Generative Artificial Intelligence in Healthcare: A Systematic Review of GANs, Diffusion Models, Large Language Models, and Variational Autoencoders for Medical Applications
This systematic review analyzes 87 original research articles to provide a comprehensive overview of how GenAI models including Generative Adversarial Networks, Diffusion Models, Large Language Models, and Variational Autoencoders—are applied across various healthcare domains.
Abstract
Generative Artificial Intelligence (GenAI) is rapidly transforming the healthcare landscape by enabling novel solutions in areas such as medical imaging, drug discovery, and synthetic data generation. This systematic review analyzes 87 original research articles to provide a comprehensive overview of how GenAI models including Generative Adversarial Networks (GANs), Diffusion Models, Large Language Models (LLMs), and Variational Autoencoders (VAEs)—are applied across various healthcare domains. We investigate key aspects such as the most frequently utilized generative models, their primary applications, the datasets that support their development, and the evaluation metrics used to measure their performance. Our analysis reveals that GANs, Diffusion Models, LLMs, and VAEs are the dominant GenAI architectures employed in current healthcare research. Each included study is summarized to highlight its core contributions, offering valuable insights into the practical use of GenAI in clinical and biomedical settings. By focusing exclusively on peer-reviewed original research, this review ensures the rigor and relevance of its findings. The study serves as a foundational resource for researchers and practitioners, outlining the current landscape and identifying promising directions for future investigation in GenAI-powered healthcare innovation.