Medical informatics

Medical experts (clinicians and supporting staff) have an increasing amount of data and computational power at their disposal.

Medical experts (clinicians and supporting staff) have an increasing amount of data and computational power at their disposal. A key challenge is to make the most of these resources to support medical practice. In the Life Sciences and Health group, our projects typically target demand-driven research questions that, when solved, enable improved (decision) support for medical practitioners. As such, we focus on fundamental research questions that can be answered with methods, techniques, and ultimately, tools from mathematics and computer science that underlie translational medical research, connecting life-science models and algorithms to real-world medical practice.

To ensure broad applicability, the research is focussed on the design and application of algorithms in the computational intelligence domain, both for optimization and machine learning. A particular focus is on the design of multi-objective optimization algorithms since real-world problems are often multi objective, meaning that there are several conflicting goals with an inherent trade-off between them (for example, maximizing radiation delivered to tumour cells versus minimizing radiation delivered to healthy cells).

Our multi-objective optimization algorithms can provide unique insights into the space of interesting (that is, near-optimal) trade-offs between objectives of interest, for instance when generating treatment plans that can have both positive and negative effects. Rather than iterating the generation of single plans until an expert is satisfied, the space of all interesting plans can be computed and studied at once. Moreover, by using machine-learning algorithms our research contributes to moving toward patient-specific medicine (also known as precision medicine), by learning from the data of previously treated patients the likely good solutions for new patients.

The main enabling technology underlying our multi-objective and machine learning approaches stems from a dedicated fundamental research line on evolutionary algorithms.

Contact person: Peter A.N. Bosman
Research group: Life Sciences and Health (LSH)
Research subgroup dedicated page: Medical Informatics
Research partners: AMC Department of Radiation Oncology, AMC Department of Pediatric Oncology, Princess Máxima Center for Pediatric Oncology, NKI Department of Radiation Oncology, TU Delft Department of Electrical Engineering, Mathematics and Computer Science, Utrecht University Department of Information and Computing Sciences, Elekta, Xomnia, ASolutions B.V., Cancer Health Coach, Focal Meditech B.V., Brocacef Group N.V.