Clinical therapy guidelines and protocols (CGPs) are a promising tool for quality assurance in medicine. Supporting their use through intelligent systems increases acceptance and reduces the workload. A prerequisite for this translating CGPs into a form that can be processed automatically. Using methods of information extraction can significantly contribute to this step. Nonetheless, there are still many challenges left. At the same time, there is already extensive experience from previous projects that is not yet being fully utilised for creating machine readable CGPs. 

Experiences from previous projects

Previous research provides experience in the field of therapeutic guideline modeling, with and without the use of information extraction, as well as guideline validation, verification, and execution. Combining these available open source modules, will allow developing a comprehensive solution for creating CGPs in formal notation and natural language.

Further development, exemplary use cases and evaluation

To achieve this goal, the focus is on further developing information extraction and transformation methods to support the transformation of CGPs in natural language into formal models, as well as creating exemplary use cases and evaluating a prototype together with experts from the application area.

Simplified creation of new CGPs

The outcome of the project are sample models and newly developed methods that take the creation of natural language CGPs and their formal models of practical application a step further. This not only simplifies the creating formal models, but also creating new CGPs in natural language.