On-site campus AI, virtual patients and augmented reality: the future of learning in the healthcare professions

“Around 6.3 million people work in healthcare professions across the German-speaking world. The healthcare professions cover a very broad spectrum, ranging from nursing through midwifery and physiotherapy to human and veterinary medicine. Whilst the professions that work with and on people have different areas of specialization and approaches to patient care, training in German-speaking countries is generally becoming increasingly academic in nature. This also opens up new opportunities for collaborative learning across healthcare professions, for example through shared teaching modules. The fundamental mechanisms of training are also very similar. The focus is on strengthening competencies, the entrustment of professional activities, developing manual skills, and increasingly enhancing the ability to act practically and make safe decisions. “To support this in training, AI and technology-supported learning are increasingly being utilized,” emphasizes Dr Julius Josef Kaminski, MD, medical education researcher, and Chair of the Conference Committee on AI and Technology-Supported Learning

Campus AI language models for reliable knowledge and the first virtual patients

To help students develop better problem-based thinking, some universities are already developing campus-based language models that are populated with specialist knowledge from the healthcare professions and can be used by students as a reliable source of knowledge. “The idea behind these visionary projects is that information from electronic health records is fed into LLM-supported systems optimized for learning. Through this mirroring and the virtual depiction of a course of treatment, learners can engage with realistic scenarios as they occur in clinical practice or almost as if in a virtual hospital,” explains Kaminski. “Major challenges in such projects include, for example, the complexity of electronic health records and hospital infrastructure, which must be taken into account,” says Kaminski. Additionally, such LLM-supported systems could also assist nurses and doctors in clinical practice.  “In a chat, one could, for example, enquire when and why a particular medication was administered to Patient X, how this affected certain laboratory results, and, through dialogue, cross-check specific further recommendations for action based on data, without having to work through various paper documents,” says Kaminski.

Multimodal models: Tracking down the disease like a detective

Multimodal large language models (MLLMs) represent a further development of LLMs. These MLLMs can be optimized for educational purposes and thus process specific anonymized diagnostic images – such as diagnostic imaging or ECG tracings. Here, students can simulate highly practical scenarios using text chats and visual cues, and learn from them. “This specifically trains differential diagnostic thinking, which traditionally requires a great deal of time and patient contact, and which often only takes place late in training or during the practical phase of the profession. This contact and the gaining of diagnostic experience can thus be brought forward into the training programme virtually, within a safe environment,” emphasizes Kaminski. It is important that these applications do not immediately reveal the correct solution to students, but rather that students develop, through chat dialogue, their initial ideas regarding the condition, consider which findings they need to rule out certain conditions, and identify which findings support specific diagnoses. “A chat system of this kind requires a software architecture that encompasses an LLM, as well as other functions such as a memory, a knowledge repository and the ability to access specific support tools. The challenge lies not only in optimizing LLMs themselves as effectively as possible from a pedagogical perspective, but also in optimizing the AI harness that surrounds the AI system,” said Kaminski. Sometimes, the models perform better the less they know the solution themselves from the outset and the more they receive the correct information only bit by bit from another – supervising – system. This reduces the chances that the system provides hidden clues or explicitly reveals the correct result.

VR simulations: Training for emergencies and interdisciplinary collaboration

Digital training scenarios can vary in complexity, ranging from simple PC-based applications to complex virtual reality. VR simulations also allow specific situations in hospitals – including those involving teamwork – to be practiced, thereby imparting procedural knowledge, an understanding of workflows and practical know-how. This allows procedures – for example, for emergencies or the chain of survival – to be practiced without the pressure of real-world time constraints. “VR simulations also offer significant opportunities for interprofessional collaboration between different healthcare professions, such as nursing staff and midwives. In multiplayer mode, learners can then take on various roles within the VR simulation that they would also perform in their everyday professional lives. Instructors can provide feedback on collaboration from both an external and internal perspective. This allows for the training of critical thinking, teamwork, communication and decision-making under pressure, ultimately improving patient care,” emphasizes Kaminski. An innovative technology in VR is eye-tracking, i.e. the monitoring of eye movements. This makes it possible to gauge how challenged a learner is. It shows whether the person is still able to learn or is already feeling overwhelmed by the situation, which the teacher can then address.

AI learning tools that make learning more fun

VR simulations, augmented reality, and multimodal models also appeal to learners’ sense of enjoyment and fun. AI can make learning about health-related topics easier in a playful way. This is also possible in less complex situations – for example, content relating to healthcare professions can be used as a source for AI podcast programmes, generating entertaining podcasts to listen to. Perhaps less fun, but certainly efficient, is the open-source tool Anki. It supports learning with virtual flashcards. There are AI algorithms designed to predict the best time for each individual to revise a piece of learning material so that it is stored in long-term memory. This enables certain situations, terminology or test questions to be learnt effectively over the long term. Many learners also independently link Anki with AI solutions, using AI to optimize their own learning.

AI Fundamentals in Healthcare Professions: From Prompting to Calendar Agents

Technology-enhanced learning and the ability to apply AI are key skills of our time. These skills must be taught to teachers and learners in healthcare professions. It is therefore not a question of banning AI, but rather of adapting training programmes so that students are equipped to use AI and technology-supported learning responsibly. “The basics revolve around working with generative AI, meaning that an AI produces texts, images, podcasts and other content. Correct prompting, for example, is essential to obtain good results from the AI. Prompting can thus be integrated into and taken into account in learning tasks,” emphasizes Kaminski. Furthermore, a conceptual understanding of AI is needed so that the results generated by AI can then be assessed using human verification techniques to check for plausibility and automation biases. “In addition to knowing what I can do with AI, learners also benefit from practical examples that offer them immediate added value.  For example, AI agents can be used to create individual learning plans for students, enabling them to learn as sustainably as possible,” says Kaminski. – It is and remains important when using AI and technology-driven tools that they are not an end in themselves, but should always serve to improve healthcare provision.

AI as a key factor in the healthcare sector

To sum up, Dr Anne Lloyd, AMEE Chief Executive says:  “The evolution and use of AI is a critical factor impacting all areas of healthcare globally. It is essential that all professionals working in healthcare education are aware of the opportunities and risks associated with AI. Our Conference provides a platform for people to come together to share, knowledge and best practice, and to learn from and inspire each other.”

About AMEE
AMEE was born in Europe over 50 years ago, but has evolved into a truly global organisation, with a community that stretches across continents, spanning all areas of healthcare, to become the only organisation where healthcare education professionals can come together as a community regardless of their specialty, role or career stage. AMEE exists to promote and inspire excellence, collaboration, and scholarship across the continuum of Health Professions Education and to transform healthcare for all through excellence in education and scholarship. Ultimately, improving patient outcomes and experience.

The AMEE Conference is the organisation’s flagship sector-leading annual conference, that brings together over 4000 healthcare professionals from over 100 countries across the globe. https://amee.org/events/amee-2026/

About IAKW-AG
IAKW-AG (Internationales Amtssitz- und Konferenzzentrum Wien, Aktiengesellschaft) is responsible for the maintenance of the Vienna International Centre (VIC) and the operation of the Austria Center Vienna. With 21 halls, 134 meeting rooms and around 26,000 m² of exhibition space, the Austria Center Vienna is Austria’s largest conference centre and ranks among the leading players in the international conference industry.