Ocularis-Ensuring Healthy Vision Through Intelligent Detection

Ocular diseases can cause vision impairment or blindness if not detected early. “Ocular Disease Intelligent Recognition (Ocularis)” is a multimodal system designed for automated ocular disease detection and classification. It utilizes a dataset of 5,000 patient records, incorporating age, diagnostic keywords, and high-resolution fundus images captured from diverse imaging devices. The model focuses on classifying eight primary disease categories. To enhance diagnostic accuracy, ensemble learning techniques are employed by integrating ResNet, EfficientNet, and VGG19, leveraging their strengths for robust classification. Additionally, SHAP-based Explainable AI (XAI) provides interpretable visualizations of model predictions, aiding clinical validation. A Google Gemini LLM-powered recommendation system further assists in generating personalized insights based on the detected condition. Ocularis is designed to be a scalable and reliable screening tool for real-world healthcare applications. Future enhancements include predictive analytics for disease progression and extending the system to broader ophthalmic applications, improving early detection and clinical decision-making.