Data in Use: How POLAR and BMI2 Expand the Benefits of the LOOP BMIP

Development of the LOOP BMIP was largely completed in 2025. The platform is close to going live. It’s technical infrastructure and central user portal are now in place, standardized data exchange procedures have been established across participating institutions, and a comprehensive framework agreement for joint data projects has been developed. Key data protection questions were resolved in close coordination with the relevant authorities. Since early 2026, the platform has been undergoing an extended testing phase with selected users and will be made available to researchers at all participating institutions from third quarter of 2026 onward.

Within this infrastructure, the platform projects POLAR and BMI2, launched in early 2024, demonstrate how the value of the LOOP BMIP can be further expanded in practice: moving beyond the mere availability of clinical data toward specialized data and analysis environments for proteomic lymphoma research and AI-supported imaging. Insights from both projects will feed into the further development of the LOOP BMIP, for example in areas such as data formats, analysis pipelines, FAIR-compliant databases, and governance processes for new data types.

POLAR: Expanding value through proteomics

POLAR (ProteOmics for Lymphoma and Prognostication) exemplifies how the LOOP BMIP can generate additional impact through specialized data environments and analyses. The project is building one of the world’s largest systematic proteomic databases for lymphoma and linking it with clinical and other molecular information to better characterize lymphoma subtypes, more accurately predict disease progression, and support more targeted therapy selection.

A particular advancement lies in the reliable analysis of formalin-fixed tissue samples, which are very common in routine clinical practice. This will enable the use of modern proteomics in everyday clinical settings. Highly standardized, partially automated laboratory and analysis processes, along with a database designed according to FAIR principles, enhance the usability of the LOOP BMIP by providing quality-assured access to complex proteomic data—for both local and internationally networked research projects.

BMI2: Expanding value through imaging and AI

BMI2 (Biomedical Informatics Imaging Platform) extends the value of the LOOP BMIP by systematically integrating medical imaging into the platform and preparing it for AI-supported analysis. The project is developing a unified, SPHN-compatible infrastructure for imaging data from radiology, pathology, radiation oncology, and other disciplines, linking these with clinical and molecular information.

In 2025, the focus was on concrete use cases: large, curated datasets for chest imaging, PET/CT in oligometastatic cancers, and pediatric brain tumors were established and made usable for automated analysis methods and AI models. At the same time, metadata standards and data pipelines were developed that can be seamlessly integrated with the LOOP BMIP. In this way, BMI2 expands the platform’s usability by providing structured access to imaging data and lays the groundwork for gradually integrating AI-based image analysis into research and clinical decision-making processes.