Related Experiment Video
Updated: Jan 8, 2026
08:13
In Vivo Confocal Microscopy in the Diagnosis and Management of Dry Eye: A Focus on Imaging Protocols and Interpretation
Published on: November 11, 2025
404
Multicentre Pixel-Level Tear Meniscus Segmentation Dataset with Multimodal Imaging for Dry Eye Diagnosis
Xiaoyu Chen1, Kesheng Wang2, Kunhui Xu1
1National Clinical Research Center for Ocular Diseases, Eye Hospital, Wenzhou Medical University, Wenzhou, China.
Scientific Data
|December 23, 2025
Summary
This study introduces a novel multimodal dataset for lower tear meniscus segmentation, crucial for diagnosing dry eye disease. This AI-ready dataset will advance automated tear secretion analysis and dry eye research.
Area of Science:
- Ophthalmology
- Medical Imaging
- Artificial Intelligence
Background:
- Dry eye disease involves tear abnormalities, impacting quality, quantity, and fluid dynamics.
- Lower tear meniscus height (TMH) is a key indicator of tear secretion and stability, typically measured manually.
- Accurate TMH assessment is vital for dry eye diagnosis, but manual measurement is time-consuming.
Purpose of the Study:
- To introduce a multicentre, multimodal, pixel-level dataset for lower tear meniscus segmentation.
- To facilitate the development and validation of artificial intelligence (AI) models for automated TMH analysis.
- To support the creation of standardized medical image databases for dry eye research.
Main Methods:
- Collected 1,693 color and 1,739 infrared images from five centers in China.
- Developed a human-computer interactive approach for generating pixel-level segmentation labels.
- Created a multimodal dataset comprising images and corresponding segmentation masks.
Main Results:
- Introduced the first publicly available multimodal tear meniscus segmentation dataset.
- The dataset includes diverse imaging modalities (color and infrared) and detailed segmentation labels.
- The dataset comprises 1,693 color and 1,739 infrared images with expert-generated labels.
Conclusions:
- The developed dataset is essential for training AI models for automated TMH segmentation.
- This resource is expected to advance research in dry eye diagnosis and treatment.
- The dataset will contribute to building standardized medical image databases for ophthalmic research.
