Related Experiment Video
Updated: Jun 23, 2026
04:48
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
Attention based multi-scale edge-aware segmentation and convolutional transformer framework for automated glaucoma
C Moorthy1, D Arulanantham2, A Suresh Babu3
1Department of ECE, Dr. Mahalingam College of Engineering and Technology, Pollachi, Tamilnadu, India.
Journal of X-Ray Science and Technology
|June 22, 2026
Summary
This study introduces a novel framework for automated glaucoma detection using an Attention-guided Multi-scale Edge-aware Segmentation Network (AME-SegNet). The system achieves high accuracy in segmenting optic discs and cups, enabling reliable early glaucoma diagnosis.
Area of Science:
- Ophthalmology
- Medical Imaging
- Artificial Intelligence
Background:
- Glaucoma is a primary cause of irreversible vision loss.
- Current automated systems struggle with boundary delineation and generalizability.
- Subtle optic disc and optic cup changes characterize glaucoma.
Purpose of the Study:
- Develop a unified, anatomically guided framework for automated glaucoma detection.
- Enhance accuracy and reliability in diagnosing glaucoma from fundus images.
- Improve upon existing automated detection system limitations.
Main Methods:
- Contrast-enhanced preprocessing for improved image quality.
- Attention-guided Multi-scale Edge-aware Segmentation Network (AME-SegNet) for optic disc/cup segmentation.
- Bitterling Colony Optimization (BCO) and Convolutional Transformer (CT) for feature selection and classification.
- Honey Badger Algorithm (HBA) for automatic parameter tuning.
Main Results:
- High segmentation performance: 97.36% (optic disc) and 96.72% (optic cup) Dice scores on Drishti-GS1.
- Excellent classification accuracy: 98.63% (RIM-ONE) and 98.96% (ORIGA-Light).
- Demonstrated strong generalization capability across multiple datasets.
Conclusions:
- The proposed framework shows robust performance and high accuracy.
- Effective for automated glaucoma screening and early diagnosis.
- Highlights potential for clinical application in reliable glaucoma detection.