Interoperability of data and knowledge is fundamental to quality clinical practice and advancing research. But these efforts are limited by data gathered using different terminologies and information types.
In radiology, most of the information that conveys the results of imaging studies is in an unstructured, narrative text format, either in radiology text reports or raw images themselves, according to lead author Dr. Daniel Rubin of Stanford University.
"One approach to improving data interoperability is to use 'common data elements' as a way to collect data in a standardized, interoperable way," Rubin told AuntMinnie.com.
A common data element defines the attributes and allowable values of a unit of information and enables interoperability of that information with other systems. For example, common data elements could describe properties of a lung mass, such as its anatomic location, shape, image number, coordinates on that image, and maximum diameter, Rubin said.
"The benefit for using [common data elements] for collecting the information is interoperability of that information across systems, enabling mining of those data for quality assurance/improvement, to improve reporting (e.g., detect omissions or inconsistencies in reports), and to enable research," he said.
Rubin noted that introducing common data elements wouldn't mean that radiology workflow would need to be changed. For instance, the radiologist could still produce narrative reports, but the system could recognize that a mass was being described and then save that information as a common data element instead of just raw text. Or, a structured reporting template could use common data elements to populate lists that are provided to radiologists for use in describing abnormalities, Rubin said.
"The takeaway is that our work, which is sponsored by the RSNA, will deliver a set of standardized [common data elements] for the radiology field which, when incorporated into radiology systems (e.g., reporting), will improve the ability of systems to help the radiologist do a better job at image interpretation," as well as help with quality improvement and other applications, Rubin said.
Learn more by visiting this poster presentation on Sunday.
