Estimating similarity of surgical situations with case-retrieval-nets

Representation and recognition of surgical situations is a prerequisite for the development of context-aware surgical assistance systems. In this publication a method for recognition of surgical situations with Case-Retrieval-Nets is presented. It enables the estimation of similarity between models of surgical situations. The main advantage of this approach is the combined use of domain knowledge and reasoning algorithms to estimate similarity. Domain knowledge about human anatomy is based on a reference ontology. Evaluation is performed on situations of two cholecystectomies.