3 Princeton Students Awarded In National Science Competition, 1 Named Finalist

PRINCETON, NJ — Earlier this month three Princeton students were included in the top 300 of the Regeneron Science Talent Search.

The program is touted as the nation’s oldest and most prestigious science and math competition for high school seniors.

Benjamin Gitai, Amy Lin, and Yurai Gutierrez Morales will be awarded $2,000 and the school district will get $2,000 for each student. But more good news was in store of the district as on Friday Gutierrez Morales was announced as a top 40 finalist. The student now has a shot at winning the top prize of $250,000.

Now in its 84th year, the competition identifies extraordinary young minds who blend scientific talent, curiosity and bold leadership to drive meaningful change for society. Regeneron Science Talent Search alumni have gone on to win 13 Nobel Prizes, and 23 MacArthur Fellowships, and have founded numerous world-changing companies, including Regeneron.

More than $3 million in awards will be distributed throughout the Regeneron Science Talent Search in total. Each finalist will be awarded at least $25,000, with the top 10 awards ranging from $40,000 to $250,000.

Although Gitai and Lin did not make it to the top 40, they plan to continue their research work.

Science Supervisor, Jacqueline Katz, said she was extremely proud of the students and their achievements.

“This is a new milestone for PHS. We're all super, super proud. These are the scientists that are going to help us out in the years to come,” she said.

Here are more details about the students and their incredible projects.

Yurai Gutierrez Morales

Gutierrez Morales’ project involves using a new technique to discover bacteria that might have helped the world's only known vegetarian spider change its diet to eat plant material. “I think that's a big discovery and there is more about these creatures that we still do not understand,” the student told Patch. "I’m very, very excited because I have a lot more to investigate.”

The student spent around two months on research which she conducted at a laboratory in Mexico. She is now observing the spiders at a lab in Princeton and continuing her research.

Born in the United State, Gutierrez Morales moved to Guatemala with her mother when she was seven-years-old. Then at age 14, she moved to Mexico where she had to start working to support her family and education took a backseat.

“I didn’t have a stable family or resources, so I had to work a lot in Mexico,” Gutierrez Morales said. “When I was working, I saw a lot of students and that made me really sad, because I wanted to study again.”

At 17, she moved back to the United States and enrolled in Princeton School District. At age 20, Gutierrez Morales is the oldest student in her class. But everything else pales when it comes to getting the education she so desired.

“I keep telling my teachers I’m too old to be here. But I'm very grateful to be studying again. I’ve always been interested in animals and I’m very excited to start research about them,” Gutierrez Morales said.

She credits her teacher Mark Eastburn from Princeton School District’s Research Program for encouraging her and pushing her to continue her research.

Amy Lin

Lin developed a machine learning-based computational model, potentially replacing expensive and dangerous experiments, to predict chemical materials' melting temperatures, an important property in material discovery for clean energy applications.

“I didn't want to just build these sort of AI algorithms that could predict this important material property, but I also wanted to find out which other material properties could contribute to best determining the melting temperature,” Lin said.

“For example, for compounds containing metals, I found that a lot of the electron-containing properties, or like, electron-related properties, were really important to predicting the melting temperature.”

Lin took a few hours per week to work on her project from September until around March last year.

The teen got interested in machine learning and AI early on when she started a research project to try to understand and predict how her dog would behave in response to her environment.

She designed machine learning in such a way that the model was able to extract very specific information about her dog. Lin found that her pet liked the colors blue and purple, and she liked low-frequency sounds. “That makes sense because she sings when I play the low bass notes on the piano. And she also likes to sniff out this purple yard sign. So that was really cool,” Lin explained. “That taught me how machine learning could be used in our everyday lives and how it can be used to make scientific discoveries. So I think that was really inspiring to me.”

Although she wasn’t named in the top 40 in the competition, Lin wants to continue her research in the future as she feels it could be beneficial.

“Scientifically what I hope to do in the future is include more material properties that we can use to predict the mountain temperature. A really popular area right now is obviously, like quantum materials and quantum mechanics. So, I was wondering, ‘can we use quantum properties to predict the mountain temperature?’ Because those are also especially important properties,” Lin said.

"In the future, I hope to be able to turn this into a really efficient and useful tool for material science, and also include other functionalities, like predicting other material properties, and also contributing a bit to materials discovery.”

Benjamin Gitai

Gitai’s project resulted in the creation of a computer vision AI model for the measurement and prediction of ankle replacement surgery.

His project tries to create pre-op measurements and automate those using AI. Additionally, the project also tries to predict if ankle replacements that are existing, were going to fail later. The aim is to try to prevent people from getting injured.

But working on the project didn’t come easy for Gitai. He had to figure out how to actually get measurements, because as of now, they're all being done by hand by doctors, which can lead to slight errors, but also be inefficient for time.

“We ran through issues very quickly where we realized that the only way you can really create an outline of a bone from an image is using AI. Because bones are porous, so they have lots of these little jagged edges and they have holes in them, which means the standard software can't really recognize the shape of a bone. Instead, it just kind of creates random little circles all throughout that general area,” Gitai explained.

However, the student was able to create a pipeline with just a couple of images, and in a few minutes, he was able to train a whole system that could recognize and outline any bone he wanted.

“With just a couple of images and a little bit of time and effort, you can make it do pretty much anything you want with any bone you want to. I made the whole thing open source.”

Gitai spent around 30 hours a week for about seven weeks on the project last summer. "There's more to be done in the future,” he said.

"I don't actually view my project as being quite done yet. Originally I was trying to get a bunch of patient data, but due to HIPAA, the only data I was able to work with for this part of the project was publicly available patient data. My hope is that in the coming months, I'll be able to get back to work on this. I think I have an idea of how I can get access to some patient data, and then we can start looking at trying to make those predictions about the actual failure of it.”

Apart from pursuing his project, Gitai will be running track at Princeton University next year, which he is very excited about.

Have a correction or news tip? Email sarah.salvadore@patch.com