Delete search term


Quick navigation

Main navigation

Data-driven decision support for intracranial aneurysms and hospital catering using Bayesian networks

At a glance


Clinical decisions in medicine and management decisions in facility management are regularly made on the basis of little evidence or extrapolations and are also influenced by subjective and economic aspects.

While data is generated exponentially in medicine due to increasing digitization, there is no framework for integrating this data into clinical routine. The number of patients diagnosed with intracranial aneurysms has increased significantly in recent decades due to better diagnostics. Enormous efforts are being made to identify predictive factors of disease progression and treatment outcomes. These factors are in a complex relationship with each other, which must be adequately taken into account in the modelling.

Ever-increasing costs in the healthcare system require economic action without impairing the quality of care. Hospital catering is expensive, but also a major factor for patient satisfaction. In addition to the purely economic optimization, numerous qualitative factors such as sustainability and employee satisfaction are also central. Their mutual dependencies, which are often not directly apparent, are of particular interest here.

The increasing collection of data in the health sector allows a systematic analysis of such questions. Bayesian networks (BNs) are a specific type of graphs that allow the variables and their dependencies to be represented in an easily accessible way - both for calculations and for interpretation. In particular, the application of BN enables the causal and probabilistic modeling of numerous factors that are in a complex, hierarchical context. The management of intracranial aneurysms is therefore an ideal framework to test BNs. In the present project, extensive data collection on the gastronomy of various health institutions is also examined. In addition, simulated data is analyzed to improve model quality.

The resulting models in the form of BNs will allow conclusions to be drawn about how different factors influence each other directly or indirectly. In this way, systematic foundations for clinical decisions and management decisions can be developed and a contribution can be made towards patient-centered and resource-optimized services in healthcare institutions.

Further information