Courtesy of Abhiram Ruthala
An infographic describing the implications of different waves captured by the Muse headset, included in a video produced by the group to present their project.
Sophomore Ovee Dharwadkar’s interest in neuroscience originated when she attended a summer program on neuroscience in 2024. The camp sparked her interest in the subject, but more importantly, it gave her connections to other motivated peers across the U.S.
After gathering a group of high school students interested in computational neuroscience — a field which involves connecting the brain to a computer — Dharwadkar decided to tackle the 2024 Congressional App Challenge, which tasked competitors with designing an app to solve a real-world problem.
In December, she learned she had received second place in the 2024 Congressional App Challenge for California’s 16th Congressional District.
Her entry, an app called iDiagnosis, uses a Muse headset and an API to get real-time EEG data to diagnose the user for depression. The Muse headset feeds the app real-time information about the neurological activity of the user, which is then filtered to provide a final diagnosis.
The first step to starting the Congressional Challenge was choosing a suitable problem that she wanted to solve. She started brainstorming alongside seniors Tarun Ramakrishnan, Abhiram Ruthala and Kenneth Sarip — the latter two are students from other schools.
“We were hoping to find a problem that wasn’t super big like climate change, where we couldn’t necessarily make a noticeable impact, but rather one that’s closer to us,” Dharwadkar said. “We decided to provide a more easily accessible method of diagnosing depression among teens.”
The next step was developing the software needed to make the program work as intended. One of the largest challenges that Dharwadkar faced during the project was acquiring adequate data to use for her app.
When researchers conduct testing on humans or animals, they must first get government approval of their project. However, according to Dhardwadkar, they didn’t have the resources to get that, so they instead used a clinical data set from professors to train their algorithm and then ran tests on their family members.
To actually make the diagnoses, they analyzed brain waves. For example, a person taking a test may exhibit alpha waves, while a person who is sleeping may exhibit delta waves. According to Dharwadkar, studies have found that patients with depression exhibit less alpha and delta waves than normal.
“I didn’t think that it would work,” Dharwadkar said. “When I finally pressed the run button, I was expecting to have so many bugs. I was getting ready to just go to sleep defeated, but then a miracle happened: There weren’t any significant bugs — [the full program] didn’t quite work yet, but it was really gratifying to see things start to come together.”
As the other members of her team were all seniors, they were often very busy working on their college applications, so Dharwadkar took the lead in designing a majority of the project.
So far, she has completed the backend of the application; she has successfully designed the algorithm, which currently has a 90% accuracy rate. According to Dharwadkar, the challenge involves coming up with an idea for an application and showing the feasibility of the project.
In the future, Dharwadkar hopes to expand her design into an actual app with its own corresponding hardware. She aims to increase both its accuracy and accessibility.
“During this process, I really learned the importance of choosing a problem that actually interests you,” Dharwadkar said. “It makes the process much more fun and it makes it go much smoother.”
