Many radiographers don’t understand how smart computer systems diagnose problems


A new study shows that many UK radiographers have a limited understanding of how new smart computer systems diagnose problems detected on scanners such as X-rays, MRIs and CT scans.

Artificial intelligence (AI) is about to be more widely introduced in radiology departments. This research shows that we need to educate radiographers so they can be sure of the diagnosis and know how to discuss the role of AI in radiology with patients and other healthcare professionals.

Clare Rainey, Principal Investigator

Radiographers are the specialists that patients meet at the time of the examination. They are trained to recognize the variety of problems found on medical scans, such as fractures, joint problems and tumors, and are traditionally seen as the bridge between patient and technology. There is a severe nationwide shortage of radiographers and radiologists, and the NHS is set to introduce AI systems to help with diagnosis. Now, research presented at the UK Imaging and Oncology Conference in Liverpool (with concurrent peer-reviewed publication – see below) suggests that, despite the impressive performance reported by AI system developers, many radiographers don’t know not know how these new smart systems work.

Clare Rainey and Dr Sonyia McFadden from the University of Ulster asked reporting radiographers about their understanding of how AI works (a “reporting radiographer” provides formal reports on X-ray images). Of the 86 radiographers surveyed, 53 (62%) said they were confident However, less than a third of respondents would be confident in communicating the AI ​​decision to stakeholders, including patients, caregivers and other healthcare professionals health.

The study also found that if the AI ​​confirmed their diagnosis, 57% of respondents would overall have more confidence in the result, however, if the AI ​​disagreed with their opinion, 70% would ask for an additional opinion. .

Clare Rainey said:

This survey highlights issues with British radiographers’ perceptions of AI being used for image interpretation. There is no doubt that the introduction of AI is a real step forward, but it shows that we need resources to go into x-ray education to ensure that we can get the most out of this technology. Patients must have confidence in how the radiologist or radiographer arrives at an opinion“.

Modern forms of AI, where computer systems learn as they go, are popping up in many places of daily life, from self-learning robots in factories to self-driving cars and self-landing planes. Now the NHS is preparing to introduce these learning systems into its imaging services, such as X-rays and MRIs. These computerized systems are not expected to replace the final judgment of a trained radiographer, but they can provide a high level first or second opinion of x-ray results. This will help reduce the time needed for diagnosis and treatment, as well as give a “belt and suspenders” to the human decision.

Clare Rainey said:

“It is not strictly necessary for radiographers to understand everything about how these AI systems work; after all, I don’t understand how my television or smartphone works, but I know how to use them. However, they must understand how the system works makes the choices it makes, so they can both decide to accept the results and be able to explain those choices to patients.”

Clare Rainey unable to travel to Liverpool, this work was presented at the UKIO by Dr Nick Woznitza. Dr. Woznitza commented:

“AI is really a range of techniques, which can have an exciting impact on what scans can tell us. My own group is working on how AI is applied to lung scans, which has the potential to help diagnose conditions ranging from lung cancer to COVID“.

UKIO Chairman Dr Rizwan Malik (Bolton NHS Foundation Trust) commented:

“Radiographers are positive about the introduction of AI, but like any new technology, there is a learning process. As the authors point out, this requires more investment in targeted education and training The introduction of artificial intelligence promises the NHS more efficient and cost-effective use of radiology resources, as well as a more reassuring experience for patients. We need to ensure that this investment in education and training is widely available to all radiographers to ensure we get the most out of the technology.”


British Congress of Imaging and Oncology (UKIO)


Comments are closed.