It is my great honor to welcome you to this collection of selected poster abstracts from the Mayo Clinic artificial intelligence (AI) summit, held in Rochester, Minnesota, July 2-8, 2025. The summit brought together clinicians, scientists, engineers, and thought leaders from across the world, united by a shared vision: to harness the power of AI for the advancement of medicine and the improvement of human health.
The 10 abstracts presented here exemplify the richness and diversity of that vision. They showcase how AI is being deployed across multiple domains—from radiology and pathology to molecular oncology and nephrology—each tackling distinct challenges, yet collectively contributing to a common goal of more precise, efficient, and equitable care. Importantly, these studies remind us that innovation in AI is not only about improving algorithms but also about thoughtfully embedding them into clinical workflows, enhancing interpretability, and ensuring that patient outcomes remain at the center of all progress.
Several of the posters highlight AI’s transformative role in medical imaging. One study report how deliberate prompt engineering can convert large language models into accurate extractors of structured findings from radiology reports, effectively bridging the gap between unstructured narratives and actionable clinical data. Another contribution illustrates how graph-based deep ensemble learning can overcome the heterogeneity of lung adenocarcinoma histology, improving both diagnostic accuracy and efficiency in pathology. These examples underscore how next-generation AI methods are redefining diagnostic practice in ways that align with, and augment, the expertise of health care professionals.
Equally compelling are the advances in molecular and multi-omics medicine. By leveraging sophisticated feature selection and predictive modeling strategies, one team demonstrates that complex cancers such as cholangiocarcinoma can be stratified with remarkable accuracy using only a minimal set of molecular markers. This kind of work not only accelerates precision oncology but also brings the promise of advanced molecular diagnostics closer to routine and accessible clinical use.
What unites all of these efforts is a focus on translation. Whether through refining prompts, integrating diverse neural architectures, or designing streamlined assays, the researchers featured here are reporting how AI can be responsibly advanced from the research bench to the patient bedside. Their attention to transparency, reproducibility, and clinical applicability reflects a field that is maturing and mindful of its responsibilities to patients, practitioners, and society.
The Mayo Clinic AI Summit has always sought to foster collaboration across disciplines and geographies. This year’s contributions reaffirm that the future of medical AI will be shaped not by isolated breakthroughs, but by shared innovation, collective expertise, and the courage to challenge convention. The work presented here provides both a snapshot of the current frontier and a glimpse of what lies ahead—a future where AI is seamlessly integrated into medicine, enabling us to deliver care that is more precise, proactive, and compassionate.
It is with pride and gratitude that I commend these abstracts to you. May they spark new ideas, inspire collaboration, and continue to advance the role of AI in fulfilling our ultimate mission: to meet the needs of the patient, first and foremost.
Potential Competing Interests
The author report no competing interests.