If there was something that could change the world, it would be a cancer-detecting app on your smartphone - all powered with AI. We know that artificial intelligence is something that will be a part of our future, but now a group of researchers from Stanford are coming out with some exciting things.
The researchers published their findings in the latest issue of Nature this week, training a neural network with 129,450 photos that displayed to the system over 2000 different types of skin conditions. Mixing one of Google's own image recognition algorithms, the researchers trained the neural network to identify both malignant, and benign skin lesions.
The team also worked with 21 human dermatologists, showing the experts the same, common and deadly forms of skin cancer, and asked what treatment they would recommend. Comparing their answer to the AI, the humans performed at the same level. Sebastian Thrun from Stanford's AI Lab wrote in a blog: "We realized it was feasible, not just to do something well, but as well as a human dermatologist. That's when our thinking changed".
One of the most interesting parts of this research, is that the team believes there would be a future where we could have an app on our phones with the AI built in. They said "it will be relatively easy to transition the algorithm to mobile devices", which could lead us onto a path of in-home testing of diseases like cancer.
Susan Swetter, Professor of Dermatology, along with one of the paper's co-authors, said: "rigorous prospective validation of the algorithm is necessary before it can be implemented in clinical practice, by practitioners and patients alike". Andrew Esteva, another co-author, wrote: "Everyone will have a supercomputer in their pockets with a number of sensors in it, including a camera. What if we could use it to visually screen for skin cancer? Or other ailments?"