Abstracts

RETURN TO ABSTRACT LISTING


E2362. Impact of Artificial Intelligence in Choosing a Career in Radiology: A Multisite Student Survey
Authors
  1. Hena Cheema; Hospital of the University of Pennsylvania
Objective:
To assess medical student attitudes towards artificial intelligence and in choosing careers in diagnostic and interventional radiology.

Materials and Methods:
Medical students at 4 major hospital training sites were offered voluntary participation in a web-based questionnaire consisting of 15 questions with Likert sub-elements that aimed to evaluate students’ perception of AI in radiology and impact in choosing a career in radiology. All participation and responses were anonymous. No personal or identifying information was collected beyond 4 general demographic questions about level of training, gender, race, and age range.

Results:
A total of 112 (63 female, 49 male) responded, including 84 medical students, 5 dual-degree (ie MD/PhD or other), and 6 undergraduates; 47% were Caucasian, 33% Asian, 9% Hispanic, 4% African American, and 6% Other. Within the past 6 months, 93% reported exposure to AI (80% read a general article, 73% via watching the news, and 73% during discussion in school). Only 9% of students reported following the field of AI closely, with 10% weekly, and 21% monthly. When asked if “AI will replace diagnostic radiologists” 47% of all respondents agreed; among female respondents, 84% agreed. Thirty-three percent of respondents reported “feeling confident about job prospects in radiology with AI.” In considering radiology as a potential career, 43% of students reported concern about AI, and in this setting, 19% of all respondents could see themselves pursuing a career in diagnostic radiology; 21% in interventional radiology. Seventy-five percent of respondents said they would prefer a mentor who is knowledgeable about AI; 86% of students were interested in learning more about the field of AI; 77% agreeing AI should be integrated into radiology residency, and 69% of students would be more interested in applying to a program with AI in the curriculum.

Conclusion:
We highlight a unique opportunity for improved radiology and medical education about AI. Medical students across demographic groups have suggested concern about the role of AI in radiology and its impact on future job availability in radiology. We show that despite these concerns, medical students are open to and looking for a training that specifically incorporates AI into the curriculum.