Taxonomy A hierarchy of pre-defined categories which used to classify things.

MeSH (Medical Subject Headings) Created and updated by the United States National Library of Medicine (NLM), it is used by the MEDLINE/PubMed article database and used in their ontologies by clinicaltrials.gov, cancer.org and much of US healthcare in general). MeSH is the National Library of Medicine’s controlled vocabulary thesaurus (taxonomy). It consists of sets of terms naming descriptors in a hierarchical structure that permits searching at various levels of specificity.

Ontology The study and definition of the categories of things that exist or may exist in certain instances within a particular domain. An ontology is a “knowledge map” or a representation of the shared background knowledge of a community for use in software development. Biologists use medical ontologies in their work for the cataloguing, naming, describing; and in support of semantically rich querying. Web services utilize ontologies in the automated discovery of information resources. An ontology allows for the support of much greater granularity while still preserving the document structure by using a controlled vocabulary for the tagging data within a trial and across all trials that create a series of “maps” (ontologies) that can be interrogated by search. Documents that are poorly formatted text have a high noise to information ratio. They are poor candidates for traditional application of relational database architecture.

XML or Ontology based Software Development methodologies provide for more flexibility in creating and maintaining document based clinical trial data. But, data with a high noise to information ratio will remain that way no matter how one stores it. Work, often hand work, needs to be performed to change that ratio. XML has an advantage here in that you can easily define the information you want to add and add it without negatively impacting the original data and irrespective of the storage mechanism.

Traditional application of client server era database technologies takes data from a document contained within an ontology, and inserts it in columns and rows in a database without a rich enough association with other data that could be (but not always) found in the same document or document category. These data to data relationships are complex and require ontology based tagging to support search.

Etienne Taylor
CTO
Clinical Data Labs, Inc.

http://etiennetaylor.com

Share →

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>