Stanford
Clinical Mind AI

Clinical Mind AI at AMEE 2026: Learning Together Across Countries and Contexts

Clinical Mind AI was presented at AMEE 2026 in Vienna through the workshop “Instructor-authored AI-simulated patients: hands-on design and assessment with Clinical Mind AI.” The session brought together Anna Svenningsson, Karolinska Institutet (Sweden); Mauricio Saavedra, Pablo Gutiérrez, and Vicente Bustos, Universidad Autónoma de Chile (Chile); Kuan-Hsun Lin, National Defense Medical University (Taiwan); Marcelo García-Dieguez, Universidad Nacional del Sur (Argentina); and Marcos Rojas, Project Director of Clinical Mind AI, Stanford University (United States).

The workshop introduced participants to the Clinical Mind AI project and to the platform itself: a web-based authoring environment in which instructors can create and deliver customizable virtual-patient activities, define feedback and assessment approaches, and collect data for teaching and research. Participants also explored how instructors can use the platform in different ways and at different points in a curriculum, from preparation and classroom learning to simulation, assessment, and independent clinical reasoning practice.

But the workshop was also designed as an opportunity to learn from what educators are already doing across very different settings. From Karolinska Institutet, Anna Svenningsson shared work spanning medicine, midwifery, and physiotherapy, including the exploration of virtual patients for practicing difficult conversations in pediatrics and providing feedback around communication frameworks. Her presentation also highlighted an important feature of this international work: adapting activities to different languages and embedding implementation within a structured evaluation framework.

At the Universidad Autónoma de Chile, three faculty members showed how Clinical Mind AI is being used across different programs and stages of implementation. Mauricio Saavedra opened with the university’s experience in kinesiology, describing how the platform has been integrated into Practice II and Practice III as students move from foundational interviewing and virtual-patient practice toward clinical placements, more complex simulations, and feedback. The same cohort has used Clinical Mind AI in different ways over time, including classroom role-play, practice at home, voice-based virtual-patient interviews, and observed encounters followed by debriefing.

Pablo Gutiérrez then presented the work underway in dentistry, where faculty identified a need for more opportunities for individual diagnostic-reasoning practice with immediate feedback. The teaching team developed 35 cases, 55 virtual patients, and 29 assessment rubrics, alongside adaptations of the platform for dental education. An initial radiology pilot is now informing plans for a broader longitudinal program spanning multiple courses and years of the dentistry curriculum, with evaluation and research embedded in the implementation.

Vicente Bustos shared the experience in medicine and psychiatry, where limited opportunities for direct clinical interviews create a particular need for additional practice. Clinical Mind AI has been used to develop psychiatric clinical cases and vignettes in which students can interact with simulated patients through voice, while faculty remain responsible for implementation, modeling, supervision, and evaluation. The team is now considering further integration across the curriculum, interdisciplinary applications, and the inclusion of perspectives from people with lived experience.

From National Defense Medical University in Taiwan, Kuan-Hsun Lin demonstrated a very different implementation: simulated surgical ICU encounters conducted in Traditional Chinese, including deliberately challenging communication situations such as a drowsy patient or a patient unable to speak. The cases were designed not only around diagnosis, but also around how learners introduce themselves, explain information, communicate with patients and families, and document the encounter, illustrating how local educational priorities can shape both simulation and assessment.

Marcelo García-Dieguez, from Universidad Nacional del Sur in Argentina, presented how Clinical Mind AI is beginning to be incorporated into a medical program built around an integrated spiral curriculum, active and progressive learning, competency-based education, early clinical contact, and community orientation. He highlighted opportunities to use the platform longitudinally as students progress from developing the doctor-patient relationship and problem-focused interviewing to more complex clinical reasoning, differential diagnosis, management decisions, and integrated care planning in later stages of training.

What made the AMEE workshop especially meaningful was not simply showing one technology or one model for using AI in education. It was seeing how a shared academic platform can take on different forms when educators bring their own learners, languages, curricula, clinical realities, and research questions to it. An idea developed in Sweden can provoke a new question in Chile; an assessment approach from Taiwan can inspire an educator in another context; and experiences from Argentina can help the wider community think differently about longitudinal implementation.

That is an important part of what Clinical Mind AI is trying to build: not a single way of teaching with AI, but an international community in which educators can experiment, evaluate, share what they learn, and generate evidence together.

More News Topics

More News