Generative AI-Enhanced Framework for Predicting and Explaining Adverse Drug Reactions (ADR): A Multi-Modal Machine Learning and LLM Integration Approach

Generative AI enables multimodal framework to predict and explain the occurrence of adverse drug reaction (ADR). The framework exploits the capability of integrating (Large Language) LLMs with visual cues to enhance the classification accuracy and produce clinically interpretable results. Ads shown to be cross-lingual categorised and less dependent on second language boundary phenomena within the contexts of this work indicate potential for real world pharmacovigilance applications.