Deep learning algorithm diagnoses skin cancer as well as seasoned dermatologists

Deep learning algorithm diagnoses skin cancer as well as seasoned dermatologists

There's been an encouraging development with Artificial Intelligence detection of skin cancers. The Google neural net is 'helping catch skin cancer now, thanks to some scientists at Stanford who trained it up and then loosed it on a huge set of high-quality diagnostic images. During recent tests, the algorithm performed just as well as almost two dozen veteran dermatologists in deciding whether a lesion needed further medical attention.


“Advances in computer-aided classification of benign versus malignant skin lesions could greatly assist dermatologists in improved diagnosis for challenging lesions and provide better management options for patients,” said coauthor Susan Swetter, professor of dermatology at Stanford. “However, rigorous prospective validation of the algorithm is necessary before it can be implemented in clinical practice, by practitioners and patients alike.”


8 Replies

  • Somehow I think 'veteran dermatologists' will use this more as a confirmatory/second opinion tool than a stand alone opinion. It's exciting technology however but I'm not yet confident enough to totally trust the decision to a machine (however advanced).


  • Indeed - but for inexperienced dermatologists, it may prove to be a very useful filter/teaching tool. It will in any case provide a 'second opinion', once it is fully available.

  • As a person who goes to the dermatologist every 6 months and occasionally has had biopsies that turned out benign, this is exciting. In the big picture, you would not need a dermatologist most of the time. I suspect that revenue would drop with the number of biopsies decreasing. You could get a nurse to take the body scans. I would think that this would make quality skin cancer checkups readily available for all and cheap.

    I was lamenting in an earlier post this week that radiologists just provide a laundry list of findings without putting them in perspective - what is of concern and what is not? If AI can be developed to crank out the laundry list, a lot of radiologists would be out of business.

  • This is very exciting news - science fiction comes to life. Maybe this machine will have a better bed side manner than some physicians.

    Sandy Beaches

  • LOL

  • Thanks! That made me laugh. After reading some of the experiences people on this forum have had with their doctors, I realize how lucky I am with the CLL doctor that I have. But I have also met some seriously bad ones!

  • I'm certainly in favour of any reliable technology that aids and improves in the detection of skin cancer. I do think however that having had a melanoma, should the machine say 'No', I'd be asking,...yes, but what does the Dermatologist say?'


  • Anton van den Hengel, Professor of Computer Science, University of Adelaide, reviews the importance of this breakthrough:

You may also like...