DRDD: AI-Powered Multi-Modal System for Diabetic Retinopathy Disease Detection

The disease named diabetic retinopathy destroys the eye's light sensitive tissue and cause blindness. It often occurs due to diabetics. This disease can be detected by using automated system that trained using machine learning algorithms with the needed dataset. In many approached system, majorly retinal images were only used for the detection of this disease, meanwhile our work focus on integrating the result obtained from both retinal images and clinical data of the patients. This enhances the accuracy rate and efficiency of diagnosis. This study introduces the fusion model that integrates deep learning and data analysis to enhance the accuracy level in detection. This model uses EfficientNet-B3 and Vision Transformer to find abnormalities in eye from eye image and TabNet for analyzing clinical data and improve the accuracy of the result. Generative AI model powered by GPT-4 to provide a personalized suggestion to patients based on their health condition. This integrated approach achieves an accuracy of 94.7% in disease detection.