Radiologists, medical students and surgeons all agree that AI should be incorporated into diagnostic radiology, but for the most part their perceptions of the technology are drastically different, according to a new study published in the European Journal of Radiology.
In fact, students see AI as more of a threat to radiology than radiologists themselves, with about a quarter of students surveyed pointing to the technology as the reason they aren’t considering pursuing the profession. The authors believe more education is needed to properly inform students and medical professionals about the coming impact of AI.
“The findings indicate a need to educate students about AI earlier in medical school and to create more of a dialogue among radiologists on the topic,” Jasper van Hoek, with University of Bern’s hospital in Switzerland, and colleagues wrote, Nov. 9 in the study. “Such efforts could lead to better education of the students about AI and thus attune them to a more positive view of this emerging technology within the field of radiology.”
Much has been said about AI’s potential impact on radiology. There were plenty of scholars, such as neural network expert Geoffrey Hinton, who warned of its looming threat to replace the physical radiologist; but for the most part, literature on the topic has sought to ease those fears, focusing on its ability to augment, rather than replace the profession.
Hoek and colleagues sent an online questionnaire to 170 people (59 radiologists, 56 surgeons and 55 students), asking various questions related to the future of radiology, career choices and work environment in relation to AI.
Overall, all three groups were supportive of using AI in radiology. Radiologists were more supportive of using AI, but they did fear that other specialties may take over reading some of their exams, referred to as “turf losses.” Medical students and surgeons (whose work relies heavily on radiology) were more skeptical of AI.
But overall, those surveyed seemed “uncertain” about the future of the profession, the authors wrote. The small data sample coupled with the study’s focus solely on Switzerland limit the generalizability of the findings.
“Following the assessment of most radiological publications—in our review of them—AI will not be a threat but rather a welcome addition to the radiological workflow,” the authors concluded.
“One must say that the results from our study might be worrisome. Students, and especially the best students, might not choose to go into radiology. Thus, better education about this supposed fear seems to be necessary.”