Related Experiment Video
Updated: Aug 13, 2025
08:22
Application of Optical Coherence Tomography to a Mouse Model of Retinopathy
Published on: January 12, 2022
4.2K
iERM: An Interpretable Deep Learning System to Classify Epiretinal Membrane for Different Optical Coherence
Kai Jin1, Yan Yan1, Shuai Wang2
1Department of Ophthalmology, The Second Affiliated Hospital of Zhejiang University, College of Medicine, Hangzhou 310009, China.
Journal of Clinical Medicine
|January 21, 2023
Summary
A new deep learning system, iERM, accurately grades epiretinal membranes (ERM) using optical coherence tomography (OCT) scans. This interpretable AI tool aids in clinical diagnosis and treatment decisions for ERM.
Area of Science:
- Ophthalmology
- Medical Imaging
- Artificial Intelligence
Background:
- Epiretinal membranes (ERM) are prevalent in individuals over 50.
- Assessing ERM severity from optical coherence tomography (OCT) images is challenging due to limited reliable analysis methods.
Purpose of the Study:
- To develop an interpretable, two-stage deep learning (DL) system named iERM for automatic ERM grading.
- To enhance the accuracy and clinical utility of ERM assessment using OCT imaging.
Main Methods:
- Trained the iERM system using human segmentation of key features for improved classification and interpretability.
- Utilized a dataset of 4547 OCT B-scans from four OCT devices across nine international medical centers.
Main Results:
- The iERM system demonstrated improved grading performance (1-5.9%) compared to traditional DL models.
- Achieved high accuracy rates of 82.9% (internal) and 87.0%, 79.4% (external), comparable to retinal specialists.
Conclusions:
- The iERM system offers a benchmark for improving DL model performance and interpretability in ERM grading.
- This AI-driven approach shows potential for precise guidance in ERM diagnosis and treatment planning.
