Technical Innovation in Health Data

Making Complex Health Data Usable, FAIR, and Predictive

Real-world health data is messy, fragmented, and often trapped in silos. Before it can be used to generate meaningful evidence, it must be transformed into something trustworthy and interoperable. In this theme, we develop the technical foundations needed to unlock the potential of high-dimensional health data.

We design and implement solutions for data wrangling, FAIR data pipelines, federated analytics, and machine-learning–based prediction modeling. Our work includes harmonizing routine clinical data, integrating MRI and imaging features, and creating secure workflows that protect patient privacy while enabling large-scale collaboration.

Much of our technical innovation originates from real-world MS research, but the tools we build are broadly applicable across healthcare domains. As use-case leaders in the Flanders AI Research Program and collaborators on projects such as BRANDO, we continuously push the boundaries of what responsible, high-impact health data science can achieve.

Key focus areas

• Data integration, cleaning, and standardization

• FAIR data (Findable, Accessible, Interoperable, Reusable)

• Federated analysis and privacy-preserving computation

• Machine learning and trustworthy prediction modeling

• Real-world MS data: clinical, imaging, and longitudinal records
  
  

Highlighted initiatives

• Flanders AI Research Program – Real-World Evidence Use Case

• BRANDO dataset collaboration (BCTRIMS)

• Predictive modeling projects in multiple sclerosis
  
  

Outputs

• Peer-reviewed publications (2024–2025 ML and radiomics studies)

  • De Brouwer et al. 2024. Machine-learning-based prediction of disability progression in multiple sclerosis: An observational, international, multi-center study (link)

  • Pirmani et al. 2025. Personalized federated learning for predicting disability progression in multiple sclerosis using real-world routine clinical data (link)

  • Khan et al. 2025. Leveraging Hand-Crafted Radiomics on Multicenter FLAIR MRI for Predicting Disability Progression in People with Multiple Sclerosis (link)

• Data pipelines and analysis workflows

• FAIRification tools and standard operating procedures