Skip to main content

Student Venture Aims to Detect Depression with App

This article originally appeared on the Tsai CITY website and is republished with permission.

A person’s facial characteristics and the way he or she talks can tell researchers a lot about underlying psychological conditions. And with the widespread use of cellphones to help read these visual and linguistic cues, a group of Yale students at the recent Yale Healthcare Hackathon decided to connect the dots. The result is Bene, an app-under-development that uses artificial intelligence technology to record and analyze users’ responses to a series of trivia-like questions (example: “What are the ingredients in s’mores?”) to draw correlations to specific disorders such as depression. Bene won the challenge posed by the Tsai Center for Innovative Thinking (Tsai CITY) to “use digital technology to improve the patient experience,” winning $1,500 and additional support from CITY.

The 10-member all-Yale team includes four MD/MBA candidates, three MD candidates, two undergraduates, and a PhD student.

“We’re trying to address mental health in low-resource settings,” says Nicholas Chedid, an MD candidate at Yale School of Medicine.

The team created a working prototype in 24 hours using an existing Microsoft cloud service platform that had AI-related modules. The idea, they say, is to get users talking.

“Ideally, we want people to respond in complete sentences,” says Anusha Raja, a joint MD/MBA candidate at Yale School of Management and Yale School of Medicine. “The first question we ask is ‘How is your day?’”

How does the app distinguish between real and fake emotions? In part, says Raja, it’s how fast or slow someone is speaking. “A person with depression will take longer to answer a question,” she says. Plus, Chedid adds, machine learning improves with time—the more interactions it records, the better it can identify changes in expression and tone.

Most of the team members knew each other in advance—and in the high-pressure environment of the hackathon, they benefitted by knowing their individual skillsets and capitalizing on them. Dharshan Anandasivam, another MD/MBA candidate, says, “The atmosphere and the community of the hackathon lends itself to an immersive experience. Everyone is working toward a common goal.”

The team is continuing to develop Bene and they plan to apply to upcoming $25,000 Yale entrepreneurship prizes, such as the Thorne Prize for Social Innovation in Health and Education and the Aetna Foundation Prize for Health Equity Innovation.