The Integration of AI and Smart CT Technology

Editor’s note: Ryan Lee, MD, MBA, MRMD, is the Chair of Department of Radiology at Einstein Healthcare Network, now part of Jefferson Health, in Philadelphia. He moderated an expert forum sponsored by Bracco Diagnostics Inc., the US subsidiary of Bracco Imaging S.p.A., and Applied Radiology. This article is based on Dr. Lee’s presentation and the group’s resulting discussion.

As the chair of radiology at Einstein Healthcare Network (now part of Jefferson Health) in Philadelphia, Pennsylvania, Ryan Lee, MD, MBA, MRMD, often considers how artificial intelligence (AI) can improve patient care. While many radiologists may think AI is useful only for image interpretation, Dr. Lee says it can intersect with the specialty in many other ways.

“One of the really cool things about the development of sophisticated technologies such as AI is that it allows the radiologist to participate even more with the patient,” he said.

Dr. Lee stressed that AI could impact the entire spectrum of patient care, both inside and outside of radiology. For example, AI software that uses natural language processing in conjunction with appropriateness criteria can enhance clinicians’ ability to determine the best study for a given patient’s symptoms.

AI-driven clinical decision support tools can also help clinicians decide when to choose another course of action. At Einstein, radiologists collaborated with a CDS vendor to determine when to order computed tomography (CT) scans for pediatric head trauma cases.

“A few clicks based on the PECARN (Pediatric Emergency Care Applied Research Network) traumatic brain injury algorithms tell us whether we should perform a CT of the brain or not. This is an example of how CDS can potentially reduce unnecessary utilization,” said Dr. Lee.



AI and contrast injection

There is significant potential for using AI on smart CT injector platforms, such as Bracco’s EmpowerCTA®+ Injector Systems, which can optimize contrast administration and personalize contrast dosing for patients.

“Another use for AI is in personalizing contrast dose for a particular patient using a patient's body weight and body habitus, and perhaps their comorbidities and other information provided in the medical record. There is a lot of potential in looking at different types of software and injector combinations in order to achieve the best study for our patients,” Dr. Lee said.



Richard Hallett, MD, chief of cardiovascular imaging at Northwest Radiology, a large private radiology group in Indiana, also believes AI will support radiologists in their daily activities and improve the efficiency of the CT department, thus driving patient care to a higher level.

“AI is a way to make us better and more efficient, and it's a way to make the department run better so patients get better care. We can spend more time doing the things that actually require a higher level of radiology thinking.,” he said.



Ehsan Samei, PhD, a professor of radiology at Duke University and chief physicist at Duke University Medical Center, agreed with Drs. Lee and Hallett that AI will only augment radiologists’ work. “AI is learning from what we have already been doing. It’s not a magical wand that we turn on and suddenly make things better. AI gives us an opportunity to learn from the things that we already know,” he explained.

Dr. Samei recommended using AI to reduce variabilities in contrast dose. “We can use AI to collect information consistently so the level of enhancement is independent of contrast rates so we can focus on what is actually wrong with the patient. Then we can make our approach more personalized, patient-specific, and consistent,” he said.



Ultimately, Dr. Lee expects the integration of AI and radiology to drive better patient care. “The possibilities are endless in how we as radiologists can help take care of the patient in their entirety,” he concluded.

© Anderson Publishing, Ltd. 2024 All rights reserved. Reproduction in whole or part without express written permission Is strictly prohibited.