ASIC: Algorithmic Surveillance of ICU patients

The demand for intensive care medicine will strongly increase over the next 10—15 years facing unmet medical needs, such as early diagnosis of acute respiratory distress syndrome (ARDS). Within the SMITH consortium, the ASIC use case (algorithmic surveillance in intensive care) will develop a system for continuous analyses of data obtained from the hospital patient data management system (PDMS) in order to enable model-based ‘algorithmic surveillance’ of the state of critically ill patients. The ASIC use case, therefore, develops the integration of two key elements for data-driven clinical support: a data – driven Diagnostic Expert Advisor and a Virtual Patient model.

The project comprises Big Medical Data analysis and a hybrid deep learning system which make use of the infrastructure and the data sharing among the participating hospitals and companies.  Moreover, we are working on the development and improvement of mechanistic systems medicine models and physiological models complemented by machine learning in a hybrid virtual patient model. Ultimately, the virtual patient model will enable individual prognoses to support therapy decisions, clinical trials and training of future clinicians.