Q&A with Gemma Walsh: tracking radiographer attitudes towards AI

Gemma Walsh is a PhD research radiographer at City St George’s, University of London, a winner of a ‘We are the City’ Rising Star Award and a contributor to the recently launched City St George's Research Radiographer Group (CRRAG). Her area of focus is the topic that’s on everybody’s minds – AI in radiography.

Gemma Walsh accepting the award for ‘best radiographer scientific abstract’ at the European Congress of Radiology (ECR), in Vienna, March 2024 from Dr Andrew England

Gemma Walsh accepting the award for ‘best radiographer scientific abstract’ at the European Congress of Radiology (ECR), in Vienna, March 2024 from Dr Andrew England

With opinions divided between concern about the technology replacing staff and hope that it might help take some of the burden off teams’ shoulders, it can be difficult to know whether to celebrate AI’s advent or not. But one thing is clear: safe and effective adoption will require the full-throated support of the profession.

To better understand radiographers’ hopes and fears about the AI’s impact, Gemma helped establish the R-AI-diographers study, which surveys professionals from across Europe about what effect they think it will have on their roles.

Synergy asked her to share more details about the survey’s findings and its conclusions.

What is the R-AI-diographers study?

A major European study, R-AI-diographers, funded by the Society of Radiographers through the College of Radiographers Industry Partnership Scheme (CoRIPS), has provided the most detailed picture to date of how radiographers perceive the short, medium and long-term impact of artificial intelligence on their roles and professional identity.

Using a cross-sectional online survey design, the research explored how AI might influence the three main areas of radiographer professional identity: patient-centred care, technical expertise and wider professional responsibilities.

While the European study received responses of more than 2,200 participants, the largest study of radiographers in relation to AI, the dataset also included 322 valid responses from radiographers working or studying across the UK.

Even though most respondents (70.8 per cent) reported having no formal AI education or training, the overall sentiment was one of cautious optimism. More than 60 per cent expressed a positive outlook on the medium to long-term impact of AI on their careers, and a substantial majority (81.2 per cent) rejected the idea that AI would replace radiographers in the future.

Radiographers were particularly hopeful about AI’s potential to streamline workflows and enhance patient care and outcomes. However, concerns remained, which were more pronounced on the qualitative analysis of open-ended style responses. Respondents highlighted the risk of deskilling and voiced scepticism about the uncertainty of realising AI’s potential in efficiency gains, reflecting wider debates about the gap between the hypothetical and real-world value of AI tools.

The findings suggest that radiography is a profession that is open to technological change, but is aware of its challenges. Respondents emphasised the need for accessible AI education as a tool to build confidence and ensure AI strengthens rather than undermines radiographic practice. The study concludes with highlighting the need for greater radiographer involvement in the development, evaluation and implementation of AI systems, to ensure these tools are trustworthy, ethical and fit for clinical use, while also enabling radiographers to actively shape the future direction of their profession.

Gemma Walsh presenting the European results of this study at the RCR Global AI conference, London, February 2025

Gemma Walsh accepting the award for best research abstract in the ‘general interest’ category at the inaugural Royal College of Radiologists (RCR) Global AI conference, London, February 2025

Have perspectives on AI changed in recent years?

As the first study to explore how AI may reshape radiographer responsibilities, direct comparisons with past perspectives are difficult. What is clear, however, is that radiographer-specific AI education remains scarce, echoing findings reported as early as 2021 (Rainey et al, 2021).

In this study, most respondents (70.8 per cent) had still not accessed any formal AI education. When asked what would best prepare them for working with AI, the strongest theme was the need for dedicated AI education and training. This lack of structured learning continues to act as a major barrier to confidence, contributing to persistent concerns about deskilling, resistance to change and the wider impact of AI on the profession. These concerns are consistent with previous research showing that fear of replacement is highest among those with limited AI knowledge, while individuals with more advanced understanding tend to hold more positive attitudes towards new technologies, because they feel better prepared to engage with them (Huisman et al, 2021).

Despite these concerns, the workforce expressed a striking level of optimism about the future. Four in five (81.2 per cent) respondents agreed that AI would support radiographers rather than replace them, and that the profession would evolve alongside these technological advances, as it has done before with previous technological change. This confidence is encouraging for both staff retention and the appeal of radiography as a career to future recruits, signalling a profession that sees AI as an enabler rather than a threat.

When asked whether AI adoption would push radiographers toward a more patient-centred focus or a more technology-driven role, respondents rejected both extremes. This suggests a strong commitment to the dual nature of radiographic scope of practice, valuing both patient care and technical expertise as core to their professional identity. These are areas radiographers are unlikely to relinquish.

At the same time, respondents recognise that their technical skills will need to expand as AI becomes more embedded in practice. This aligns closely with the updated HCPC Standards of Proficiency, which now emphasise digital knowledge and skills as fundamental to safe and effective practice. Importantly, radiographers were clear that key responsibilities, particularly those linked to patient safety, such as radiation protection and ensuring the quality of imaging and treatment, must remain firmly within radiographer control, not delegated to AI systems. This is also aligned with current clinical practice accountability frameworks, which propose human accountability for erroneous use of these technologies.

Core members of the City Research Radiographer Group. (L-R) Dr Emily Skelton (CRRaG Deputy Director), Mrs Gemma Walsh (doctoral researcher), Dr Christina Malaateniou (CRRaG Director and project CI), Mr Nikos Stoggiannos (doctoral researcher) & Mr Ben Potts (doctoral researcher)

Do you expect the study to influence the way radiographers use AI going forward?

Ensuring a sustainable future for the profession will depend on radiographers working in genuine partnership with AI, and that requires clear professional leadership. Establishing dedicated radiographer AI leadership pathways or AI champion roles is becoming increasingly important to oversee the safe, effective and ethical deployment of AI across imaging and therapy services. These roles would provide practical assurance of the trustworthiness, clinical relevance, reliability and fitness for the purpose of the new AI tools in medical imaging and oncology.

Local AI champions would also play a central role in building confidence by coordinating staff education and training, leading research and evidence-based practice. They would also support by: managing post-market surveillance; communicating how AI systems work to colleagues and patients; understanding where limitations lie and how to escalate concerns and to whom; and how to integrate these tools safely into day-to-day practice. This kind of structured support is essential for creating an AI-enabled working environment in which radiographers feel empowered rather than threatened.

Ongoing research from the CRRAG team at City St George’s, University of London is now exploring what these AI leadership roles could look like in practice, and how they might be embedded within radiography services. Their work is helping to shape a future in which radiographers lead, not follow, the digital transformation of their profession.

Access the full R-AI-diographers publication online now.

Find out more about AI in radiography

The SoR has sought feedback on the use of AI in radiography from a range of SoR members and advisory groups, including the SoR informatics group, the College of Radiographers (CoR) patient advisory group, the SoR research advisory group and the CoR Education and Career Framework (ECF) review writing groups.

This feedback has been used to develop recommendations and guidance focused on the different areas of radiographic practice.

Artificial intelligence guidance for clinical imaging and therapeutic radiography workforce professionals can be accessed online here.

Read more

Image captions