Artificial intelligence in radiology: Friend, not foe, say experts concerned about student perceptions of AI
There have been countless conversations centered on the role of artificial intelligence as it pertains to the field of radiology, with arguments both supporting and decrying its use. But when these conversations trickle down to medical students, the tone in which they are presented could have a ripple effect for years to come.
This is exactly what experts sought to address in a new paper published in Current Problems in Diagnostic Imaging. Corresponding author of the paper Navid Faraji, MD, of the Division of Musculoskeletal Radiology at University Hospitals Cleveland Medical Center, and colleagues cited “unfounded concerns” among medical students who have indicated that they would not pursue a career in radiology because they feel the use of artificial intelligence will eventually render the profession obsolete.
In an effort to perhaps dissuade skepticism among students who are on the fence about the future of radiologists, the experts offered a detailed overview of the use of AI in the field, including the benefits, pitfalls and considerations that should be made relating to the technology’s future role in imaging.
“The reality is that it is much more likely that AI will be heavily integrated into a radiologist's clinical practice to focus on improving quality and efficiency rather than replacing the radiologist as a whole,” the authors wrote.
Benefits of AI
They started with the good—AI's ability to offer better workflow optimization, structured reporting, streamlined administrative tasks and the fact that AI systems can analyze data beyond what is visible to the human eye, making it a valuable diagnostic tool in healthcare.
None of this happens without radiologists though, the authors explained, as they play an integral role in the development, training, testing and validation of AI applications.
“This requires radiologists to have a firm understanding of the application of the tool and if there is a true clinical need, evaluate these tools as it relates to the scope of their practice, and use their expertise to guard against over-reliance of technology.”
Radiologists can work with developers to share their expertise on clinical needs, curate appropriate medical imaging data or assist in supervised learning by completing various tasks such as image annotation to establish ground truth for the purposes of training an algorithm. All of these things combined give radiologists an opportunity to “have a major impact” on the integration of AI into clinical practice, the authors suggested.
AI Considerations
The authors also touched on the pitfalls of AI—legal ramifications to consider, HIPAA compliance with data curation and sharing, application accuracy and performance and whether AI can achieve generalizability in real-world, clinical practice.
A Complimentary Tool to An Attractive Specialty
Several applications have already made their way into workflows, but we have only scratched the surface regarding AI’s potential, the authors suggested, adding that AI will “ultimately support radiologists.”
“AI will make radiology more impactful in healthcare, and thus a more attractive specialty for potential trainees.”
The paper can be viewed 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.