RSNA 2022: As radiology confronts economic pressures, will AI ride to the rescue?

This week’s RSNA 2022 meeting concluded as radiology faces a crossroads. Economic pressures ranging from rising image volumes to staffing shortages to looming reimbursement cuts are threatening the specialty in existential ways.?

The question is, can radiology find a way out of the crucible before a crisis erupts??

The challenges are manifold. The backlog of scans that developed during the COVID-19 pandemic is just starting to clear, but is running smack into rising imaging volumes as the Baby Boom generation ages and requires more care.?

The Great Resignation that has plagued the rest of society has also had a ripple effect in healthcare, leading to severe job shortages for allied health professionals like radiologic technologists. While the supply of radiologists is more closely managed, recruiters are reporting that today’s job market for radiologists is unlike any they’ve ever experienced, according to Robert Schaffer of Radiology Business Solutions . Some sites are reporting that they have postponed the acquisition of new capital equipment because they simply don’t have the radiologists to read the increased volume a new scanner would produce.

Meanwhile, radiology can’t look for any help from the federal government. The specialty is currently slated to see cuts for Medicare and Medicaid reimbursement of 10%-11% starting January 2023, and it’s going to take an act of Congress (literally) to stave off the cuts – which would be devastating for providers, according to Sandy Coffta of Healthcare Administrative Partners .

There is a white knight on the horizon – artificial intelligence (AI). Originally seen as a threat to radiologists (remember when one pundit said we should stop training radiologists?), AI is now being positioned as a tool that can help radiologists deal with workflow challenges and operate more efficiently.

Nowhere was that more evident than at RSNA 2022. Despite rumors that venture funding for start-up AI firms is beginning to dry up, there was still an abundance of new AI vendors on the technical exhibit floor. The conference’s AI Showcase saw a mix of longtime radiology vendors dipping their toes into the waters of AI to AI companies that are familiar faces starting to generate revenue to companies exhibiting at the RSNA show for the first time.?

And while many AI developers demonstrated algorithms for traditional clinical tasks like image analysis and interpretation, a newer trend is emerging of AI applications that can analyze radiology operations, identify bottlenecks and workflow challenges, and help departments and imaging centers adapt. And this is where things get interesting.

AI developers have recognized that the main challenge that radiologists believe they face is not identifying hidden pathology on a scan, but getting through a worklist that seems to grow exponentially by the hour. At RSNA 2022, there was far less talk about analyzing clinical images and determining bone age and much more talk about breaking down workflow and business operations so imaging facilities can work more efficiently.??

Indeed, there are many other signs that AI is evolving from a hypothetical technology that may or may not steal jobs from radiologists and into a business tool that – like PACS and enterprise imaging – is becoming a critical part of how radiologists do their jobs. In a video interview from RSNA 2022, Dr. Paul Chang of the 美国芝加哥大学 described how he’s spending much more time “herding cats” in AI – dealing with nuts-and-bolts issues like data governance. And that’s a good sign, he believes.

That’s not to say there’s not still work to do with AI. Integration into radiology workflow is still a problem, according to Herman Oosterwijk of OTech Consulting in a video interview. Where does AI data come from, where does it go, and how does it impact workflow?

So, while RSNA 2022 highlighted the many challenges facing radiology, it also offers rays of hope, much like the rays of sunshine that occasionally break through the clouds of McCormick Place on a late November day. Future meetings will likely see the continued evolution of the technology as AI firms sharpen their focus on what the discipline really needs – as opposed to what they can train algorithms for.?

And while AI may or may not end up saving radiology, right now it’s our best hope.?

#radiology #RSNA22 #ImagingAI