Is AI Replacing Radiologists? Doctors Weigh In on AI’s Impact

The rise of artificial intelligence has sparked debate across numerous professions, prompting questions about the future of work. Even roles once considered safe from automation, like those in healthcare, are now being scrutinized. Specifically, the question of whether AI will replace skilled radiologists is gaining traction, fueled by rapid advancements in medical imaging technology.

Since the 2010s, breakthroughs in AI have led to the development of models capable of diagnosing conditions with accuracy comparable to healthcare professionals. Medical imaging is one of the fastest-growing domains in healthcare, and the commercialization of AI-powered imaging tools is accelerating. Between 1995 and 2024, the US Food and Drug Administration authorized 950 AI products, with a significant surge – 690 authorizations – occurring between 2016 and 2024 alone, compared to just 33 in the preceding two decades.

Detecting Breast and Brain Abnormalities: Current Realities

Recent investigations into the use of AI in the UK’s National Health Service (NHS) reveal a more nuanced picture than outright replacement. AI products are being developed to detect abnormalities like tumors in breast X-rays and vessel blockages in brain CT scans – crucial indicators of breast cancer and stroke. However, current UK regulations restrict full automation due to a lack of robust evidence supporting its effectiveness. Instead, these tools are primarily used to support the decisions of consultant-level professionals.

Radiologists interviewed in a recent study expressed mixed feelings about AI’s performance. Although hospital audits may suggest high accuracy, professionals often find discrepancies between AI analysis and their own clinical judgment. Dr. A, a consultant neuroradiologist, noted, “The AI is theoretically useful, but actually in practice… I found it not as accurate as, or doesn’t necessarily correlate with, what my analysis would be.” Dr. D, a consultant stroke physician, highlighted a specific issue: “[An image]… comes through, where [AI] has clearly interpreted bone, which is white on CT, as being blood, which is also white on CT.”

A crucial skill emerging for healthcare professionals is the ability to selectively evaluate AI outcomes, recognizing potential biases and ensuring accurate interpretation. As Dr. A explained, “it’s very easy to look at that [the pictures] face value and say, ‘OK, this is what it’s telling me, and therefore this is correct’. … but you need to be able to selectively choose what is relevant, and that is a skill in itself—not to get overwhelmed by the information that you’re given and to know what is relevant.”

Currently, AI functions as a decision-support tool, augmenting rather than replacing existing tasks. Radiologists report that AI can prompt a second look at potentially abnormal areas, even those initially missed. Dr. S, a consultant stroke physician, shared, “When it [AI] picks up any abnormalities, it makes us think twice, basically to build sure that that area is either abnormal or not abnormal.” Dr. J, also a consultant stroke physician, added, “Sometimes I have missed very small areas, for example, and the AI has picked it up.”

Easing the Burden: Task-Level Automation and Workforce Shortages

While complete automation isn’t yet a reality, task-level automation – automating specific aspects of image analysis – is becoming increasingly feasible. This could be particularly beneficial given the current workforce shortages in radiology. Rather than leading to redundancies, automation could alleviate pressure on training and recruitment. Dr. D, a consultant stroke physician, believes, “We’re so grossly understaffed in the UK for radiology that I don’t think we need a reduction [of radiologists]. We probably don’t need a huge amount more [radiologists], because the diagnostic work will slowly drop off.”

The potential for automation extends to interventional radiology, where real-time imaging guides procedures like tumor removal and clot retrieval during stroke. Dr. L, a consultant interventional neuroradiologist, noted, “[AI] is very useful for streamlining the workload for stroke intervention, and also for aneurysm work.”

However, task-level automation also presents challenges. If AI takes over routine image analysis, radiologists may have fewer opportunities to maintain and refine their skills for more complex cases. Dr. J, a consultant breast radiologist, expressed concern: “That’s a big worry … If AI does all the easy stuff, you don’t know what normal looks like anymore, and that becomes difficult, because you should be trained on what’s normal, or a combination of both [normal and abnormal] . If AI automates half the analysis, you become less good at assessing, because you’re not seeing so many and not so familiar with the bigger range.”

A Shifting Landscape: Work Practices and AI Integration

The evolving relationship between medical imaging work and AI mirrors trends observed in other sectors, such as accounting, finance, and manufacturing. Rather than mass job displacement, the structure and practices of work are changing at a more gradual pace than initially predicted. Research suggests that the benefits of AI – such as increased efficiency and reduced workload – are strongest with moderate use, rather than complete automation or a lack of integration.

While more dramatic implications are possible if automation intensifies, they are not inevitable. Some organizations have even scaled back automation efforts due to cost and integration challenges, as seen with Amazon’s decision to discontinue its “Just Walk Out” technology in some grocery stores.

The future of work in radiology, and healthcare more broadly, remains uncertain. However, current evidence suggests that AI is more likely to augment the roles of skilled professionals than to replace them entirely. Continued research and careful implementation will be crucial to navigating this evolving landscape and ensuring that AI serves to enhance, rather than diminish, the quality of patient care.

Yuxuan Wu is the Editor’s Choice award winner in Vitae’s 2025 Three Minute Thesis competition sponsored by The Conversation UK.