Combining YAWL and DBNs for Surgical Phase Detection

To provide assistance functions in context of surgical interventions, the use of a surgical phase detection plays an important role. By assessing the progress of an on-going surgery, a tailored (i.e., context sensitive) decision support for medical practitioners can be enabled. Subsequently, this provides opportunities to prevent errors, injuries, negligence or malpractices. In this work, a surgical phase detection, combining Yet Another Workflow Language (YAWL) with Dynamic Bayesian Networks (DBNs) is proposed. Thereby, YAWL is used to model the relationship of surgical phases; DBNs are used to allow for the detection of surgical phases of interest. The approach is presented for the application example of a cholecystectomy (removal of the gallbladder).