Integrating Edge Information into Ground Truth for the Segmentation of the Optic Disc and Cup from Fundus Images
Yoga Sri Varshan, Hitesh Gupta Kattamuri, Subin Sahayam, Umarani Jayaraman
TL;DR
This paper addresses boundary-focused optic disc and cup segmentation in fundus images by introducing edge information into the ground truth. It extracts optic disc and cup edges via a 2D Laplacian, stacks these edge maps with region masks to form a multi-channel ground truth, and trains U-Net variants with a focal loss biased toward boundary targets. Across REFUGE and Drishti-GS datasets, edge-informed models yield substantial gains in Dice scores and reduced Hausdorff distances, demonstrating improved boundary fidelity over baseline architectures. The approach is simple, robust, and broadly applicable to 2D segmentation tasks requiring precise boundaries, with activation maps suggesting clearer boundary-aware learning as the mechanism behind the improvements.
Abstract
Optic disc and cup segmentation helps in the diagnosis of glaucoma, myocardial infarction, and diabetic retinopathy. Most deep learning methods developed to perform segmentation tasks are built on top of a U-Net-based model architecture. Nevertheless, U-Net and its variants have a tendency to over-segment/ under-segment the required regions of interest. Since the most important outcome is the value of cup-to-disc ratio and not the segmented regions themselves, we are more concerned about the boundaries rather than the regions under the boundaries. This makes learning edges important as compared to learning the regions. In the proposed work, the authors aim to extract both edges of the optic disc and cup from the ground truth using a Laplacian filter. Next, edges are reconstructed to obtain an edge ground truth in addition to the optic disc-cup ground truth. Utilizing both ground truths, the authors study several U-Net and its variant architectures with and without optic disc and cup edges as target, along with the optic disc-cup ground truth for segmentation. The authors have used the REFUGE benchmark dataset and the Drishti-GS dataset to perform the study, and the results are tabulated for the dice and the Hausdorff distance metrics. In the case of the REFUGE dataset, the optic disc mean dice score has improved from 0.7425 to 0.8859 while the mean Hausdorff distance has reduced from 6.5810 to 3.0540 for the baseline U-Net model. Similarly, the optic cup mean dice score has improved from 0.6970 to 0.8639 while the mean Hausdorff distance has reduced from 5.2340 to 2.6323 for the same model. Similar improvement has been observed for the Drishti-GS dataset as well. Compared to the baseline U-Net and its variants (i.e) the Attention U-Net and the U-Net++, the models that learn integrated edges along with the optic disc and cup regions performed well in both validation and testing datasets.