Ontology-based Approaches to Medical Data Integration

Medical data come in a variety of forms and are stored in different types of databases. Some are structured relational database stores, while others are semi-structured and unstructured, such as the emerging NoSQL data stores. A lot of valuable data can be found in the form of unstructured text, such as clinical notes and discharge letters. To analyze and discover hidden patterns and extract knowledge from the data, they should be integrated. In the field of medicine, many ontologies have been created to provide a common basis for information exchange and to improve semantic interoperability. In this paper, we provide an overview of ontology-based integration approaches for various sources of medical data. We also identify current challenges and provide directions for future research.