Pocket Ophthalmic Clinic – Ai-Powered Smartphone Application For Remote Eye Disease Detection And Management
Published 2025 - 43rd Congress of the ESCRS
Reference: PP18.04 | Type: Poster | DOI: 10.82333/30es-fp69
Authors: Shima Eghtedari 1 , Farhad Nejat* 2
1Biomedical engineering,Vision health research clinic,Tehran,Iran, Islamic Republic Of, 2Ophthalmology,Vision health research clinic,Tehran,Iran, Islamic Republic Of
Purpose
A-EYES is an AI-based mobile application designed to revolutionize ophthalmic care by enabling smartphone-based diagnosis and management of eye diseases. It aims to provide an accessible, accurate, and cost-effective solution for detecting conditions like dry eye disease, keratoconus, pterygium, conjunctival nevus, strabismus, amblyopia and cataract. The app minimals the need for specialized tools and ensures rapid screening, offering patients and healthcare providers a reliable teleophthalmology tool. Our mission is to bridge the gap between eye care accessibility and early disease detection through AI-driven mobile health solutions anywhere, anytime.
Setting
For each disease, we gathered over 1,000 smartphone images using a Samsung mobile phone from patients at Vision health research clinic, ensuring diverse representation. All images were collected with patient consent and ethical approval from the Semnan University of Medical Sciences ethics committee. This dataset forms the foundation for training and validating our deep learning models, enhancing its accuracy and generalizability in real-world ophthalmic applications.
Methods
A-EYES utilizes AI deep learning models trained on labeled eye images to detect ophthalmic conditions. Users capture images of their eyes using their smartphone cameras, and the app processes them through a pre-trained convolutional neural network (CNN). The algorithm analyzes features indicative of specific ocular surface diseases and provides a probability-based diagnosis. Additionally, A-EYES includes smartphone-based vision tests, such as visual acuity, color vision, and Amsler grid tests, ensuring a comprehensive ophthalmic evaluation. The model is validated using diverse clinical datasets.
Results
Preliminary testing of A-EYES demonstrated high accuracy in detecting dry eye disease, keratoconus, and pterygium, cataract, strabismus, amblyopia with AI predictions closely matching ophthalmologist diagnoses. The app successfully identified pathological patterns from smartphone images with an accuracy exceeding 90% for every eye conditions. User feedback highlighted the app’s ease of use, rapid response time, and potential to reduce unnecessary clinic visits enhance sustainability. Ongoing improvements and clinical trials aim to enhance model robustness and expand disease early detection and management and also monitoring in time.
Conclusions
A-EYES represents a breakthrough in mobile ophthalmology, offering a reliable AI-driven solution for early eye disease detection and management. By enabling smartphone-based screening without additional hardware, it enhances accessibility to eye care anytime, anywhere, particularly in remote and underserved areas. The app has the potential to transform teleophthalmology, reducing diagnostic delays and supporting timely interventions. Future developments will refine AI accuracy, expand diagnostic capabilities, and integrate with telemedicine platforms.