E1441. Impact of Artificial Intelligence on United States Medical Students’ Choice of Radiology
  1. Kristen Reeder; Frank H. Netter MD School of Medicine at Quinnipiac University
  2. Hwan Lee; University of Pennsylvania Perelman School of Medicine Department of Radiology
Emerging use of artificial intelligence (AI) in radiology has been accompanied by concerns about job replacements. We aimed to examine United States (US) medical students’ opinions on the use of AI in radiology, factors influencing these opinions, and the impact on application to radiology residency.

Materials and Methods:
463 medical students across 32 US medical schools participated in an anonymous online survey to measure the following: impact of AI on choice of radiology, opinion on role of AI in radiology, prior exposure to radiology, and prior exposure to AI.

AI significantly lowered students’ preference for ranking radiology (P<0.001). One-sixth of students who would have chosen radiology as the first choice did not do so because of AI, and one-third of those who chose radiology as the first choice were still concerned about AI. Among the 188 students who considered radiology as one of their top three specialty choices, ranking radiology lower due to AI was associated with greater subjective concerns about AI (P<0.001), less perceived understanding of radiology (P=0.016), and predicting a decrease in job opportunities due to AI (P<0.001), but not with preference for diagnostic vs. interventional radiology (P=0.996). The level of concern towards considering radiology as a specialty due to AI was significantly increased by discussion with other medical students (P=0.003). Education on AI during clinical radiology rotations was considered the most beneficial way to learn about AI, regardless of students’ preference for radiology.

Educational means to address medical students’ concern about AI will help attract the future radiology workforce to meet the increasing demand for medical imaging.