advancing academic writing for medical students
Nicolas S. Bodmer
Chief scientific officer
medignition AG
19.09.2025
Scientific writing is a major hurdle for many medical students. To address this challenge, medignition, in collaboration with the University of Zurich, has developed a digital tool that combines e-learning modules with an AI writing assistant. The aim is to help students gain autonomy in writing while easing the supervision burden on faculty.
Writing as a source of stress
For students at the University of Zurich, preparing a Master’s thesis is often the most stressful part of their studies. In a survey of 24 medical students, 75% reported elevated stress during thesis preparation, 58% cited scientific writing as the main stress factor, and 42% struggled with writing in English. Only 13% felt adequately supported. These figures underline the pressing need for structured, scalable support in academic writing.
E-learning meets AI
The tool combines structured e-learning content with an AI-driven writing assistant. The e-learning modules introduce the foundations of scientific writing, including manuscript structure and reporting standards. The AI assistant provides real-time, iterative feedback on structure, grammar, style, and flow. In the pilot evaluation, 71% of students rated the e-learning modules as helpful, and 75% found the AI assistant helpful or very helpful. The System Usability Scale (SUS) score was 76 out of 100, well above the benchmark of 68, confirming both practicality and acceptance among students.
Results and next steps
Since its launch, the tool has been used by 125 students, contributing to the completion of 71 Master’s theses. Building on this success, development continues in three directions: expanding AI support to additional sections such as results and discussion, deploying locally hosted large language models to ensure data security, and introducing systematic AI-based checks for formatting and structure to further reduce faculty workload.
A responsible path forward
The project demonstrates how AI can make a measurable difference in medical education by reducing student stress, increasing autonomy, and alleviating faculty pressure. At the same time, successful integration requires universities to define clear frameworks for data protection, ethics, and academic standards. With these guardrails in place, AI-supported writing has the potential to become a cornerstone of medical education in Switzerland and beyond.