Free-text radiology reports and structured documentation processed through a commercially available differential diagnosis-generating CDS tool provide alternate methods of synthesizing data to form an interpretation. This study represents initial steps to combine these methods by testing the performance of extracting CDS-specific features from radiology reports using natural language processing (NLP).

Learning Objective: After participating in this session, the learner should be better able to:
Understand challenges and potential solutions of representation of radiologist interpretations in structured and unstructured forms.

Authors:

Justin Rousseau (Presenter)
Dell Medical School at The University of Texas at Austin

Juan Diego Rodriguez, University of Texas at Austin
Jared Abrams, University of Texas at Austin
Alex Liu, University of Texas at Austin
R. Nick Bryan, Dell Medical School at The University of Texas at Austin

Presentation Materials:

• Abstract/Manuscript - 1