Erin Smith is hard at work on an app to help detect the early signs of Parkinson’s disease. Already, she’s traveled to the headquarters of Google and Twitter and won $10,000 to continue her research.
She’s also a 17-year-old junior at Shawnee Mission West High School.
Smith, who lives in Lenexa, is developing an app called FacePrint, which measures how muscles in the face move, particularly in patients with Parkinson’s disease. Her work is already getting her national attention, especially from the Michael J. Fox Foundation.
“Erin is innately curious and so curious that obstacles don’t get in her way. She goes over them, under them, around them so that she can find the answer,” said Brenda Bott, coordinator of the Shawnee Mission School District’s biotechnology program.
Never miss a local story.
Recently, she took part in the BuiltByGirls Future Founder challenge in San Francisco, where girls presented their technology projects and business plans to a panel of other teen girls and Silicon Valley entrepreneurs at Twitter headquarters. Her FacePrint app and presentation were good enough to garner a $10,000 prize.
Smith already has a provisional patent for her work, and she’ll use her prize to pay for costs associated with developing it further, as well as getting a full patent on her work.
She got the idea for her app from a TV show she watched with her family years ago called “Lie to Me.”
“There’s this detective … and he can determine if people are lying based off of a distinctive facial action coding system,” Smith said. “Then about a year ago … I was watching a video by the Michael J. Fox Foundation, and I noticed that whenever Michael J. Fox or another Parkinson’s disease subject would laugh or smile, it came off as really distant or unfelt, (and I thought), ‘What if the facial action coding system could have healthcare implications?’”
She talked with local people who care for Parkinson’s patients and found they had noticed the same effect. Delving into it further, she read research saying that Parkinson’s affects the part of the brain responsible for spontaneous facial expressions.
“I established the scientific validity, and then I started working with people in the community, and that human element pushed me forward,” she said.
This fall, she launched a study with the Michael J. Fox Foundation where Parkinson’s patients watch specific Super Bowl commercials and try to replicate emoji emotions while being filmed by webcams to capture their facial movements.
Smith runs that footage through a special algorithm from a company called Affectiva that works with facial recognition software to generate her data. Partnering with the foundation gets her data to use from all over the world.
“The idea now is to, rather than reinvent the wheel, write a coding program to mine that data,” Bott said. “Her idea is very innovative, but it’s typical of what her generation is doing.”
Smith hopes to take the data she’s collecting and pinpoint how different muscles interact to see where the difference in facial movements lie between people with and without Parkinson’s. The app isn’t ready to go on the market just yet, but Smith hopes it will be by the end of this school year.
“There’s almost continuously something I have to fix or overcome,” she said. “You’re running into problems almost daily, so it’s just kind of changing your mindset from, ‘It’s going to be perfect,’ to, ‘I’m going to do this repeatedly… and the short-term details are going to be flexible and change.’”
BuiltByGirls and the Michael J. Fox Foundation aren’t the only ones who’ve noticed Smith’s talents. She’s gotten recognition from the National Center for Women and Technology, the International BioGENEius Challenge, Google and Seventeen Magazine.
For the future, she’s considering partnering with healthcare non-profit organizations to get her app distributed worldwide, especially to developing countries, to improve early detection of Parkinson’s to get people started on therapies earlier.
Smith said she also wants to see if there are applications for her app for other medical conditions as well.
“Everyone I’ve spoken to is really excited about the possibilities, because current diagnosis methods are pretty—like non-technically speaking—they’re pretty bad. And I think the Parkinson’s patients are really excited to see such a non-invasive, inexpensive way of diagnosis,” Smith said.