Related Experiment Video
Updated: Jan 13, 2026
07:06
Binocular Dynamic Visual Acuity in Eyeglass-Corrected Myopic Patients
Published on: March 29, 2022
3.2K
An Automated Diagnosis of Myopia from an Optic Disc Image Using YOLOv11: A Feasible Approach for Non-Expert ECPs in
Nicola Rizzieri1, Luca Dall'Asta2, Maris Ozoliņš1,3
1Department of Optometry and Vision Science, The Faculty of Science and Technology, University of Latvia, Jelgavas Street 1, LV-1004 Riga, Latvia.
Life (Basel, Switzerland)
|October 29, 2025
Summary
This study developed an automated AI tool using deep learning to detect myopia from fundus images by analyzing the optic disc. The accessible system aids early myopia diagnosis for eye care practitioners, improving patient outcomes.
Area of Science:
- Ophthalmology
- Artificial Intelligence
- Medical Imaging
Background:
- Myopia prevalence is increasing globally, necessitating early detection to prevent vision loss.
- Current diagnostic methods may not be universally accessible or scalable.
- Automated tools can enhance early myopia screening in diverse settings.
Purpose of the Study:
- To develop an accessible, automated deep learning system for myopia detection from fundus photographs.
- To analyze the optic disc region for myopia indicators using AI.
- To create a tool usable by eye care practitioners without computer science expertise.
Main Methods:
- Utilized a deep learning model based on the YOLO (You Only Look Once) architecture (versions 8 and 11).
- Developed a pipeline to extract the optic disc from fundus images using a custom-trained YOLOv8 model.
- Implemented a classification algorithm to determine myopia presence based on optic disc features.
Main Results:
- The system achieved high diagnostic accuracy, sensitivity, and F1 scores in myopia detection.
- Lightweight models like YOLOv11-nano demonstrated performance comparable to larger variants (AUC 97.5% vs. 97.3%).
- The AI tool proved effective in identifying myopia from optic disc analysis.
Conclusions:
- AI-based screening tools can be feasibly integrated into clinical practice for myopia detection.
- The developed system offers a scalable and cost-effective solution for early myopia diagnosis.
- This approach empowers eye care practitioners with advanced diagnostic capabilities without requiring specialized technical skills.
