Research themes
- Topological analysis
- Streaming graphs
- Biomedical image analysis
- Long tail/imbalanced statistical learning
- Belief theory
- Digital twins
- Simulation
Research activity
Dr Cardoen focuses on designing scalable, interpretable algorithms to enable novel scientific discovery from biomedical imaging data at multiple scales ranging from diffuse optical tomography (DOT) to confocal and superresolution microscopy (SRM, STED, SMLM, dSTORM). He leverages and extends concepts from several fields: fuzzy computing (belief theory, information fusion) to simulation (discrete event simulators, agent based simulation), graph algorithms, vector fields, recurrent neural networks, and structural causal discovery from complex data.
Application domains range from degenerative (Alzheimer, ALS, ageing) to metabolic (diabetes) and infectious disease, with a specific focus on elucidating subcellular function from complex data. Whenever possible he prefers to design algorithms that push beyond the empirical resolution limits of modalities and that are adaptive or robust to complex noise models.