AI-Enabled Medical Chatbots: Advancements in Patient Query Handling and Automated Healthcare Delivery

Abstract

Artificial Intelligence (AI) is revolutionizing healthcare through the application of sophisticated Large Language Models (LLMs), facilitating rapid symptom assessment and enhanced disease identification. This study investigates the performance of multimodal LLMs, specifically llama-4-scout-17b-16e-instruct and llama-4-maverick-17b-128e-instruct, tailored for the analysis of medical images, alongside their counterparts optimized for text-based diagnostic support. These models were evaluated using real-time X-ray imagery and patient-reported symptom descriptions, with assessments focusing on diagnostic precision, response clarity, processing efficiency, and contextual richness. Findings reveal that vision-specialized models demonstrate high accuracy in image interpretation, though with relatively slower processing times, while text-oriented models provide lucid insights, with occasional limitations in handling intricate scenarios. By advancing real-time multimodal analysis independent of pre-existing datasets, this research underscores the potential of synergizing vision and text functionalities to enhance the accuracy and responsiveness of AI-driven chatbots, paving the way for scalable, effective healthcare interventions in practical settings.