IE 11 is not supported. For an optimal experience visit our site on another browser.

After cousin's death, teen creates program to reduce surgical errors

The program, which identifies when a doctor is making a mistake, helped the 17 year old win a Regeneron Science Talent Search prize.
Growing up, Pravalika Gayatri Putalapattu visited the Regeneron Science Talent Search as part of their public day so she could see the various projects students submitted to the contest and collect their profile cards.
Growing up, Pravalika Gayatri Putalapattu visited the Regeneron Science Talent Search as part of their public day so she could see the various projects students submitted to the contest and collect their profile cards.Courtesy the Society for Science

In 2016, Pravalika Gayatri Putalapattu’s cousin, Sirisha, 13, underwent brain surgery to have a cancerous tumor removed. During that procedure in India, one of the doctors made a mistake and Sirisha ended up dying. Grappling with that loss felt tough for Pravalika, but she turned her tragedy into something that might help others.

“This error was preventable, which means that there exists sort of an alternative universe where she didn’t have to die during the surgery. Dealing with that loss was pretty difficult,” the 17-year-old high school senior from Centerville, Virginia, told TODAY Parents. “I wanted to make a program that could help other families from having to deal with this type of loss.”

She started researching creating a system that could alert laparoscopic surgeons in real time if they were poised to make an error. Her resulting project garnered her a seventh place award and a $70,000 prize in the Regeneron Science Talent Search 2022. She used machine learning to train a program to work on a DaVinci machine, the “standard” laparoscopic surgical machines. The program she created works now for laparoscopic gallbladder removal surgery. She selected this type of surgery because there were many videos annotated by surgeons that her machine-learning program could study.  

“I needed to look at existing surgery video that are annotated by frame, which tool is in each frame and what action is being performed in that frame,” she said. “That’s how (the program) learns — by making those associations.”

But Pravalika hopes that it will eventually work for other laparoscopic procedures.  

When Pravalika Gayatri Putalapattu won seventh place in Regeneron Science Talent Search she realized she would have a profile card like those she collected from past competitions.
When Pravalika Gayatri Putalapattu won seventh place in Regeneron Science Talent Search she realized she would have a profile card like those she collected from past competitions.Courtesy the Society for Science

“My program is able to identify surgical tools in each frame and the way that they’re moving,” she explained. “So hopefully I can work on other procedures that are performed by the Da Vinci machine.”She’s received some feedback on her program, which fuels her continuing improvements to it.

“Overall the general consensus is that this is a useful idea in theory but implementing in practice might be a little tricky because how do you let the surgeon know they’re performing an error without further distracting them?” Pravalika said.

She considered an audio message, such as a beep, but learned that some doctors consider sound too distracting. Now, she thinks she might use a message where the surgeon looks during surgery.  

“The surgeon is watching a video screen and controlling the tools with a remote controller,” she said. “I was thinking about putting a banner on the top of the screen whenever the surgeon makes a mistake. So somewhere where it’s really close to where they’re already looking but not obstructing anything important.”

Navigating how surgeons interact with the program has been one of the more difficult aspects of her project.

“One of the biggest challenges for actually getting this implemented is figuring out how exactly to work with the human, surgeon, how to set up that human machine interface,” Pravalika said.

Pravalika Gayatri Putalapattu participates in Olympiad math, reads, draws, hangs out with her friends and explores different restaurants and parks in her area.
Pravalika Gayatri Putalapattu participates in Olympiad math, reads, draws, hangs out with her friends and explores different restaurants and parks in her area.Courtesy the Society for Science

She hopes to continue work on her program and improve its accuracy. It’s 93.5% accurate right now but she believes it should be better before it’s in use. She’s still not sure where she’s going to college but wants to major in computer science and physics to pursue quantum computing in graduate school. Winning a prize as part of the Regeneron Science Talent Search felt “surreal” for her.

“It’s not just a science fair. You send in a research paper and details about your project and you get judged on your project. But they also ask you general questions; they test your scientific knowledge,” she said. “I was not expecting to be a finalist at all so it was really a cool moment. And my parents were super proud of me.”