Related Experiment Video
Updated: Jun 21, 2026
06:16
LipidUNet-Machine Learning-Based Method of Characterization and Quantification of Lipid Deposits Using iPSC-Derived Retinal Pigment Epithelium
Published on: July 28, 2023
Automated quantification of lens epithelial cell density using deep learning: validation and large-scale clinical
Poramaporn Luangprasert1,2, Chutimon Sindhuprama3, Piyorod Srisawad3
1Department of Ophthalmology, Ramathibodi Hospital, Mahidol University, Bangkok, 10400, Thailand.
Scientific Reports
|June 19, 2026
Summary
An AI tool accurately quantifies lens epithelial cell density (LECD) from micrographs. This automated method for counting lens epithelial cells (LECs) is more reliable than manual methods, aiding cataract research.
Area of Science:
- Ophthalmology
- Biomedical Engineering
- Artificial Intelligence
Background:
- Lens epithelial cells (LECs) are vital for maintaining lens transparency.
- Lens epithelial cell density (LECD) is hypothesized to correlate with cataract formation and severity.
- Current methods for quantifying LECD are limited, hindering research.
Purpose of the Study:
- To develop an AI-driven model for automated LEC counting.
- To assess LECD, distribution, and influencing factors using AI.
- To compare AI-based cell counting with traditional manual methods.
Main Methods:
- Developed AI software using deep learning for LEC enumeration from light micrographs of anterior lens capsules.
- Validated the AI model against manual cell counting.
- Employed statistical analyses including ANOVA, Bland-Altman, MAPE, and ICC for performance evaluation.
- Compared AI results with inter-observer agreement between expert graders.
Main Results:
- The AI model demonstrated high accuracy (98.1%) and excellent agreement with manual counts (ICC=0.994).
- AI agreement surpassed inter-observer agreement (ICC=0.972) among experts.
- The AI tool provided a rapid, objective, and repeatable method for LECD analysis.
- Over 43,000 cells across 20 images were analyzed.
Conclusions:
- AI-based automated LEC counting is a reliable and efficient alternative to manual methods.
- This AI tool can advance research into the relationship between LECD and cataracts.
- The software offers an objective and repeatable approach for LECD analysis in ophthalmology.
