The National Institute for Health and Care Excellence (NICE) has released draft guidance on how AI tools can be used to examine X-rays and help identify potentially missed fractures.
While the SoR welcomes an opportunity to safely test new technologies, it has emphasised the importance of human skill in image interpretation and in image reporting.
It is a legal requirement that the final report under the Ionising Radiation (Medical Exposure) (Amendment) Regulations 2024 is undertaken by an appropriately educated and trained entitled professional.
In addition to a trained healthcare professional reviewing the X-ray, an independent NICE committee has said one of four AI tools can be used to help detect fractures on X-rays in urgent care.
Enhancing patient care
SoR supports NICE’s draft guidance on using AI to help detect fractures in urgent care when used to complement the expertise of reporting radiographers and radiologists, and where the evidence base demonstrates efficacy.
Charlotte Beardmore CBE, executive director of professional policy at the SoR, said: “These AI tools can potentially improve accuracy but will not replace the expertise of radiographers and radiologists. With the current pressures on diagnostic services and high vacancy rates, AI can assist professionals by reducing the potential for missed fractures and enhancing patient care.”
The recommendation allows TechCare Alert, BoneView, RBfracture or Rayvolve to be used in urgent care settings in England while further evidence is generated.
'Something the machine can't do'
NICE suggested that using AI technologies may help reduce variation in care across the country, reduce the number of fractures that are missed at initial presentation, and prevent further injury or harm to people during the time between the initial assessment and a decision on further treatment.
Richard Evans, CEO of the SoR, told BBC Radio 4’s Today programme: “AI provides an extra set of eyes to look at the images that are produced. Trained healthcare professionals are still going to be a vital part of the chain, the radiographer or the assistant practitioner will take the images, but too often there are few people around in emergency departments who are fully trained to give a definitive report on those images.
“Missed fractures are a problem, for the patient, for the healthcare service generally, because people develop longer-term conditions. Sometimes fractures just aren’t visible on the images at all. This is where radiographers are uniquely able to act on clinical evidence and experience to investigate further. That’s something the machine can’t do.”
“We expect, and as we’ve always seen, technology to develop and become better at assisting the healthcare process, that’s what we all support and want to see,” added Mr Evans. “I don’t see that the human interaction the radiographer provides with the patient can be overtaken by the machine just yet, however clever they get and however much we come to rely on them, which I’m sure we will more and more.”
Working in isolation?
A NICE committee noted that the use of these technologies is considered low risk, as each image is reviewed by a professional and the AI is not working in isolation. This makes it unlikely using the technologies will lead to an increase in unnecessary referrals to fracture clinics.
An independent review carried out by the radiology department will still take place.
After the consultation, responses will be considered by the committee, which may then alter its recommendations before final guidance is published on the NICE website.
Managing the workload
Mark Chapman, director of HealthTech at NICE, said: “Every day across the NHS thousands of images are interpreted by expert radiologists and radiographers, but there is a high vacancy rate within these departments across the country and more support is needed to manage their workload.
“These AI technologies are safe to use and could spot fractures which humans might miss given the pressure and demands these professional groups work under.
“Using AI technology to help highly skilled professionals in urgent care centres to identify which of their patients has a fracture could potentially speed up diagnosis and reduce follow-up appointments needed because of a fracture missed during an initial assessment.”
A consultation on the draft recommendations has begun and comments can be submitted via the NICE website until Tuesday 5 November 2024.
Innovation with staff involvement
The Ionising Radiation (Medical Exposure) (Amendment) Regulations 2024 No 896 (2024) state that clinical evaluation of X-rays requires a trained person, therefore AI technologies cannot be used autonomously without human interpretation. The technologies could, however, speed up the flow of people through the care pathway and reduce the likelihood that a fracture is not detected.
For the detection of musculoskeletal fractures there are a range of various classes of CE marked technology and AI-based software on the market and in use in healthcare. Their intended use cases are variable, and more evidence is required to establish trustworthiness across a range of settings and populations. The SoR supports this innovation with the involvement of clinical staff and patients and careful consideration of the emerging evidence base.
All results will continue to be reviewed by entitled professionals, as this is a requirement of the regulations. The Society remains committed to ensuring the safe and effective use of AI in radiography, alongside appropriate training and support for radiographers.
(Image: Photo by Douglas Sacha via GettyImages)