New knowledge discovery regarding causes and consequences of paediatric critical illnesses via complex analyses of DNA methylation

Intensive care is a data rich discipline with very large databases of raw data routinely registered for individual patients from a myriad of information sources. These comprise not only one-time demographic and admission information but also time series of diverse physiological signals, medications and treatments, laboratory analysis and with the advent of "omics", also genetic and epigenetic information. Patients are also routinely followed-up after their discharge from the intensive care unit, and consequently additional data on their long-term evolution is collected. Analysis of these vast databases is a challenging feat, which nowadays requires the use of advanced data-driven techniques, such as artificial intelligence and machine learning. These techniques allow exploiting the hidden patterns in raw data in order to unravel actionable knowledge. This results in either “knowledge discovery” when the hitherto hidden patterns are revealed in a clinically understandable way or in “predictions” when the patterns remain hidden but result in prognostic performance that equals or surpasses expert clinicians. With the ultimate goal of gaining a better understanding of the long-term consequences of intensive care and therefore of improving patient care, in this project we apply artificial intelligence for prediction and knowledge discovery with emphasis on the one hand on epigenetic data and on the other hand on tasks relevant to intensivists in their daily practice.