Related Experiment Video
Updated: Sep 24, 2025
Assessing Early Stage Open-Angle Glaucoma in Patients by Isolated-Check Visual Evoked Potential
Published on: May 25, 2020
Artificial Intelligence for Screening of Multiple Retinal and Optic Nerve Diseases
Li Dong1, Wanji He2, Ruiheng Zhang1
1Beijing Tongren Eye Center, Beijing Key Laboratory of Intraocular Tumor Diagnosis and Treatment, Beijing Ophthalmology and Visual Sciences Key Lab, Medical Artificial Intelligence Research and Verification Key Laboratory of the Ministry of Industry and Information Technology, Beijing Tongren Hospital, Capital Medical University, Beijing, China.
Artificial intelligence accurately screens 10 retinal diseases using ocular fundus images. This deep learning system, RAIDS, offers a faster, more sensitive alternative to ophthalmologists, especially in underserved regions.
Area of Science:
- Ophthalmology
- Medical Imaging
- Artificial Intelligence
Background:
- Limited availability of experienced ophthalmologists hinders early diagnosis of retinal diseases.
- Artificial intelligence (AI) presents a potential solution for efficient, real-time retinal disease screening.
Purpose of the Study:
- To develop and validate a deep learning (DL) algorithm for simultaneous recognition of multiple retinal diseases from ocular fundus images in clinical practice.
Main Methods:
- Developed the Retinal Artificial Intelligence Diagnosis System (RAIDS) using 120,002 ocular fundus photographs to identify 10 retinal diseases.
- Validated RAIDS on a prospective dataset and compared its performance against ophthalmologists using data from the Beijing Eye Study and Kailuan Eye Study.
Main Results:
- RAIDS demonstrated high sensitivity (89.8%) in detecting any of 10 retinal diseases in a prospective dataset of 208,758 images.
- RAIDS achieved accuracies from 95.3% to 99.9% for differentiating 10 retinal diseases, outperforming certified ophthalmologists in sensitivity for 7 diseases.
- RAIDS required 96-97% less time for image assessment compared to ophthalmologists.
Conclusions:
- The DL system (RAIDS) accurately distinguishes 10 retinal diseases in real time.
- This AI technology can help mitigate the shortage of experienced ophthalmologists, particularly in underdeveloped areas.
