How artificial intelligence is transforming biomedical research and education

Postdoctoral researcher and medical physicist Mikael Brix studies and develops novel imaging methods in the University of Oulu Medical Faculty to better patient diagnostics. Brix integrates AI into his research in multiple ways. Large language models serve as sparring partners for refining scientific writing and conducting literature reviews. In programming, he sometimes delegates routine coding tasks to AI, especially when the solution is already clear, thus saving time without compromising his own skill development.

From neural networks to critical thinking

Mikael Brix thinks that the most striking application of AI is in medical imaging, where convolutional neural networks are used for automatic organ segmentation. What once took days per patient can now be completed in seconds, dramatically accelerating research workflows and opening new possibilities for large-scale epidemiological studies.

Despite these advantages, Brix remains cautious. He emphasizes that AI’s impact depends heavily on how it’s used. “If you give AI too much control over your thought process, you risk ending up with exotic and questionable research designs,” he warns. He sees the researcher’s role as a critical filter, ensuring that AI enhances rather than distorts scientific inquiry.

Importantly, Brix believes that AI cannot replace the core of scientific motivation, the drive to understand and shape a field. Nor can it grasp the full complexity of a research project. “If you hand over the big picture to AI, you’ll eventually face chaos,” he says. He hopes future AI systems will better communicate their uncertainties, perhaps through Bayesian approaches, allowing researchers to make more informed decisions.

AI in structural biology and drug discovery

Lari Lehtiö, Professor of structural biology at the University of Oulu Faculty of Biochemistry and Molecular Medicine, has witnessed firsthand how AI has revolutionized the research field. His team uses AI to predict protein structures and complexes, a practice that has become routine since the release of AlphaFold2 in 2021. He emphasizes that these predictions don’t replace experimental methods, but they provide valuable hypotheses that guide research and inspire new experiments.

Professor Lehtiö is also involved in developing an AI-based tool for protein crystal recognition at Biocenter Oulu. This system helps researchers focus on the most promising samples, streamlining the process of experimental structure determination and improving data management.

The impact of AI in his field has been profound. “It’s changed how we do research,” Professor Lehtiö says. “We can now plan experiments based on highly accurate models, which was unthinkable a few years ago.” However, he stresses the importance of critical evaluation. AI models must be validated using biophysical methods, and researchers must be aware of their limitations to avoid overinterpretation.

Looking ahead, Lehtiö sees AI playing a major role in structure-based drug development. Predicting interactions between macromolecules and small molecules at scale could dramatically accelerate early-stage drug discovery. But even here, he advises caution: “These models are powerful, but they’re still just tools. We need to understand them deeply to use them wisely.”

Human-centered AI in healthcare education

Kristina Mikkonen is a professor of nursing science (PhD, RN) and leads a research group of Health care education and competence (HealthEduCom) in the Research unit of health sciences and technology. She brings a unique perspective to AI, focusing on its role in healthcare education and interdisciplinary innovation. Professor Mikkonen uses AI to analyze large datasets, support systemic thinking, and design immersive learning environments. These include extended reality (XR) simulations where students and professionals can practice clinical decision-making in safe, interactive settings.

AI also plays a role in language learning and integration, helping culturally and linguistically diverse healthcare professionals adapt to new environments. In teaching, Professor Mikkonen uses AI-driven tools to create hybrid scenarios that foster teamwork and prepare learners for real-world complexity.

“The impact has been transformative,” she says. “We’re no longer limited to classrooms or clinical placements. AI allows us to simulate authentic healthcare situations and connect disciplines in new ways.” For Professor Mikkonen, AI is a collaborator, a tool that expands possibilities while leaving room for human judgment and creativity.

As AI is increasingly utilized, she sees the researcher’s role as a bridge-builder between technology and human needs. “We must ask: Does AI truly support learning? Does it foster inclusion and safety? Does it empower professionals rather than replace them?” She believes AI cannot replicate empathy, ethical judgment, or the relational aspects of care, which are central to healthcare and education.

Her vision for the future is one of hybrid intelligence, where humans and AI complement each other. She hopes AI will evolve toward more transparent, ethical, and human-centered applications, enabling global collaboration and inclusive education while keeping human values at the core.

Join the conversation at Kontinkangas Campus Science Day 2025

The themes of this article will also be explored at the upcoming Kontinkangas Campus Science Day 2025, scheduled for October 28th in Leena Palotie Hall under the title “Are you hAIped yet? – AI in Biomedical Discoveries.” The event is your opportunity to engage with cutting-edge research and explore how AI is transforming biomedicine. Whether you're a student, researcher, educator, or simply curious about the future of science, this event offers a chance to learn, connect, and get inspired.

So – are you hAIped yet?

Text: Mikko Karpale and Pirjo Åström with the help of Microsoft Copilot