Related Experiment Video
Updated: Aug 8, 2025
07:32
Standardized Histomorphometric Evaluation of Osteoarthritis in a Surgical Mouse Model
Published on: May 6, 2020
12.2K
Measurement method of tear meniscus height based on deep learning.
Cheng Wan1, Rongrong Hua1, Ping Guo2,3
1College of Electronic Information Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing, China.
Frontiers in Medicine
|March 3, 2023
Summary
A new deep learning algorithm automatically measures tear meniscus height (TMH), a key indicator for dry eye disease. This automated method offers high accuracy and consistency, aiding clinical diagnosis.
Area of Science:
- Ophthalmology
- Medical Imaging
- Artificial Intelligence
Background:
- Tear meniscus height (TMH) is crucial for diagnosing dry eye disease.
- Traditional TMH measurement methods are manual, subjective, time-consuming, and labor-intensive.
Purpose of the Study:
- To develop an automated TMH measurement method using deep learning and image processing.
- To improve the accuracy and efficiency of TMH assessment for dry eye diagnosis.
Main Methods:
- A deep learning segmentation algorithm based on the DeepLabv3 architecture, incorporating elements of ResNet50, GoogleNet, and FCN networks.
- Training and evaluation using 305 ocular surface images, with performance metrics including intersection over union and dice coefficient.
- Comparison of automated TMH measurements against manual measurements using linear regression.
Main Results:
- Achieved high segmentation performance for the tear meniscus (IoU: 0.896, Dice: 0.884) and corneal projection ring (IoU: 0.932, Dice: 0.926).
- The proposed model demonstrated superior performance compared to existing segmentation models.
- Automated TMH measurements showed high consistency with manual measurements (r²=0.94).
Conclusions:
- The developed deep learning algorithm enables accurate and automated measurement of TMH.
- This method significantly assists clinicians in the diagnosis of dry eye disease.
- The approach offers a more objective, efficient, and reliable alternative to traditional TMH measurement techniques.
