Semantic technology for management of data recorded during therapy of cardiac arrest patients

Abstract

The topic of this thesis is the integration of data recorded during cardiopulmonary resuscitation (CPR). In the process of CPR many important data should be recorded, such as: time of the cardiac arrest witnessed, number of shocks given to the patient and parameters for each shock. These data will be used for the coming analysis and producing the report of the therapy. The analysis and comparison of the CPR data and reports may improve the survival rate. However the formats and terminologies of CPR data vary greatly between EMS (Emergency Medical Services) systems in different places. So it is inconvenient to access, compare and transfer data between different EMS systems. In this thesis we use semantic way to solve the heterogeneity problem. An ontology is defined to provide a standardized representation of the heterogeneous data from different EMSs. In this ontology we defined standardized CPR terminologies and report style, and the “dialect” from different EMS systems can be mapped onto this ontology. An interface is defined to map the different databases from different EMS systems onto this ontology. And we provide a uniform way based on the ontology to query all of the databases as if they are one huge database.