Eileen Cheever, PA-C, reviews some highlights from her lecture " AI in Dermatology" on Thursday, November 7, 2024, at the 5th Annual Elevate-Derm West Conference at the Westin Kierland Resort in Scottsdale, Arizona.

In this video

In conversation with Jennifer Connor in Elevate Studios, Cheever explains that artificial intelligence began as slow pattern recognition trained on "ambiguous data sets," and has since evolved into machines that learn from information fed to them as problems with solutions. Generative AI, she notes, is already woven into everyday life — the suggested words that appear as you type a text or email. As for whether AI will replace clinicians, her "tongue-in-cheek" answer is no: a robot won't replace you in clinic, but "somebody that is using AI to make themselves more efficient" might.

She walks through how AI is being used in dermatology today, mostly in research and laboratory settings rather than real-world clinics — helping identify suspicious pigmented lesions and flag features worth biopsying, and, in what she calls the most interesting study in her deck, translating biopsy reports into "remarkably accurate, patient-friendly language" that was more likely to help than harm patients. She also names the field's real challenges: a lack of standardization in a picture-heavy specialty, underrepresentation of skin of color in data sets built largely on white skin, ongoing accuracy concerns, and privacy.

  • AI won't replace clinicians, but clinicians who use AI to work more efficiently may replace those who don't — so embrace the technology while knowing its limitations.
  • Most current dermatology applications are still in laboratory and research settings; real-world clinic studies will give the most accurate picture of how AI can actually help.
  • AI shows promise in flagging suspicious pigmented lesions and in translating biopsy reports into patient-friendly language with remarkable accuracy.
  • Key limitations include lack of image standardization, underrepresentation of skin of color in data sets, accuracy concerns, and patient privacy.
  • Stay current by attending conferences and webinars covering AI, and watch for AI integrated into journal subscriptions and newsletters.