There can be few healthcare professions that stand to be altered as fundamentally by artificial intelligence (AI) as radiography. Put simply, AI is the perfect tool to automatically highlight abnormalities, segment anatomies, and compare results to reference values.
Academic radiographer Dr Christina Malamateniou is clear that the future of radiography is going to be dominated by AI. Does she see that as a threat to the profession? Her answer is typically nuanced.
“Radiographers will not be replaced by AI, the radiographers who know about AI will replace those who don’t. The role of training and education on AI should therefore be emphasised for a smooth AI integration.”
And Christina should know. She has taught medical imaging and research methods in seven universities and three countries (including as visiting lecturer at the Technological Educational Institute of Athens and University College Dublin).
She has been recently appointed Visiting Professor of Radiography in the HESAV University of Lauzanne, in Switzerland.
Having worked as an academic and researcher in medical imaging for the past 19 years, she was previously an MRI radiographer at the University Hospital of Athens, where she qualified.
She gained her PhD from Imperial College London, specialising in perinatal MRI at the Robert Steiner MRI Unit at Hammersmith Hospital.
Along the way she has picked up the Clinical MRI prize and Leonard Levy prize from the British Institute of Radiology for work presented at the UK Radiological Congress in 2006 and has been a senior fellow of the Higher Education Academy since 2018.
Christina has been a visiting lecturer for postgraduate medical imaging courses at King’s College London since 2016 and guest lectured at UCL in 2018. She is currently the director of the postgraduate programme in radiography at City, University of London, and supervises master’s and PhD students.
As the chair of the Society of Radiographers’ AI Working and Advisory Group, Christina is ideally placed to judge the impact of the new technology on the radiography workforce.
“AI is just a tool that can be used for better or worse – and that is where professionals’ responsibility lies. It’s important to educate ourselves and understand what AI means. Many see it as a threat to jobs,” she says.
But she also points to her research, which found a significant gender difference in perceptions of AI: “Women showed significantly less confidence in using AI equipment [than men]. That doesn’t mean they are less competent, it means they felt less confident. Because radiography is heavily female dominated, this might have implications. If women feel less confident, they might choose to change careers.”
While AI is being integrated across all radiography and radiotherapy applications, Christina says it is in areas such as breast and lung cancer screening where its ability to generate usable images at low radiation doses comes to the fore.
She feels strongly that radiographers have a pivotal and exciting role to play in the development of AI-based applications: “It’s important to be involved in the decision- making for AI tools. Radiographers are the interface between technology and the patient and this is the most exciting position to be in now. If they can explain AI to the patient, there will be more trust. Explainable AI is one of the priorities in clinical practice and they can be the AI champions for radiology.”
The expert patient
While AI is moving forward at pace, Christina cautions that patients can often be forgotten in the rush to improve throughputs and manage the backlog Covid-19 has created in radiography.
“The patients we have today are not the patients we had 50 years ago,” she says. “We have a highly educated population. People have access to Google and academic papers and, very often, they come to us with questions we can’t answer. We should be involving them more in decisions relating to their diagnosis and care and allow them to be part of the design of AI tools. A patient with cancer knows much more than the professionals about the experience of cancer.
“I am a member of the British Standards Institute and we are trying to create auditable standards for AI in healthcare. Companies that develop AI technology will have to ensure proper accreditation and validation of their product, which will allow it to successfully pass any external evaluations for procurement.
“If you design things that don’t really work for the patients or the healthcare professionals who use them, the cost of trying to adjust the technology to suit them is huge. So you’d better get their input early on.”
Prenatal imaging
Prenatal imaging is an important area for Christina and she is involved in some fascinating research into parents’ experiences.
“We are trying to understand the experience of parents who undertake a fetal ultrasound or MRI. How does this affect the bonding of the parents, particularly those who are becoming parents for the first time?”
The research team that includes PhD student Emily Skelton is conducting surveys and interviews before and after scans to understand how healthcare professionals are perceived.
Was enough time available to discuss the results? Were parents’ fears allayed? Were they prepared for bad news?
“We are doing this in both normal and abnormal pregnancies. For both mums and dads because dads have been left out of that narrative,” she says.
While the prenatal scan can be a joyful occasion if all is well, it is also the point at which problems with the baby’s development are identified. Christina recognises the huge pressures on sonographers to maintain a delicate balance: “The duty of the healthcare professional is to ensure they protect the mother and baby. I would not like to be in the sonographer’s position where they tried so hard to deliver excellent service during the pandemic – it is hard to balance the patient’s safety and the humanisation of healthcare given all the imposed restrictions. Safety should always come first though.”
Powerful role models
Being a widely respected and successful leader in radiography, it troubles Christina that there are so few women in similar positions. “There are many more women in radiography than men – in the UK, the ratio is three to one – but the leadership positions are disproportionately held by men. This might be because women don’t put themselves forward so easily,” she wonders, citing the influence of ‘imposter syndrome’.
“I’ve never felt there was a glass ceiling. Because I had very strong role models, I just put myself forward for leadership positions.
“My role models started with my mum and there have been some amazing radiographers in Greece and the UK, and then researchers who have really fought for more research in radiography,” she says.
“It’s important that young women who come into radiography realise their potential. I’m trying my best to advocate for that and, hopefully, I will become a role model for women in my profession and beyond.”
Helping others
Christina is keenly aware how the pandemic put enormous pressures on frontline workers, and also on radiographer academics, researchers and students.
“I am in awe of the efforts everyone put in to adjust teaching and research to ensure continuity, and of our academic staff and student efforts to continue the learning during this challenging time. Everything we do in healthcare, and in radiography in particular, is about helping others. Radiographers showed again their true colours of love, kindness, loyalty and integrity and went above and beyond to do just that. I am hugely proud of all of my colleagues but I hope we will not be asked to do similar sacrifices in the future and things will resume some level of normality”
