Ramin Javan MD, Presenter: Nothing to Disclose

Richard Duszak MD, Abstract Co-Author: Nothing to Disclose

Asim F. Choudhri MD, Abstract Co-Author: Nothing to Disclose

Devin Traer Caywood MD, Abstract Co-Author: Nothing to Disclose

A simple prototype decision support tool within PACS was designed and implemented, for simplifying making recommendations regarding intra-abdominal 'incidentalomas'. This prototype can be applied to any aspects of radiology where multi-step flowcharts exist.

Emerging decision support tools utilize a vast array of patient and literature data to facilitate improve decision making and the practice of evidence-based radiology. Expert consensus algorithms, such as those of the American College of Radiology (ACR) Appropriateness Criteria (AC) are ideally structured to serve as the backbone of decision support concepts. Given the frequency with which they are identified, we have chosen intra-abdominal 'incidentalomas' from the ACR AC, which provide guidance for a very commonly seen concept in day-to-day imaging.

Decisions support is better received and more likely to increase efficiency when it is easily accessible and easy to use. An end-user interface, available whileviewing images within PACS, rather than on the web or as a separate application would facilitate use, particularly if structured in a simple and intuitive. A PC-based plug-in is created, which could ideally be incorporated into PACS during image interpretation. After choosing one of four available organs (liver, spleen, kidney and pancreas), the system facilitates lesion counting and size and attenuation measurements to facilitate a differential diagnosis and recommendations, which can be inserted into the report as a statement. Microsoft Visual Basic 6.0 was used to create this application. The flowcharts used are available in the application under separate tabs.

Decisions support tools can theoretically be implemented in radiology wherever flowcharts exist. Ideally, an application can be created once the decision tree is provided as its input, as long as the flowchart data and the 'conditional' boxes or 'if' statements are indicated, similar to a programmer's flowchart. Parts of this process can be automated using Computer-Aided Detection (CAD), such as detection of organ, density and lesion size.

Javan, R, Duszak, R, Choudhri, A, Caywood, D, Simulation of a Smart Decision Support Tool within PACS for Intra-abdominal 'Incidentalomas' Based on ACR Appropriateness Criteria.  Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL. http://archive.rsna.org/2011/11016351.html