - Published Version
Restricted to Registered users only

Download (5MB) | Request a copy
Official URL: https://doi.org/10.1109/ICIP.2018.8451543

Abstract

We present a framework for robust optic disc segmentation using convolutional neural networks. Optic disc is an important anatomical landmark in the fundus image used for the diagnosis of ophthalmological pathologies. Our objective is to develop a system for unsupervised, early and robust detection of diseases such as glaucoma. We introduce the FineNet, which generates a high-resolution optic disc segmentation map (1024x1024) from retinal fundus images. The network is trained on three publicly available datasets, MESSIDOR, DRIONS-DB, and DRISHTI-GS. The proposed framework generalizes well as it performs reliably even on test images that have a significant variability. For experimental evaluation, we perform a five-fold cross-validation and achieve accurate optic disc localization in 99.4% of cases. Moreover, for optic disc segmentation we achieve an average Dice coefficient and Jaccard coefficient of 0.958 and 0.921, respectively.

Item Type: Conference Paper
Additional Information: 25th IEEE International Conference on Image Processing (ICIP), Athens, GREECE, OCT 07-10, 2018
Keywords: Convolutional neural networks; fundus image; optic disc; glaucoma; Jaccard coefficient
Department/Centre: Division of Electrical Sciences > Electrical Engineering
Depositing User: Id for Latest eprints
Date Deposited: 07 Feb 2019 04:43
Last Modified: 07 Feb 2019 04:43
URI: http://eprints.iisc.ac.in/id/eprint/61609

Actions (login required)

View Item View Item