!! Collaborate Guideline Development - Medical treatment at its best
Within the project CliWE (Clinical Wiki Environments), KnowWE is extended by plugins to allow for the collaborative development of Computer-Interpretable Guidelines (CIGs). Clinical guidelines are based on evidence-based medicine and improve patient outcome by providing standardized treatments. Their computerization allows for decision-support systems at the point of care, or even the automated application by closed-loop systems in the setting of Intensive Care Units. 

The goal of CliWE is to create a platform that supports the engineering of CIGs by spatially distributed domain specialists. Therefore, the graphical CIG language DiaFlux was created. Its focus lies on the direct applicability and understandability by domain specialists. By offering only a small set of intuitive language elements, the guidelines can in the best case be built and maintained by the domain specialists themselves. Currently, the extensions developed within CliWE are used in the project WiM-Vent. Its goal is to integrate medical expertise concerning mechanical ventilation and physiological models into an automated mechanical ventilator. 

In the course of this project, one knowledge engineer guides and supports one domain specialists (backed up by a committee of further experts) during the knowledge engineering process. The latest version of the guideline contains 17 DiaFlux modules, that in total contain 295 nodes and 345 edges. During its development, the testing capabilities of KnowWE are extensively used. So far, about 1.100 continuous integration builds were automatically executed. Especially the empirical testing feature is applied to define and process local test cases, as well as ones that are created using external tools, e.g., a Human Patient Simulator. Those simulated patient sessions can then be replayed in KnowWE for introspecting and debugging the guideline execution. A highlighting of the taken paths within the DiaFlux models serves as an accessible means of explanation for the domain specialists.