Transformation operators for easier engineering of medical process models

The need for high-quality models is increasingly recognized for driving and documenting complex medical processes such as cancer therapies. A medical environment for such processes has to deal with a great multiplicity of dimensions such as different pathologies, different hospital departments, different agents with different concerns and expertise, different resources with a wide spectrum of capabilities, and so forth. The variety of needs along those multiple dimensions calls for multiple, complementary and consistent facets of the composite process model, each addressing a specific dimension. Building multi-dimensional process models is in our experience hard and error-prone. The paper describes various operators for composing process model facets in a coherent way or, conversely, for decomposing process models into specific facets that abstract from details irrelevant to a specific dimension. These operators are grounded on the formal trace semantics provided by our process language and its supporting analysis toolset. The paper shows how these operators may help modeling, analyzing, documenting and enacting complex processes. Their use is illustrated on simplified examples taken from real cancer therapies.