Building TrustworthyDigital Twins for Patient-Centred Medicine
3R Swiss Imaging Network SA
3R Swiss Imaging Network is a leading European private radiology group with 22 centers across Switzerland in both French and German-speaking regions, staffed by 60 radiologists and serving 300,000 patients annually. We operate a unified PACS, RIS, and datalake infrastructure, ensuring seamless data integration and interoperability across all sites. Since 2021, we have implemented 10 AI solutions in clinical practice, processing approximately 100,000 imaging studies per year. Our R&D department pioneers AI-driven clinical research, focusing on multimodal solutions for personalized medicine and workflow optimization. We maintain electronic patient consent for research and data sharing to the cloud, supporting ethical and secure innovation. 3R fosters industry partnerships and international collaborations, advancing impactful healthcare innovation.
Role within TWIN-X
3R Swiss Imaging Network brings TWIN-X into the place where most imaging really happens: community radiology. We run diagnostic centers across six Swiss cantons, Valais, Vaud, Geneva, Fribourg, Neuchâtel and Zürich, which gives the project something university hospitals alone cannot offer, namely imaging data from everyday practice, recorded on many different scanners and reflecting a broad mix of patients. Our main job is to put the project's AI to an honest test. We check whether the models stay accurate and fair once they leave the lab and meet new sites, new equipment and new patient groups. We also lead the work on showing how well these tools carry over to ordinary clinical networks, and we contribute our radiology expertise to the reporting templates, the data preparation and the secure Swiss cloud setup behind the research. What we care about is simple: AI that earns its place in the clinics where patients are actually seen.
Key tasks:
- Supply real-world imaging data from community radiology sites, spanning many scanner makes and a wide patient mix
- Lead the independent, multi-site testing of the project's AI, checking accuracy, calibration and fairness from one site to the next
- Lead validation of the cardiovascular and oncology prototypes in real clinical conditions
- Lead the project's flagship paper on how well these tools hold up across community radiology networks
- Cover the supporting work: reporting templates, data anonymisation, the secure Swiss cloud, and the ethics and data-sharing approvals for the Swiss sites
