Photo Credit: Society for Science

Ellen Xu, a 17-year-old science prodigy, made history when she designed the first diagnostic test for Kawasaki disease using artificial intelligence (AI). The rare disease struck her sister years ago and pushed Ellen to create a better solution. At the most prestigious young adult science competition in the nation, Ellen’s project earned her third place and a grand prize of $150,000.Kawasaki disease has no existing test method that could diagnose it accurately without relying on physicians’ years of training or research. It is characterized by fever-like symptoms which can be mistaken for other conditions if not diagnosed on time. If left untreated, children may develop long-term heart complications just like what happened to Ellen’s sister before she was correctly diagnosed.

That’s why Ellen decided to use deep learning to create an accurate diagnostic test with only a smartphone image as input. She used a convolutional neural network – an algorithm that mimics how our eyes work – and programmed it to analyze images for potential Kawasaki disease symptoms. However, this type of algorithm needs lots of data references in order to process images quickly and effectively; so Ellen crowdsourced them from medical databases all around the world.

In the end, her efforts paid off: Ellen achieved 85% specificity in identifying between Kawasaki and non-Kawasaki symptoms with just a smartphone image. This success earned her third place at Regeneron Science Talent Search Medicine & Health Project which came with a hefty reward of $150,000!

It is truly inspiring that such an innovative solution was created by someone as young as 17 years old! Congratulations again to Ellen Xu for making history with her revolutionary AI diagnostic test for Kawasaki disease!

Sources: Good News Network