4 key reasons AI won’t replace radiologists

Countless articles have been written over the years about whether or not AI technologies will ever replace radiologists for good, leading to a mass extinction like something out of the Jurassic period. According to a new commentary in Radiology: Artificial Intelligence, however, radiologists have no reason to fear being replaced—as long as they are willing to embrace AI and adapt to these changing times.  

Author Curtis P. Langlotz, MD, PhD, department of radiology at Stanford University, wrote about this subject in detail. These are four reasons, according to his commentary, that AI will not be replacing radiologists:

1. Radiologists have been down this road before

Langlotz noted that this isn’t the first time radiologists have been faced with “this supposed awful adversary.” Computer-aided detection systems exploded in popularity in the 1990s, though research suggests they didn’t necessarily make a big impact on radiologist accuracy one way or another.

“The recent rush of novel AI algorithms should prompt introspection about past failures of AI to live up to its promise,” Langlotz wrote. “Today’s AI tools have achieved regulatory clearance based on their performance at a small number of health care organizations. Perhaps the incremental accuracy of these new AI methods will reduce false-positive findings and blunt the ‘cry wolf’ effect, but the generalizability of these algorithms to the diversity of radiology practices remains an open question.”

2. Radiologists know how to adapt and take charge

History, he added, is full of moments when the specialty was supposed to be in trouble—but it just never actually happened. Some thought MRI technology was going to replace radiologists, for example, because physicians would immediately know everything they needed without any further explanation. What actually happened, though, was that radiologists ended up learning a great deal about MRI scanners—their strengths, their weaknesses, and it has helped the specialty provide more value than ever. A repeat of that exact same series of events is now underway, with radiologists taking the lead on AI and once again demonstrating what the specialty can bring to the table.

3. An accurate algorithm ≠ an accurate radiologist

“We often compare AI algorithms to radiology experts based on the ability to identify a single disease or a small set of diseases,” Langlotz wrote. “These assessments dramatically oversimplify what radiologists do. A comprehensive catalog of radiology diagnoses lists nearly 20,000 terms for disorders and imaging observations and over 50,000 causal relations.”

Algorithms that can help diagnose common conditions are a “major step forward,” he notes, but a radiologist is looking for numerous conditions all at once while also keeping an eye out for anything else suspicious that might show up in a patient’s test results.

“AI is impressive in identifying horses,” Langlotz added,” but it is a long way from recognizing zebras.”

4. Autopilot didn’t replace pilots

One of the more accurate metaphors for how AI will impact radiology, according to Langlotz, is how autopilot has impacted pilots. The pilot may turn on autopilot while in the cockpit, allowing it to handle “tedious or repetitive tasks,” but what if the system has a malfunction or there’s a horrible storm on the horizon? The pilot is right there, able to take over as needed.

Overall, Langlotz concluded, AI is destined to “profoundly change” the practice of radiology—but there will still most certainly be a need for radiologists.

“The ethereal notion of an artificial general intelligence destined to replace us is just as fanciful today as attaching human qualities to submarines,” he wrote. “As we are lifted by the latest AI bubble, ‘Will AI replace radiologists?’ is the wrong question. The right answer is: Radiologists who use AI will replace radiologists who don’t.”