Related Experiment Video
Updated: Jun 27, 2025
10:14
Author Spotlight: Ex Vivo OCT-Based Multimodal Imaging of Human Donor Eyes for Research into Age-Related Macular Degeneration
Published on: May 26, 2023
3.1K
Progression or Aging? A Deep Learning Approach for Distinguishing Glaucoma Progression From Age-Related Changes in
Sayan Mandal1, Alessandro A Jammal2, Davina Malek3
1From the Department of Electrical and Computer Engineering, Pratt School of Engineering (S.M., F.A.M.), Duke University, Durham, North Carolina, USA.
American Journal of Ophthalmology
|May 4, 2024
Summary
A novel deep learning algorithm effectively detects glaucoma progression using optical coherence tomography (OCT) scans, outperforming traditional methods. This advancement aids in identifying structural changes without a reference standard.
Area of Science:
- Ophthalmology
- Medical Imaging
- Artificial Intelligence
Background:
- Glaucoma is a leading cause of irreversible blindness.
- Accurate detection of glaucoma progression is crucial for timely intervention.
- Optical coherence tomography (OCT) provides high-resolution imaging of the optic nerve and retinal nerve fiber layer.
Purpose of the Study:
- To develop and validate a deep learning (DL) algorithm for detecting glaucoma progression using OCT images.
- To address the challenge of identifying glaucoma progression in the absence of a traditional reference standard.
Main Methods:
- A retrospective cohort study involving eyes with glaucoma and healthy controls.
- Development of a weakly supervised time-series learning model (Noise-PU DL) using convolutional neural networks (CNN) and long short-term memory (LSTM) networks.
- The DL model was trained to identify glaucoma progression, accounting for age-related changes and test-retest variability.
Main Results:
- The study included 8,785 follow-up sequences from 3,253 eyes.
- The DL model achieved a hit ratio of 0.498, significantly outperforming ordinary least squares (OLS) regression (0.284) at 95% specificity.
- This indicates superior sensitivity in detecting glaucoma progression.
Conclusions:
- A deep learning model successfully identified longitudinal structural changes indicative of glaucoma progression in OCT B-scans.
- The developed DL algorithm offers a promising tool for glaucoma monitoring, even without a definitive reference standard.
