EdTech Fellow, Doctoral Candidate at TUM
December 17, 2025
Seventh Episode: Johannes Reifenrath
In this seventh episode, Johannes Reifenrath discusses his medical thesis and contributions to an Interdisciplinary, Longitudinal and Virtual Patient management app (ILVI). Listeners can now enjoy the podcast directly here below, and further explore his use of machine learning models in oncology as well as his insights on the limitations and opportunities of LLMs in medical education.
What we need is the skill to correlate symptoms with differential diagnoses, and the reasoning skill to choose a diagnostic pathway to confirm or rule out these differential diagnoses.
Johannes Reifenrath
Interview
1. Can you briefly introduce yourself and your current role at the TUM?
My name is Johannes, and I studied Medicine here at the TUM. I am currently wrapping
up my medical thesis, before starting my residency in the department of oncology next
Month.
2. What is your primary area of research, and what initially drew you to this field?
My thesis focuses on developing a method for label-free tumor diagnostics in brain cancer. I first combined Raman spectroscopy with fluorescence microscopy to obtain the specific spectra of a tumor marker protein, and then I trained two machine learning algorithms to distinguish tumor cells based on their spectral fingerprint of that protein. In vitro, we can thereby distinguish cells with high sensitivity and specificity. What I enjoy about my research is that it allows me to push the boundaries of my curiosity and creativity beyond the clinal context, where I usually I have to follow treatment guidelines.
3. I read that you also have a strong interest in Medical Education, and that you have been taking part in the development of a so-called ILVI app, which is to my understanding, a smartphone app that allows medical students to interact with virtual patients. What does ILVI stand for? And what are promising opportunities that arise from this one for students and teachers in medical education?
As young doctors, we are expected to hit the ground running. However, 71% of young doctors in Germany describe deficits in linking their knowledge to clinical reasoning. So what we need is the skill to correlate symptoms with differential diagnoses, and the reasoning skill to choose a diagnostic pathway to confirm or rule out these differential diagnoses. That is especially true now that patients and doctors have access to LLMs – but research shows that LLMs cannot yet faithfully recapitulate the iterative process of diagnostic reasoning; that still remains a core skill of physicians. And this is exactly what our Interdisciplinary and Longitudinal and Virtual Patient management app – short: ILVI – promises: a platform to train diagnostic reasoning skill. ILVI emulates the diagnostic spectrum of a university clinic, and is, to our knowledge, the first German app that uses a real-world financial budget and a time limit as gamification strategies. We are excited to leverage ILVI to upscale clinical case discussions in the core medical curriculum. With ILVI, students can virtually diagnose patients prior to the lecture; and during the lecture, educators can view the students’ responses in an evaluation tool that facilitates case discussions.
4. What future developments or outcomes do you hope to achieve in your research for educational technologies?
I believe, doctor-patient interactions are one of the most crucial yet often precarious situations in medical caregiving. Especially, history-taking is an essential skill for physicians, but there are few opportunities to train this in a secure environment where you get constructive feedback. We are prototyping a chatbot that can be integrated into ILVI to simulate a patient for history taking. We are also training a second LLM to rate the user’s interactions based on an established medical framework and give constructive feedback. I would like to see how this can improve history-taking skills. In addition, I would be excited to explore the use of multidimensional digital twins to enhance the realistic experience of doctor-patient interactions.
5. Could you also share with us something about your interests outside of your academic work?
Outside of my academic work, I really value spending time outdoors. I like to decelerate by going hiking in the Alps—it’s one of the few places where I can completely switch off. I’m especially fascinated by the alpine flora with its incredible diversity.
