Abstract Archives of the RSNA, 2011
LL-INS-TH5B
Using the Semantic Web for Radiology Decision Support of Focal Liver Disease
Scientific Informal (Poster) Presentations
Presented on December 1, 2011
Presented as part of LL-INS-TH: Informatics
Ihsan Djomehri, Abstract Co-Author: Nothing to Disclose
Daniel L. Rubin MD, Presenter: Grant, General Electric Company
Our goal is to develop a system leveraging Semantic Web technologies to deliver knowledge to radiologists “just-in-time” as they are interpreting images to help them tackle the challenge of accessing the vast amounts of radiology knowledge. Our system may reduce variation in radiological practice by providing radiologists the knowledge they need.
We developed an ontological representation of radiological knowledge of focal liver disease diagnosis, encoding the information from two independent knowledge sources (Dahnert’s textbook and STATdx from Amirsys Inc.). The model encodes diagnoses, clinical contexts, radiology image features of liver lesions, and probabilistic relationships among entities. We built a computer reasoning application that accesses this knowledge resource (ultimately to be deployed on the Semantic Web), receives input about the clinical context and image features, and outputs the most likely diagnoses. We evaluated the system by providing it several cases of known diagnosis.
Our ontological model of radiology successfully captured all the major pieces of information in the knowledge sources. Our decision support application generated a correct differential diagnosis based on observed imaging features in each of the test cases. We are now conducting a study to evaluate the ability of our tool to reduce variation in performance of radiologists interpreting cases of focal liver disease.
We have demonstrated the feasibility of encoding radiological knowledge in a human-readable, machine-accessible ontological form. We also showed the feasibility of using this knowledge to deliver decision support. Our methods may help to improve radiologist diagnostic performance.
Our methods could improve the interpretation of imaging studies and improve patient care.
Djomehri, I,
Rubin, D,
Using the Semantic Web for Radiology Decision Support of Focal Liver Disease. Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL.
http://archive.rsna.org/2011/11008276.html