Ellen Xu may only be in high school, but she is already contributing to the San Diego region’s prestige as a leader in medicine and health innovation.
On March 14, the 17-year-old Del Norte High School student was the third-place finisher – winning a hefty $150,000 prize – in the Regeneron Science Talent Search, a national program of Society for Science that is the country’s oldest and most prestigious science and math competition for high school seniors.
“Congratulations to Ellen on winning third place in the Regeneron Science Talent Search 2023,” said Maya Ajmera, president and CEO of Society for Science and executive publisher of Science News. “Ellen is an inspiring researcher whose work to develop a new, more reliable way to diagnose Kawasaki disease has the potential to help doctors provide better care to children, and materially improve health outcomes.”
For her award-winning project, Xu developed a convolutional neural network – a subset of machine learning used for image analysis – to help diagnose Kawasaki disease (KD) – the leading cause of acquired heart disease and coronary artery aneurysms in children between the ages of 1 and 5.
KD is diagnosed based on five visual signs that can easily be confused with other diseases. To improve identification, Xu pre-trained her model with images of KD and lookalike diseases she found online as well as images provided by parents of KD children. To generate a larger dataset, she altered the images by adding a variety of random photographic transformations.
Xu’s work indicates that her model can distinguish between children with and without clinical manifestations of KD with 85% specificity using a smartphone photo of the child. Xu authored an article on her research that was published in the peer-reviewed journal Nature Scientific Reports.
Bright Young Minds
Since 1942, Society for Science has been awarding the nation’s bright young minds in its talent search. Each year, nearly 2,000 student entrants submit original research in critically important scientific fields of study and are judged by leading experts in their fields. The search focuses on identifying, inspiring and engaging promising young scientists who are creating the ideas that could solve society’s most urgent challenges.
In 2017, Regeneron became the third sponsor of the Science Talent Search with a 10-year, $100 million commitment that nearly doubled the overall award distribution to $3.1 million annually, increasing the top award to $250,000 and doubling the awards for the top 300 scholars to $2,000 and their schools to $2,000 for each enrolled scholar to inspire more young people to engage in science.
“My experiences participating in the Science Talent Search changed my life, helping convince me to devote my life to trying to use science to cure disease,” said George D. Yancopoulos, M.D., Ph.D., co-founder, president and chief scientific officer of Regeneron, and a 1976 Science Talent Search finalist and top winner. “I can only hope this year’s students will be similarly inspired to become the next generation of scientists, engineers and innovators that will develop and advance solutions for the world’s greatest challenges.”
The March 14 award ceremony brought the top 40 finalists to Washington D.C. to present their projects before a panel of scientific professionals, and to network with the judges and each other.
Xu said that her “favorite part” of the competition was “hanging out with the finalists in the lounge. It was just really fun realizing we have so much in common despite our interests being so different.”
The competition is known for producing projects from a wide range of scientific topics. This year, those topics included everything from cancer research to climate change to the space race and more.
Neel Moudgal from Saline, Michigan won first place and $250,000 for creating a computer model that can rapidly and reliably predict the structure of RNA molecules using easily accessible data to make it easier to diagnose and treat certain diseases. Second place and $175,000 went to Emily Ocasio of Fairfax, Virginia., who used artificial intelligence to determine whether humanizing language was used by the Boston Globe in the years spanning 1976-84 when describing homicide victims. She found that Black victims received less humanizing coverage than white victims.
In addition to her interest in using technology to solve medical and health problems, the multi-talented Xu is also an engineer. Over Thanksgiving break, she designed and built a low-cost, fully functional, nine-foot-wide geodesic dome in her backyard. To prepare for the project, she studied geodesic dome calculations and stereometry and even printed a miniature 3D model.
Xu is also co-editor-in-chief of Del Norte High School’s award-winning literary magazine, founder of a nonprofit that teaches cybersecurity to girls of military families and a nationally and internationally ranked saber fencer.