Improved optic disc and cup segmentation in Glaucomatic images using deep learning architecture

Abstract

Glaucoma is an ailment causing permanent vision loss but can be prevented through the early detection. Optic disc to cup ratio is one of the key factors for glaucoma diagnosis. But accurate segmentation of disc and cup is still a challenge. To mitigate this challenge, an effective system for optic disc and cup segmentation using deep learning architecture is presented in this paper. Modified Groundtruth is utilized to train the proposed model. It works as fused segmentation marking by multiple experts that helps in improving the performance of the system. Extensive computer simulations are conducted to test the efficiency of the proposed system. For the implementation three standard benchmark datasets such as DRISHTI-GS, DRIONS-DB and RIM-ONE v3 are used. The performance of the proposed system is validated against the state-of-the-art methods. Results indicate an average overlapping score of 96.62%, 96.15% and 98.42% respectively for optic disc segmentation and an average overlapping score of 94.41% is achieved on DRISHTI-GS which is significant for optic cup segmentation.

Improved OD and OC Fragmentation using Deep Learning for the Progression Recognition of Glaucoma

“…Here, the performance of OD and OC fragmentation using multi-label Au-net has been evaluated. The public dataset DRISHTI-GS1 [20,21] has been used for this experiment. This dataset has 101 images.…”

Segmenting the Optic Disc Using a Deep Learning Ensemble Model Based on OWA Operators

“…The fusion of the predictions of individual models can improve the segmentation accuracy [8]. In [9], a CNN model was trained using probability masks instead of binary masks to segment OD and OC. Each probability mask was obtained by fusing segmentation masks made by multiple experts.…”