Artificial intelligence (AI) will replace radiologists — but only the ones who don’t embrace it, according to a leading authority in radiology.

“The radiologist who is using AI is going to be replacing the radiologist who has not embraced AI,” said Rasu Shrestha, M.D., MBA, in a podcast published by Imaging Technology News (ITN).

Shrestha, who recently joined Atrium Health as chief strategy officer in charge of leading the development of its enterprise across the Carolinas, will speak at the upcoming Society for Breast Imaging/American College of Radiology Breast Imaging Symposium on April 4 at 10:30 am on the topic of “Artificial Intelligence in Radiology — Are We Doomed?” In his current roles with Atrium Health, he also leads strategic development of innovation.

For more than a decade, Shrestha directed the division of radiology informatics at the University of Pittsburgh Medical Center  (UPMC); is currently a member of the board of directors of the Healthcare Information and Management Systems Society (HIMSS) and chairs its Innovation Committee; and served as co-chair of the 2019 Health Datapalooza.


AI Could Help “Humanize” Patient Care

Artificial intelligence can be a huge enabler, he said in the ITN podcast: “In many ways (it can) make us superhuman. AI can augment the most humanistic elements of who we are as clinicians.”

Shrestha explained in the ITN podcast that smart algorithms can be made to do the tedious work — the mundane elements of medical practice — “that we are really struggling with today, that are causing a lot of burnout; lot of heartache; and are causing us to not focus attention on our patients.”

To get the most from AI, however, radiologists must pay attention to data science, he said. “We have really missed the boat in terms of evolving not just the medical school curriculum, but even in our residency and fellowship, and perhaps CME,” said the former director of UPMC radiology informatics.  “It is important for us to get into the weeds n terms of understanding what data science is all about.”

Radiologists need to understand that AI is here today.  “We are already leveraging AI in many different ways,” Shrestha said in the ITN podcast. 


How To Do Artificial Intelligence Right

AI, “if done right,” he said, could help humanize the healthcare — to help physicians overcome barriers that have — ironically — been put in place by other forms of technology.  Too often clinicians have their backs to their patients, Shrestha explained, “because they are entering notes into the electronic medical record system; they are documenting; and they are creating proof for billing.

“Those are transactional elements that are required. That is how healthcare is run today,” he said in the podcast, “but it is very artificial.”

AI might remove some of that “artificialness” by handling tedious and time-consuming tasks.  Doing so will give physicians the opportunity to connect better with their patients. It should not, however, interfere with physicians’ authority.

AI should “help us with our decisions,” he said.  “But the decisions should be ours.”

During his April 4 presentation at the SBI symposium, Shrestha hopes to separate rhetoric about AI from reason, “as I call it the hope to hype to here,” he said in the ITN podcast.  “We need to do that with evidence, with science, with a level of rigor.  At the same time we need to do it with clinicians and with patients in what I call the living lab of our everyday practice of delivering healthcare.”


Greg Freiherr is a contributing editor to Imaging Technology News (ITN). Over the past three decades, he has served as business and technology editor for publications in medical imaging, as well as consulted for vendors, professional organizations, academia, and financial institutions.


Related content:

PODCAST: Radiologists Must Understand AI To Know If It Is Wrong 

PODCAST: Fitting Artificial Intelligence Into Cardiology 

VIDEO: Artificial Intelligence in Radiology — Are We Doomed?

Technology Report: Artificial Intelligence 

PODCAST: How Technology Is Changing Cardiology 

Subscribe Now