Readers who have an active interest in the future of electronic health records (EHR) systems will find this paper essential. It presents the idea that electronic health records can be organized around anatomical concepts. The paper then discusses what concept-based anatomical knowledge representation involves and how concept-based EHR systems might materialize.

Typically, health record systems deal with patient data at the document level. Yet, to improve the medical decision making process, clinicians and other users of medical records need data organized at the information level. To do this, concept-based multimedia health records can be processed with analytical engines that are guided by relevant domain knowledge; this requires the concepts and the relationships among them to be defined, so they can be organized in a coherent fashion. The biomedical domain has ontologies available, such as the National Library of Medicine Unified Medical Language System, that capture and represent anatomical concepts and the relationships among those concepts. The benefits of concept-based electronic health records could potentially lead to care process improvements, a reduction in the number and impact of medical errors, and a decrease in costs.

A number of challenges exist in successfully designing and implementing concept-based electronic health records, such as the need to extend the existing ontology for concepts. This paper should serve as a springboard for debate and research on how to meet these challenges in order to be able to extract additional value from EHR systems when they finally become a common reality.