February 21, 2018

TRI Co-Sponsored CTSA Workshop Using AI to Mine Research Data

As an ontologist, UAMS’ Mathias Brochhausen, Ph.D., teaches computers the meaning of words. He takes a term, such as “informed consent,” and writes a definition in language the computer can interpret.

“Here’s something about ontologies that is very cool: Ontologies are actually an artificial intelligence product,” he said.

Brochhausen’s expertise and enthusiasm for ontology/AI solutions will be in play Feb. 26 and 27. He will be hosting and collaborating with a national group of researchers to create semantically-enabled products that support access to more data across research institutions.

The work will be performed during the 2018 Clinical and Translational Science Ontology Group Spring Workshop, “Ontology of Informed Consent: An Approach to Specimen and Data Sharing.”

The workshop is supported by Department of Biomedical Informatics, where Brochhausen is an associate professor, the UAMS Translational Research Institute (TRI), and the NIH National Center for Advancing Translational Sciences (NCATS) Clinical and Translational Science Awards (CTSA) Program.

As a piece of code or software, an ontology can use logical inferences to bypass barriers caused by the absence of agreement in biomedical data representation. For example, Brochhausen said, consider the wide variation in informed consent language that tells researchers whether they may re-contact participants.

“Using ontologies, we can define re-contacting, and everything that falls under that definition the computer will automatically sort into re-contacting,” he said.

The field of ontology has strong roots at UAMS and the Department of Biomedical Informatics. The discipline has also continued to grow along with the department; Brochhausen now oversees the research group Biomedical Ontologies Arkansas (BOAR).

“It’s an exciting time for UAMS to host this workshop,” he said. “The spirit of collaboration with clinical scientists, which is so strongly emphasized by the TRI, really makes UAMS stand out.”