Summary
Objectives: Biomedical ontologies have been very successful in structuring knowl edge for many
different applications, receiving widespread praise for their utility and potential.
Yet, the role of computational ontologies in scientific research, as opposed to knowledge
management applications, has not been extensively discussed. We aim to stimulate further
discussion on the advantages and challenges presented by biomedical ontologies from
a scientific perspective.
Methods: We review various aspects of biomedical ontologies going beyond their practical successes,
and focus on some key scientific questions in two ways. First, we analyze and discuss
current approaches to improve biomedical ontologies that are based largely on classical,
Aristotelian ontological models of reality. Second, we raise various open questions
about biomedical ontologies that require further research, analyzing in more detail
those related to visual reasoning and spatial ontologies.
Results: We outline significant scientific issues that biomedical ontologies should consider,
beyond current efforts of building practical consensus between them. For spatial ontologies,
we suggest an approach for building “morphospatial” taxonomies, as an example that
could stimulate research on fundamental open issues for biomedical ontologies.
Conclusions: Analysis of a large number of problems with biomedical ontologies suggests that the
field is very much open to alternative interpretations of current work, and in need
of scientific debate and discussion that can lead to new ideas and research directions.
Keywords
Biomedical ontologies - biomedical informatics - spatial ontologies - artificial intelligence
- mathematical morphology
