In a paper published in National Library of Medicine by Ping Yu,Hua Xu, Xia Hu, and Chao Deng showcases the transformative potential of generative AI and LLMs in healthcare and medicine.
Generative artificial intelligence (AI) and the utilization of large language models (LLMs), as seen in the example of ChatGPT, hold substantial promise in the transformation of data and information management within the healthcare and medical sectors. However, there exists a notable shortage of literature aimed at guiding non-AI professionals in effectively incorporating these technologies. To address this crucial need, this study embarks on a comprehensive scoping literature review, intending to provide insights into the integration of generative AI and LLMs into healthcare and medical practices.
The review elucidates the distinctive mechanisms that underlie these technologies, including Reinforcement Learning from Human Feedback (RLFH), encompassing concepts like few-shot learning and chain-of-thought reasoning. These differentiating factors set them apart from traditional rule-based AI systems. The study emphasizes the requirement for an inclusive and collaborative co-design process that actively involves all relevant stakeholders, encompassing clinicians and consumers, to unlock the transformative potential of these advancements.
While global research explores the opportunities and challenges presented by these technologies, including ethical and legal dimensions, LLMs offer the prospect of significant advancements in healthcare. These advancements span across the realms of data management, information retrieval, and the decision-making processes critical to the field. To fully harness the potential of these technologies, it is imperative to continually innovate in areas like data acquisition, model fine-tuning, prompt strategy development, evaluation, and system implementation.
Healthcare organizations are encouraged to proactively engage with these technologies, with the aim of enhancing the quality, safety, and efficiency of healthcare services while upholding ethical and legal guidelines to ensure responsible and beneficial application.
