RetinalGPT is a specialized conversational assistant for retinal image analysis, built on multimodal large language models (MLLMs) to address the limitations of existing medical AI systems.

Key Innovations:

• Performs quantitative analysis of retinal images beyond what general medical MLLMs can achieve
• Designed specifically for clinical retinal applications, improving diagnostic support
• Bridges the gap between general AI assistants and the specialized needs of ophthalmologists
• Demonstrates how domain-specific AI can outperform general-purpose models in specialized healthcare contexts

Clinical Impact: RetinalGPT represents a significant step forward in AI-assisted ophthalmology, potentially improving diagnostic accuracy, reducing the burden on specialists, and increasing access to expert-level analysis in underserved areas.

RetinalGPT: A Retinal Clinical Preference Conversational Assistant Powered by Large Vision-Language Models