A Survey on Advancing Medical Diagnostics and Healthcare with Generative AI: Techniques, Applications, and Future Prospects

Generative artificial intelligence (AI) is reshaping modern healthcare by augmenting diagnostic accuracy, accelerating imaging workflows, and enabling data-driven clinical decisions. This paper surveys the recent evolution and clinical integration of generative models—such as GANs, diffusion architectures, and transformer-based systems—in medical diagnostics, with a focus on radiology. Across 219 peer-reviewed studies and 47 clinical implementations, these models demonstrate improvements in image clarity, artifact reduction, and rare disease detection, while reducing documentation burden and enhancing operational efficiency. The review also examines challenges including adversarial vulnerabilities, hallucination risks, and computational constraints, alongside policy recommendations for safe deployment. Quantitative analyses reveal that these tools can reduce diagnostic errors by up to 41% and contribute to projected cost savings of $362 billion annually by 2030. Emphasis is placed on ethical frameworks, federated data strategies, and emerging trends such as quantum generative modeling and neuro-symbolic systems, establishing generative AI as a pivotal enabler of precision medicine.