Thanks to new developments in artificial intelligence technology scientists say that it’s now possible to identify certain types of skin cancer in photographs with just about the same accuracy as specially trained cancer doctors. A team of cancer researchers at Stanford University reported that these findings were “incredibly exciting” and would start to be tested in clinics. If everything pans out, they believe that artificial intelligence has the power to revolutionize healthcare as we know it by giving people the ability to turn their smartphones into a way to scan for skin cancer anywhere!
The artificial intelligence software was originally developed by Google and was able to differentiate between photos of cats and dogs. The software was then repurposed and was shown 129,450 photos of different skin conditions and was told what each one was. It’s this way that the software learned to spot the characteristics of the most common types of skin cancer, carcinoma, and melanoma which is the most deadly of the two.
In an experiment to test the AI, it was put up against 21 skin cancer specialists and doctors, and the findings were very impressive. Dr. Andre Esteva who was one of the researchers on the team said, “We find, in general, that we are on par with board-certified dermatologists.” However, because the computer software isn’t human, it cannot make a complete diagnosis, as diagnoses are confirmed by taking a biopsy of the skin tissue. Dr. Esteva also said the cancer detection system now needs to be tested amongst doctors in a clinic.
One of the most exciting parts of the research is the potential for the algorithm to be applied to mobile devices so people can scan for cancer anywhere. In order to achieve this, however, the app would have to be built and put through more rounds of testing for its accuracy on various mobile devices. But don’t worry, cancer research isn’t going totally tech yet. Regarding the use of AI in detecting skin cancer Dr. Jana Witt, from the UK charity Cancer Research said, “Using artificial intelligence to help diagnose skin cancer is very interesting, as it could support assessments by GPs and dermatologists. It’s unlikely that AI will replace all of the other information your clinician would consider when making a diagnosis, but AI could help guide GP referrals to specialists in the future.”