Many radiology trainees in the U.K. are interested in working on projects, activities, and teaching related to artificial intelligence (AI), according to survey results published on 19 January in Clinical Radiology.
Researchers led by Dr. Obaid-Ul Hashmi from the East of England Imaging Academy in Norwich also found that almost all respondents felt that AI should be taught during their training. However, only one respondent reported that their training program had implemented instruction in AI.
"Despite the current limited availability of AI-based activities and teaching within U.K. training programs, U.K. trainees' attitudes towards AI are mostly positive with many showing interest," Hashmi and colleagues wrote.
AI's use in radiology continues to grow. In the U.K., the Royal College of Radiologists (RCR) introduced AI in its 2021 curriculum. However, debate continues on how best to implement AI into practices and teaching curricula, as well as how AI could affect radiology careers.
Previous reports from around the world suggest that trainees and professionals alike are optimistic about AI in radiology. In their study, Hashmi and fellow researchers wanted to survey U.K. radiology trainees to better understand their attitudes toward AI, including the level of demand for education, learning needs, previous experiences, and career impact.
The team collected data in a two-month period in 2021 from a total of 149 trainee radiologists, with "at least one response" from all U.K. training programs. The researchers found that 83.7% of trainees reported being interested in using AI. However, 71.4% also reported having no prior experience working with AI, while 79.9% said they would like to be involved in such projects.
The researchers also found that 98.7% of respondents felt that AI should be taught during their training, with over three-fourths saying it should be taught between specialty training years one through three. However, just one respondent said that their respective training program had AI teaching in place. The top three priority areas that trainees said they want in AI teaching included basic understanding, implementation, and critical appraisal of software.
Additionally, 74.2% of trainees indicated that AI would improve the jobs of diagnostic radiologists over the next 20 years. However, the top concerns they reported include issues related to the implementation of information technology, ethics, and regulations. Less than half of trainees reported being concerned with AI replacing radiologists' jobs.
Still, the researchers highlighted that the survey showed "strong" support for further formal training, support, and involvement with AI. They also wrote that trainees could be a "potentially large, untapped resource" for researchers now involved in AI research and development.
"Positive attitudes could improve clinical adoption of AI, underpinning the importance of integrating AI teaching into the radiology curriculum. This could also be extended to medical students," they added.