Advances in AI technology for image reporting risk dehumanising patients and creating a “conveyor belt” system, the Society of Radiographers will hear today.
AI software algorithms are currently used in clinical practice to assist reporting radiographers and radiologists in reading X-ray, CT and MRI images and writing clinical reports. For patients on certain pathways, this can help speed up diagnosis times.
However, there is a risk that patients will become dehumanised in the process, radiographers will tell the SoR’s Annual Delegates’ Conference.
“Phrases such as ‘patients on and off like a conveyor belt’ risk dehumanising patient care, shifting the focus from holistic, person-centred care to purely mechanistic view of healthcare delivery,” they will tell delegates at the three-day conference in London.
Elaborating on these remarks, Tracy O’Regan, the SoR’s professional officer for clinical imaging and research, says: “There’s no simple patient. Even someone coming for a simple procedure, like a thumb X-ray, might have additional needs or care requirements. For example, they may be neurodiverse or living with dementia, which can make the procedure more complex.
“Or you might have an elderly person come in for a scan, and the only people they’ll speak to that day will be the bus driver, the receptionist and you. That person isn’t there just for that image – they’re there for the human connection.”
And, she adds, patients may be unable to follow instructions without assistance: “We always say to students: ‘You can’t just tell a patient to lie on a bed with their head on a pillow.’ They’ll lie down with their feet on the pillow – or they’ll do something else entirely. When a patient is very worried or anxious, they can struggle to comprehend instructions.
“The job of a radiographer is not just about acquiring images. That notion dehumanises the person in front of us. It’s also about knowing how to read someone – when the patient is overwhelmed; when they need you to say something; when they need you to be silent.”
The need for these skills has only been heightened by new AI technology that speeds up diagnosis for certain conditions. For example, AI has reduced the time it takes for a patient with lung nodules demonstrated on X-ray to receive a CT scan and – for some people – a subsequent diagnosis of lung cancer.
“Radiographers would not routinely give unexpected news to someone,” said Dr O’Regan. “But in those hospitals where they’re speeding up the pathway and patients are going straight to CT, this has changed.
“Morally, if we can use technology to speed up diagnosis times, we should. That waiting time for people is horrendous – time slows down, and one week can feel like six months. Speeding up the pathway makes a huge difference for patients.
“But for some patients, that potential diagnosis is a massive shock. They’re looking to the radiographer taking them to CT for counselling and support.
“Advances in technology are opening up a new need. It’s making radiographers think: ‘What am I providing to make things better for these patients? What skills or development do I need?’
“Ongoing investment in radiography education and training is essential if radiographers are to make the best use of new technology and provide the best and most timely care for patients.
“The need for these person-centred skills doesn’t lessen with advances in technology – it grows. These skills are the complex part of the job. They’re the holistic art of radiography, as opposed to the AI-enabled science of radiography. It’s an art and a science, and they must both work together.”