> The AI revolution in medicine is here.

There were limited scope CADe results showing improvements over average readers 20 years ago, and people calling it a 'revolution' then. I'm not sure anything has really shifted; the real problems in making clinical impact remain hard.