This three-day training course is designed to provide a basic introduction to the field of biomedical ontology and to enhance awareness of current developments and best practices in ontology in the life sciences.
Attendees who might find this seminar worthwhile include:
developers and users of biomedical ontologies, terminologies and coding systems,
developers and users of electronic patient record systems,
biologists and physicians interested in the possibilities of modern ontologies.
We are targeting advanced doctoral students, but welcome interested post-doc and industrial participants as well. The number of participants is restricted to about 30 to maximize possibilities for intense discussion. All participants will receive from their attendance in this tutorial hands-on training in ontology design and use.
Barry Smith: Introduction to Biomedical Ontologies
Werner Ceusters: Biomedical Ontologies and the Electronic Health Record
Olivier Bodenreider: On Mapping, Aligning and Integrating Biomedical Ontologies
Mark Musen: Case Studies in Ontology Development
I. Introduction to Biomedical Ontologies
Biomedical Ontologies have developed in an uncoordinated way, often reflecting mere relations of 'association' between what are called 'concepts', and serving primarily the purposes of information extraction from on-line biomedical literature and databases.
In recent years, we have learned a great deal about the criteria which must be satisfied if an ontology is to allow true information integration and automatic reasoning across data and information derived from different sources. Such criteria would ensure that ontologies which satisfy them would be automatically interoperable.
We will survey existing biomedical ontologies and show how they fall short of meeting these criteria and how they can be reformed in such a way as to allow true information integration. We also survey the current reform efforts under the auspices of the Open Biological Ontologies (OBO) consortium.
II. Biomedical Ontologies and the Electronic Health Record
The future of biomedical informatics, in an era of personalized, evidence-based medicine, will increasingly involve reasoning with the sorts of temporally indexed instance data we find in the Electronic Health Record (EHR). We will survey the problems which confront us when we try to use current EHR systems as a basis for such reasoning, and show how an adequate ontology of the biomedical domain can lead to a new and better management of EHR data.
Current Electronic Health Records (EHRs) are organized around two kinds of statements: those reporting observations made, and those reporting acts performed. In neither case does the record involve any direct reference to what such statements are actually about. They record not: what is happening on the side of the patient, but rather: what is said about what is happening.
In addition, existing EHRs embody information primarily in the form of general concept codes, which are tied to instances in reality only indirectly, via designations of persons and times. In contrast to this, we shall describe a regime in which the EHR incorporates direct representations of all clinically salient individuals as they are related together in reality. We show how this will allow us to achieve interoperability among different EHR systems at the level where it really matters: in regard to what is happening in the real world. It will allow us to keep track of particular disorders and of the effects of particular treatments in a precise and unambiguous way, and to engage in new types of reasoning and error checking in relation to the data encoded. We will also show a prototype implementation of an EHR/terminology system conforming to our methodology for ontology design, focusing on how such an implementation can be used to verify data entry in the EHR, to reason with the data, and to use the resultant EHR / terminology system for statistical and other purposes.
III. On Mapping, Aligning and Integrating Biomedical Ontologies
Unlike other domains, biomedicine has a long tradition of developing terminological resources and ontologies for organizing both documents and thoughts. With the increasing need to integrate vast, disparate clinical and biological information resources, these efforts are becoming increasingly important. One result is the development of repositories of cross-referenced ontologies such as the Open Biomedical Ontologies (OBO) library.
We will survey these developments in a critical manner, focusing on mappings, alignments and integrations in various domains of ontology, and paying attention also to the way biomedicine is currently being used as a testbed for Semantic Web technologies and on the associated formalisms for representing ontologies (RDF/S, OWL). Thus for example we will examine how relations among types in single ontologies can be detected from patterns of associations in annotations, and draw parallels between terminology integration in the UMLS and data integration through RDF in the Semantic Web.
IV. Case Studies in Ontology Development
We will examine a series of existing biomedical ontologies, indicating how they were initially developed and subsequently maintained, and outlining common problems and solutions. The evolution of ontologies to represent clinical practice guidelines and protocols will provide a foundation for much of the discussion. We will survey early efforts to define entities of clinical care, and trace the expansion of the corresponding models as investigators incorporated increasing numbers of distinctions concerning clinical protocols and guidelines. We will discuss trade-offs in modeling as well as work on ontology evaluation.
We will discuss how the need to create biomedical ontologies has stimulated the creation of new technology for ontology editing and management. Special attention will be paid to the Protege ontology development environment, and to its OWL plug-in, and also to the new technologies developed by the US National Center for Biomedical Ontology.
Participants should plan to arrive
in Dagstuhl between 3 and on
The registration fee is € 600 and includes lodging and full board at Schloss Dagstuhl. A non-refundable registration fee of € 20 is included in this amount. Doctoral students and other academics may apply for a grant of up to €400 to assist their participation.
Further details on the amenities at Dagstuhl and getting there may be found here.
Inquiries about the course and registration should be addressed to firstname.lastname@example.org.
Barry Smith is Julian
Park Distinguished Professor of Philosophy in the University at
Werner Ceusters is Professor of Psychiatry in the University at Buffalo and Director of the Ontology Research Group in the New York State Center of Excellence in Bioinformatics and Life Sciences. He has degrees in medicine, neuropsychiatry, informatics and knowledge engineering and served as coordinator of a series of international research projects in medical natural language processing under the Third, Fourth and Fifth Research Frameworks of the European Commission. Since then, he has also been active in standardisation bodies related to medical terminology such as CEN/TC251/WG2 and ISO/TC215/WG3. In April 1998, he started a new company - Language & Computing nv (L&C) - to exploit the results of his research. He left L&C in 2004, his main interest being now applying and testing a new theoretically-grounded approach to ontological engineering.
Olivier Bodenreider, MD, PhD, leads Medical Ontology Research in the National Library
Mark Musen is head of Stanford
Medical Informatics, an interdisciplinary research group at