Are automated CT exams on the way? Cardiologist shares concerns about radiation, regulation

“Only about a third of the world's population has access to this critical technology,” Powell said, noting that both image acquisition and interpretation currently depend heavily on trained professionals. “What if someday all of these medical devices and imaging instruments could become robotic, more and more autonomous, just like our cars are becoming more and more autonomous?"

Ed Nicol, MD, head of cardiovascular CT at Royal Brompton Hospital in London and a former SCCT president, thinks AI has tremendous potential to improve imaging workflows. However, unlike screening tests that carry little or no risk, CT exposes patients to ionizing radiation, requiring physician oversight and careful justification before scans are performed.

"I would never say never," Nicol said, noting how quickly technology evolves. "But I think it's definitely going to be a much more difficult environment."

Radiation regulations limit autonomous imaging

Unlike magnetic resonance imaging (MRI), CT examinations involve radiation exposure. Nicol said this creates legal and ethical barriers that do not exist for imaging modalities without ionizing radiation.

In the United Kingdom, CT use is governed by the Ionising Radiation (Medical Exposure) Regulations (IR(ME)R), which require medical justification before exposing patients to radiation. Similar principles apply in many other countries.

"X-rays are not without potential consequence. We know that there is an increased risk of cancer," he explained.

Those regulatory safeguards mean autonomous systems would need more than just technical capability, they would also need to satisfy legal requirements governing when imaging is appropriate.

AI still lacks clinical context

Nicol also questioned whether AI can adequately determine who should undergo imaging without understanding the broader clinical picture.

"What might it miss?" he asked. "AI cannot do everything. It needs some context sometimes. And that is what AI can't do, at least currently."

While AI algorithms have demonstrated impressive performance in image acquisition and interpretation, Nicol said physicians remain responsible for determining whether imaging is appropriate in the first place.

"We have a responsibility to say, 'Okay, it's interesting. Is it going to help us? Does it improve access? Does it improve healthcare overall? What's the downside?'"

Incidental findings can create unnecessary care

One of Nicol's biggest concerns is that wider CT screening could find more incidental findings that create a cascade of follow-up testing and can result from unnecessary imaging. He referred to the phenomenon jokingly known as "victims of medical imaging technologies," or VOMIT, where incidental findings trigger additional scans, specialist consultations and patient anxiety, despite ultimately being deemed harmless findings.

Beyond increasing healthcare costs, he said these situations can expose patients to unnecessary stress while offering little clinical benefit.

Screening decisions remain complex

Nicol said similar concerns apply to other proposals that appear straightforward on the surface but become more complicated when viewed through a clinical lens.

He cited suggestions that patients should simply receive emergency medications to self-administer if they experience chest pain. In reality, many heart attacks present with atypical symptoms, while many episodes of chest pain are caused by non-cardiac conditions such as gastroesophageal reflux disease or costochondritis.

Likewise, accurately identifying patients who should undergo cardiac CT requires careful clinical assessment rather than relying solely on symptoms or automated decision-making.

He also pointed to differences between U.S. and European chest pain guidelines, noting that definitions of appropriate testing vary depending on how clinicians define typical angina and assess pretest probability.

Although AI will continue improving workflow efficiency, scan acquisition and image interpretation, Nicol believes autonomous cardiac CT screening remains constrained by the need for physician judgment, regulatory oversight and careful patient selection.

For now, he said, the greatest opportunity for AI is likely to be assisting clinicians rather than replacing them in determining who should undergo cardiac CT imaging.