When do we add AI to radiology training programs?

When it comes to teaching new dogs new tricks, radiology training programs need to be thinking about updating their curricula and preparing for both the short- and the long-term effects of AI and machine learning, according to “Toward Augmented Radiologists,” a new commentary published online in March in Academic Radiology.

In the short term, Authors Shahein H. Tajmir, MD, and Tarik K. Alkasab, MD, PhD, of Massachusetts General Hospital and Harvard Medical School, see AI and ML being used to “predictate and preanalyze” examinati ons, taking on tedious tasks such as measuring nodules. Training programs will need to prepare trainees by teaching them how to work with these tools, taking advantage of what they can do and how to actively supervise AI tools.

Trainees need to know how to recognize when technologies aren’t working correctly, too. As Tajmir and Alkasab suggest: “Perhaps standardized teaching cases where the AI tools perform well and poorly will be developed to assess a resident’s ability to know when to leverage an AI tool’s output and when to discard them because of failure.”

The future face of imaging could differ a lot from the present, the authors note. For instance, in the future radiology trainees and attending radiologists could have their work and interactions monitored by computers looking for opportunities to correct reading errors and improve efficiency. AI could be used to select which cases trainees should read, maybe even creating its own cases to test certain aspects of the trainees’ abilities. AI also could alert educators when a trainee needs help. But beware of going too far, the authors urge, allowing appropriate watchfulness for improvements “without disrupting the development of trainees.”