This article was taken from the March 2015 issue of WIRED magazine. Be the first to read WIRED's articles in print before they're posted online, and get your hands on loads of additional content by subscribing online.
When it comes to science fair projects, most school kids just stick to papier mâché volcanoes. At 17, Florida student Brittany Wenger created an artificial brain for detecting cancer. Wenger's
program, which enables accurate recognition of complex patterns by modelling the parallel processing capabilities of biological neural networks, was the winning entry for the 2012 Google Science Fair.
Thousands of entrants from around the world take part in the annual competition for a top prize that includes a $50,000 (£32,000) scholarship and a much-coveted internship at Cern, Google or LEGO.
When Wenger applied this pattern- recognising power to a public database of non-invasive samples of breast mass tissue, her network was able to detect malignancy with a remarkable 99.11 per cent sensitivity across 7.6 million trials -- a better rate than the best commercially available method for cancer diagnosis.
Wenger is now studying computer science and biology at Duke University in North Carolina with the aim of becoming a paediatric oncologist. She continues to test her programme at one of Duke's cancer research labs and has opened it up online to the public, with several hospitals trialling the system. As neural networks continually adjust and improve their accuracy in response to receiving new information, this means that Wenger's brain will continue its learning right alongside her.
This article was originally published by WIRED UK
