Boundary-Aware Transformer for Optic Cup and Disc Segmentation in Fundus Images

Segmentation of the Optic Disc (OD) and Optic Cup (OC) boundaries in fundus images is a critical step for early glaucoma diagnosis, but accurate segmentation is challenging due to low boundary contrast and significant anatomical variability. To address these challenges, this study proposes a novel s...

Full description

Saved in:
Bibliographic Details
Main Authors: Soohyun Wang, Byoungkug Kim, Doo-Seop Eom
Format: Article
Language:English
Published: MDPI AG 2025-05-01
Series:Applied Sciences
Subjects:
Online Access:https://www.mdpi.com/2076-3417/15/9/5165
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850030623107842048
author Soohyun Wang
Byoungkug Kim
Doo-Seop Eom
author_facet Soohyun Wang
Byoungkug Kim
Doo-Seop Eom
author_sort Soohyun Wang
collection DOAJ
description Segmentation of the Optic Disc (OD) and Optic Cup (OC) boundaries in fundus images is a critical step for early glaucoma diagnosis, but accurate segmentation is challenging due to low boundary contrast and significant anatomical variability. To address these challenges, this study proposes a novel segmentation framework that integrates structure-preserving data augmentation, Boundary-aware Transformer Attention (BAT), and Geometry-aware Loss. We enhance data diversity while preserving vascular and tissue structures through truncated Gaussian-based sampling and colormap transformations. BAT strengthens boundary recognition by globally learning the inclusion relationship between the OD and OC within the skip connection paths of U-Net. Additionally, Geometry-aware Loss, which combines the normalized Hausdorff Distance with the Dice Loss, reduces fine-grained boundary errors and improves boundary precision. The proposed model outperforms existing state-of-the-art models across five public datasets—DRIONS-DB, Drishti-GS, REFUGE, G1020, and ORIGA—and achieves Dice scores of 0.9127 on Drishti-GS and 0.9014 on REFUGE for OC segmentation. For joint segmentation of the OD and OC, it attains high Dice scores of 0.9892 on REFUGE, 0.9782 on G1020, and 0.9879 on ORIGA. Ablation studies validate the independent contributions of each component and demonstrate their synergistic effect when combined. Furthermore, the proposed model more accurately captures the relative size and spatial alignment of the OD and OC and produces smooth and consistent boundary predictions in clinically significant regions such as the region of interest (ROI). These results support the clinical applicability of the proposed method in medical image analysis tasks requiring precise, boundary-focused segmentation.
format Article
id doaj-art-00be52e84cdd4cfa9abc2fd28a2726a1
institution DOAJ
issn 2076-3417
language English
publishDate 2025-05-01
publisher MDPI AG
record_format Article
series Applied Sciences
spelling doaj-art-00be52e84cdd4cfa9abc2fd28a2726a12025-08-20T02:59:11ZengMDPI AGApplied Sciences2076-34172025-05-01159516510.3390/app15095165Boundary-Aware Transformer for Optic Cup and Disc Segmentation in Fundus ImagesSoohyun Wang0Byoungkug Kim1Doo-Seop Eom2AI Development Team, Sensorway, 140 Tongil-ro, Deogyang-gu, Goyang-si 10594, Republic of KoreaDivision of Computer Science and Engineering, Sahmyook University, 815 Hwarang-ro, Nowon-gu, Seoul 01795, Republic of KoreaInstitute of Convergence Science, Korea University, 145 Anam-ro, Seongbuk-gu, Seoul 02841, Republic of KoreaSegmentation of the Optic Disc (OD) and Optic Cup (OC) boundaries in fundus images is a critical step for early glaucoma diagnosis, but accurate segmentation is challenging due to low boundary contrast and significant anatomical variability. To address these challenges, this study proposes a novel segmentation framework that integrates structure-preserving data augmentation, Boundary-aware Transformer Attention (BAT), and Geometry-aware Loss. We enhance data diversity while preserving vascular and tissue structures through truncated Gaussian-based sampling and colormap transformations. BAT strengthens boundary recognition by globally learning the inclusion relationship between the OD and OC within the skip connection paths of U-Net. Additionally, Geometry-aware Loss, which combines the normalized Hausdorff Distance with the Dice Loss, reduces fine-grained boundary errors and improves boundary precision. The proposed model outperforms existing state-of-the-art models across five public datasets—DRIONS-DB, Drishti-GS, REFUGE, G1020, and ORIGA—and achieves Dice scores of 0.9127 on Drishti-GS and 0.9014 on REFUGE for OC segmentation. For joint segmentation of the OD and OC, it attains high Dice scores of 0.9892 on REFUGE, 0.9782 on G1020, and 0.9879 on ORIGA. Ablation studies validate the independent contributions of each component and demonstrate their synergistic effect when combined. Furthermore, the proposed model more accurately captures the relative size and spatial alignment of the OD and OC and produces smooth and consistent boundary predictions in clinically significant regions such as the region of interest (ROI). These results support the clinical applicability of the proposed method in medical image analysis tasks requiring precise, boundary-focused segmentation.https://www.mdpi.com/2076-3417/15/9/5165fundus imageoptic discoptic cupboundary-aware transformer attentiongeometry-aware lossstructure-preserving data augmentation
spellingShingle Soohyun Wang
Byoungkug Kim
Doo-Seop Eom
Boundary-Aware Transformer for Optic Cup and Disc Segmentation in Fundus Images
Applied Sciences
fundus image
optic disc
optic cup
boundary-aware transformer attention
geometry-aware loss
structure-preserving data augmentation
title Boundary-Aware Transformer for Optic Cup and Disc Segmentation in Fundus Images
title_full Boundary-Aware Transformer for Optic Cup and Disc Segmentation in Fundus Images
title_fullStr Boundary-Aware Transformer for Optic Cup and Disc Segmentation in Fundus Images
title_full_unstemmed Boundary-Aware Transformer for Optic Cup and Disc Segmentation in Fundus Images
title_short Boundary-Aware Transformer for Optic Cup and Disc Segmentation in Fundus Images
title_sort boundary aware transformer for optic cup and disc segmentation in fundus images
topic fundus image
optic disc
optic cup
boundary-aware transformer attention
geometry-aware loss
structure-preserving data augmentation
url https://www.mdpi.com/2076-3417/15/9/5165
work_keys_str_mv AT soohyunwang boundaryawaretransformerforopticcupanddiscsegmentationinfundusimages
AT byoungkugkim boundaryawaretransformerforopticcupanddiscsegmentationinfundusimages
AT dooseopeom boundaryawaretransformerforopticcupanddiscsegmentationinfundusimages