Artificial Intelligence for the Clinician Episode 5: Are Radiologists Out of a Job?

This episode of Behind the Knife explores the transformative potential of artificial intelligence in radiology, focusing on the landmark Maasai randomized controlled trial published in The Lancet. The study evaluated AI as a triage tool in breast cancer screening across 100,000 Swedish women, showing that AI was non-inferior to traditional double-reading by radiologists, with a 44.2% reduction in radiologist workload, higher sensitivity, and fewer interval cancers—particularly aggressive subtypes. The AI functioned independently, flagging high-risk cases without showing markings during reading, preserving radiologist autonomy and allowing them to operate at the top of their license. The hosts emphasize that AI should not replace radiologists but rather redistribute their roles toward higher-value tasks like clinical context integration and complex case interpretation. They contrast this with concerning reports of AI inaccuracies in surgical navigation systems, highlighting regulatory gaps, automation bias, and the need for rigorous training and oversight. The episode concludes with a call for more RCTs before deployment, stronger public trust, and clinician vigilance in adopting AI responsibly.