RSNA 2017: How to provide value by creating data-enabled radiology reports

Radiologists have been working to improve radiology reports in recent years so that they can provide more value and bring significant improvements to patient care. Monday, Nov. 27, at RSNA 2017 in Chicago, Tarik Alkasab, MD, PhD, radiology service chief of informatics and IT at Massachusetts General Hospital in Boston, told attendees about the many benefits of building data-enabled radiology reports for referring physicians.

“We need to find a way to extract the value we get from the very rich imaging data we are seeing and make it more accessible in the context of the data-driven, team-oriented practice of medicine we are seeing in the 21st century,” Alkasab said. “The way I propose we move to that is to think of our radiology reports of not just a layer of text, but of two layers—a text layer and a layer that is essentially structured data that conveys the kind of information we are extracting.”

So what kinds of data gets included in a data-enabled radiology report? Alkasab explained that it is primarily measurements (liver density, nodule location), extractions (tumor volume, stroke volume) and categorizations (BI-RADS category, hemorrhage increased in size).

By including this information in its own layer, he added, it benefits referrers, research repositories, payer systems and, yes, radiologists themselves.

“One of the things that this helps us do as radiologists is be more clear about the value we are providing,” Alkasab said. “It makes it easier for us to demonstrate where radiology is increasing the value of the care that is being provided and can help us combat the notion of radiology as a cost center.”

In addition, he added, creating these data-enabled reports can help auto-generate portions of future radiology reports and it makes it easier to compare exams with one another.

Alkasab also explained that these reports require radiologists to become “data wranglers,” taking data from a variety of sources and then choosing the appropriate information for the data layer of their reports. They should also, he added, confirm that this data follows “the three Cs” of radiology reports; it must all be correct, complete and confirmed.

Alkasab concluded the session with one last key point: Radiologists should not be looking at the concept of creating data-enabled radiology reports or becoming a data wrangler as if it is additional work. As AI and deep learning technologies evolve, radiology departments will find that certain time-consuming tasks are no longer on their plate, meaning specialists have more time to build these high-quality reports and showing support for their referring physicians.

Michael Walter
Michael Walter, Managing Editor

Michael has more than 16 years of experience as a professional writer and editor. He has written at length about cardiology, radiology, artificial intelligence and other key healthcare topics.

Trimed Popup
Trimed Popup