Skepticism and optimism: Radiologists are still divided on AI integration
Radiologists as a whole are generally ready to implement artificial intelligence applications in their practice on some level, but new survey results suggest that a significant portion are skeptical of how AI will enhance readers’ decision-making processes.
In a survey of over 230 radiologists, only around one-third of respondents who currently utilize AI considered its contributions decisive. This perception did not negate the respondents’ optimism for the potential of AI though—more than 80% agreed that AI will improve multiple aspects of the radiology workflow in the future.
AI applications are at the doorstep of the medical imaging world—hundreds of applications have already received the FDA's stamp of approval—yet adoption of the technology has not kept pace with its advancements so far. That's why understanding how those who will be most affected by AI feel about its implementation is so important, authors explained in the European Journal of Radiology.
“The successful integration of AI tools into clinical workflows hinges upon the active engagement of healthcare professionals, who serve as primary stakeholders interfacing with AI systems daily,” corresponding author Michaela Cellina, from the radiology department at the Fatebenefratelli Sacco Hospital in Milan, Italy, and colleagues noted. “Consequently, understanding the perceptions and attitudes of radiologists towards this transformative technology emerges as a critical imperative.”
The questionnaire touched on respondents’ current satisfaction with using AI-based tools, perceptions of the potential for AI to improve things like detection rates and workflow, and optimism for the future benefits that could come as the technology is fine tuned. It was distributed to a wide range of radiologists with varying demographic details and experience levels.
Radiologists of all experience levels were optimistic about the future of AI, but those younger than 30 and older than 60 expressed more positive perceptions compared to their middle-aged peers. Just under 40% of respondents reported experience with AI in their current roles, but around 30% of them were not convinced that it was benefiting their practice.
More than one-third of respondents reported feeling inadequately prepared to utilize AI, especially the younger radiologists surveyed, 46% of whom said their AI literacy is lacking. Two-thirds viewed AI as an opportunity to improve the field of radiology, rather than a threat to their careers, though 85% said that radiologists should still have the final say on reports. This further reinforces the notion that AI will support not replace radiologists, the authors suggested.
“Despite the generally positive outlook among radiologists, there remains significant work to be done to enhance the integration and widespread use of AI tools in clinical practice,” the group wrote.
One solution the team proposed is improving AI literacy, not just among trainees but across all experience levels. Educational initiatives could start at the trainee level, but professional development training related to AI should continue throughout each career stage, the group suggested.
“These initiatives should impart a comprehensive understanding of AI, covering its potential benefits and associated risks, thereby equipping radiologists to effectively navigate and harness these technologies.”
The study abstract is available here.
In addition to her background in journalism, Hannah also has patient-facing experience in clinical settings, having spent more than 12 years working as a registered rad tech. She joined Innovate Healthcare in 2021 and has since put her unique expertise to use in her editorial role with Health Imaging.