Scientists have developed an algorithm that is just as good at identifying skin cancer as any dermatologist in London. Although the program isn’t quite ready for your smartphone yet, it has the potential to change the way we check moles and diagnose skin cancer.
Learning to Identify Skin Cancer
The team at Stanford University developed a neural network that learned how to recognise skin cancer. The algorithm was able to tell the difference between photographs of cancerous growths and non-cancerous moles, warts and skin lesions. When it was tested against a group of experienced dermatologists, it was just as accurate as the doctors. In some cases, the algorithm was actually better at identifying the cancerous growths than the dermatologists.
The Future of Skin Cancer Diagnosis?
The researchers suggest that this algorithm could one day be developed into a smartphone app that could identify cancerous moles using the phone’s camera. You would then be able to use the app to check your moles regularly or to put your mind at rest if you are worried about any unusual moles or changes. Your dermatologist in London might even use an app like this to confirm their own diagnosis when they examine your skin.
However, while this algorithm has the potential to change the way that skin cancer is identified in the future, it isn’t quite there yet. The researchers are still working on it and there isn’t yet a reliable app that can check your moles for you. You still need the expert eye of your dermatologist in London to know whether a suspicious mole is likely to be cancerous. One day, you may be able to ask your phone whether you should get your mole checked out, but for now you should continue to rely on your doctor.