Emergence of Large Language Models: Robustness and Adverse Effects in Medical Imaging

Large Language Models (LLMs) have revolutionized natural language processing, with emerging applications in medical imaging diagnostics through multimodal integration.This paper explores the robustness of LLMs when applied to medical imaging tasks, their potential adverse effects, including biases and misinformation, and the implications for clinical deployment.We analyze recent advancements, evaluate the challenges in ensuring safe and reliable use, and propose future research directions to mitigate risks.