Scientific Computing for the Improved Detection and Therapy of Sepsis

Sepsis is the number 3 cause of mortality in the general population and ranks number 1 cause of death in the surgical intensive care unit (ICU). Sepsis is defined as systemic inflammation in the presence of an infection. Typically half of the patients in a surgical ICU suffer from systemic inflammation. Distinction of non-infectious from infectious inflammation (sepsis) is very challenging for clinicians. The association with antibacterial resistance development precludes uncritical prophylactic antibacterial treatment.

Our objective is to identify rules for early and reliable diagnosis of sepsis in critically ill patients and to develop tools for clinical diagnosis. For this, we apply epidemiological and biostatistical methods as well as computer linguistics and text mining to clinical documentation in a data-oriented approach. We collect complementary data from targeted cellular and molecular analyses. Particular emphasis is on modeling and simulation of the pro- and anti-inflammatory process. This enables testing of hypotheses in silico and the discovery of new strategies for sepsis therapy, will enhance our understanding of inflammation, and has the potential to significantly advance critical care.

Discover more about our project at the project web site.