A research team at Osaka Metropolitan University's Graduate School of Medicine, led by Dr. Hirotaka Takita and Associate Professor Daiju Ueda, conducted a meta-analysis to evaluate the diagnostic capabilities of generative AI in medicine. The study, which analyzed 83 research papers published between June 2018 and June 2024, aimed to provide a comprehensive understanding of how generative AI performs compared to medical professionals. Among the various large language models (LLMs) studied, ChatGPT was the most frequently analyzed. The results showed that medical specialists had a 15.8% higher diagnostic accuracy than generative AI models. On average, generative AI achieved a diagnostic accuracy of 52.1%. However, the latest models of generative AI often demonstrated accuracy levels similar to those of non-specialist doctors. This finding suggests that while AI is not yet on par with specialists, it can serve as a valuable support tool for non-specialist doctors. The research highlights several potential applications of generative AI in the medical field. Dr. Takita emphasized that AI could be used in medical education to assist non-specialist doctors and to provide diagnostic support in areas with limited medical resources. This could be particularly beneficial in rural or underprivileged regions where specialist doctors are scarce. Despite the promising results, the researchers noted several areas that require further investigation. These include evaluating AI's performance in more complex clinical scenarios, using actual medical records for performance assessments, enhancing the transparency of AI decision-making processes, and verifying its effectiveness across diverse patient groups. The study was published in the journal npj Digital Medicine in April 2025. The meta-analysis synthesizes a wide range of research, providing a clearer picture of AI's role in diagnostics. However, it also underscores the need for ongoing refinement and validation to ensure AI can reliably support medical professionals in various settings. The researchers are confident that continued advancements and rigorous testing will improve AI's diagnostic accuracy and its integration into medical practices. The use of generative AI in diagnostics poses both opportunities and challenges. One of the key opportunities is its potential to reduce the diagnostic workload for non-specialist doctors, thereby improving the efficiency and quality of healthcare. However, concerns about the reliability and interpretability of AI decisions remain. Addressing these concerns is crucial for building trust among medical practitioners and patients. Another critical aspect is the ethical and regulatory framework surrounding AI in healthcare. As AI technology advances, regulatory bodies must ensure that these tools meet high standards of safety and efficacy. Additionally, training programs for healthcare professionals to effectively use and interpret AI-generated diagnoses will be essential. The study's findings have significant implications for the future of medical diagnostics. They suggest that generative AI could become a commonplace tool in medical settings, especially for non-specialist doctors and in areas with limited healthcare resources. However, the full potential of AI in diagnostics will only be realized through further research, development, and regulatory support. Industry insiders have praised the study for its comprehensive approach and the potential impact on medical education and resource-limited settings. Dr. Takita and his team's work at Osaka Metropolitan University highlights the institution's commitment to advancing the intersection of technology and healthcare, positioning it as a leader in the field of AI in medicine. Osaka Metropolitan University, established in 2022, is a public university in Japan dedicated to fostering interdisciplinary research and innovation. The university's Graduate School of Medicine is at the forefront of exploring how AI can enhance diagnostic capabilities and improve patient outcomes.