Related Experiment Video
Updated: May 6, 2026
07:51
Full-Field Optical Coherence Microscopy for Histology-Like Analysis of Stromal Features in Corneal Grafts
Published on: October 21, 2022
1.6K
MCOA: A Comprehensive Multimodal Dataset for Advancing Deep Learning in Corneal Opacity Assessment
Xinyu Ma1, Jianxia Fang1, Yaqi Wang2
1Eye Center, The Second Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, Zhejiang, China.
Scientific Data
|May 30, 2025
Summary
This study created the largest corneal opacity dataset, combining anterior segment optical coherence tomography (AS-OCT) and photographs. This resource advances AI for diagnosing vision impairment and personalizing treatments.
Area of Science:
- Ophthalmology
- Medical Imaging
- Artificial Intelligence
Background:
- Corneal opacity is a leading cause of vision impairment globally.
- Current assessment relies on subjective slit lamp examinations.
- Anterior segment optical coherence tomography (AS-OCT) offers detailed structural insights.
Purpose of the Study:
- To address the lack of large-scale datasets for AI development in corneal opacity.
- To create a comprehensive dataset for improved corneal opacity recognition.
- To facilitate AI-driven clinical decision-making and personalized treatment strategies.
Main Methods:
- Compilation of the most extensive corneal opacity dataset to date.
- Inclusion of 6,272 AS-OCT images and 392 anterior segment photographs.
- Detailed annotation of images with cornea and opacity information.
Main Results:
- Established the largest available dataset for corneal opacity research.
- Dataset combines high-resolution AS-OCT and photographic data.
- Annotations provide detailed information for AI model training.
Conclusions:
- The dataset is a significant advancement for AI in ophthalmology.
- Enables development of deep learning algorithms for corneal opacity recognition.
- Supports AI-driven clinical decisions and personalized patient care.
