High-Performance Optic Disc Segmentation Using Convolutional Neural Networks

Mohan, D and Harish Kumar, JR and Sekhar Seelamantula, C (2018) High-Performance Optic Disc Segmentation Using Convolutional Neural Networks. In: 25th IEEE International Conference on Image Processing, ICIP 2018, 7 - 10 October 2018, Athens; Greece, pp. 4038-4042.

PDF Pro_Int_Con_Ima_Pro_2018.pdf - Published Version Restricted to Registered users only Download (2MB) | Request a copy

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 Fine-Net, which generates a high-resolution optic disc segmentation map (1024 � 1024) from retinal fundus images. The network is trained on three publicly available datasets, MESSI-DOR, 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. © 2018 IEEE.

Item Type:	Conference Proceedings
Publication:	Proceedings - International Conference on Image Processing, ICIP
Additional Information:	Copyright for this article belongs to IEEE.
Keywords:	Convolution; Diagnosis; Neural networks; Ophthalmology; Optical data processing, Convolutional neural network; Fundus image; Glaucoma; Jaccard coefficients; Optic disc, Image segmentation
Department/Centre:	Division of Electrical Sciences > Electrical Engineering
Date Deposited:	06 May 2019 12:43
Last Modified:	06 May 2019 12:43
URI:	http://eprints.iisc.ac.in/id/eprint/62122

Actions (login required)

View Item