Frédérique Meeuwsen

Radboud UMC

In short, Frédérique Meeuwsen's fellowship entails the following:

Towards becoming the pathologist of the future

Pathology diagnostics are under increasing pressure due to rising patient numbers, the growing demand for comprehensive disease characterization, and an expanding range of treatment options, all while the number of pathologists continues to decline. The rise of machine learning-driven image analysis tools offers a promising solution, with AI applications transforming pathology workflows. These tools can assist in tasks such as assessing prognostic biomarkers, tumor grading, diagnosis, and quality control, thereby improving efficiency and accuracy.

However, despite their potential, few AI algorithms are in routine clinical use, and there is a notable lack of studies evaluating their impact in real-world clinical settings. Regulatory challenges and reimbursement issues contribute to a significant "translation gap" in pathology—referring to the difficulty in prospectively validating and integrating AI models into clinical workflows. Additionally, there is a shortage of experts who can navigate the complexities of AI implementation and effectively engage with diverse stakeholders. This gap can only be bridged by involving pathologists who are well-trained in both the clinical and technological aspects of computational pathology.

Frédérique Meeuwsen, who is a registered pathologist by the end of 2024 and holds a PhD in Biomechanical Engineering (TU Delft), aims to explore every aspect of the AI development cycle—spanning from problem identification to model development and successful implementation. Through collaboration with leading researchers at Radboudumc and internships at internationally recognized labs, she plans to acquire the knowledge and
practical tools needed to position herself as a pathologist of the future and play a significant role in the transformation of pathology.

Supervision: Prof. dr. Jeroen van der Laak, Radboudumc, Nijmegen.
Planned Internships:
- Harvard Medical School, Boston, USA (Prof. dr. Faisal Mahmood)
- University of Leeds, Leeds, UK (Prof. dr. Darren Treanor)
- Institute of Tissue Medicine and Pathology, Bern, Switzerland (Prof. dr. Inti Zlobec)

Publications

• MSAIPath: Predicting Microsatellite Instability From Routine Histology Slides Without Reinventing the Wheel 
  • Modern Pathology, Volume 39, Issue 1 
  • Click here to view the full publication
• Computer-aided tumor cell fraction (TCF) estimation by medical students, residents, and pathologists improves inter-observer agreement while highlighting the risk of automation bias
  • Virchows Archiv, 2025
  • Click here to view the full publication

Short summary of progress / results December 2025

My fellowship focuses on exploring the full AI development cycle within pathology and on bridging the gap between AI research and clinical implementation. During the first half of the fellowship, substantial progress has already been made. 

I am currently leading the implementation team responsible for the safe integration of an AI application into the routine pathology workflow. This tool is expected to go live in early 2026, with several additional AI tools to follow. My contributions include coordination and alignment with clinical, technical, and organizational stakeholders, ensuring compliance with (IVDR) regulatory requirements, and shaping the overall implementation strategy. A scientific manuscript describing this implementation trajectory has been submitted for publication. In addition, I have presented this work at several scientific meetings, including the European Congress of Pathology 2025 in Vienna. 

Furthermore, I have completed one of the planned international internships at the Institute of Tissue Medicine and Pathology in Bern (under the supervision of Prof. Dr. Inti Zlobec), where I gained valuable experience and established strong collegial collaborations. I have also attended multiple conferences in Europe and the United States, further deepening my expertise and expanding my international network.

Within the Computational Pathology Group at Radboudumc, I am involved in research projects across several domains, including multimodal AI, generative AI, and large language models (LLMs), resulting in multiple co-authored publications. These activities further strengthen my role as a pathologist of the future. The knowledge gained will also be shared during the first AI workshop for pathologists, scheduled for February 2026.