An online survey was built in Checkbox, version 8.3.0 (Checkbox Technology, Inc., San Francisco). The survey questions were based on previously implemented surveys on AI [7,8], with input from AI experts in medical imaging and radiotherapy. The ASRT research department provided further guidance on survey questions and structure, with support from the ASRT senior executive and leadership team.

The survey consisted of 37 total questions: 7 demographic questions, 26 closed-type questions related to AI, and 4 open-ended questions that allowed respondents to enter free-text responses. The approximate time needed to complete the survey was 15 minutes.

Conditional questioning was employed to automatically modify the order and appearance of survey questions according to the respondents’ demographic characteristics and previous responses.

The survey was sent electronically to all USA-based educators in medical imaging and radiotherapy in the ASRT database, excluding those who had opted out of receiving electronic communication from ASRT. Hence, this survey was electronically sent to 5,066 educators working in all disciplines in both clinical and educational settings. A survey link was distributed to selected participants via personalized email. During the data collection period, email reminders were sent (n=3) to maximise participation. 

All quantitative data were analysed using descriptive and frequency statistics. Graphs and tables were also created to better visualise findings. Regarding the responses received for the open-ended questions of the survey, a conceptual content analysis strategy was employed to analyse qualitative data.

The survey and analysis within this manuscript was conducted by the ASRT as part of this routine internal audit and evaluation practices; hence, this survey did not need approval by an ethics review board. The research team observed all key principles of research governance, like informed consent, anonymity, and confidentiality.