At Yale New Haven, docs see AI as pathway to entrepreneurship

Modern Healthcare – Friday, June 6, 2025
By Gabriel Perna

It wasn’t exactly Shark Tank, but a recent competition at Yale New Haven Health in Connecticut showed how artificial intelligence has paved the path for physicians to become entrepreneurs.

As tech companies, health system, insurers and investors pour money into developing AI, doctors are going beyond just using the technology. Some are creating algorithms with their free time to improve clinical care, even potentially looking to turn this interest into a full-time business as they watch the success of physician-led companies such as Abridge and Viz.ai.

“Doctors by nature are conservative because we want to make sure we do no harm, and so you have to subject new ideas to very rigorous evaluation frameworks,” said Dr. Lee Schwamm, chief digital health officer at Yale New Haven, who hosted the competition. “However, I think you’re seeing the vanguard in this group of doctors. You’re seeing people who are excited about applying this new technology to solve difficult and important clinical problems.”
Yale New Haven Health’s AI Championship

Yale New Haven’s AI Championship, which took place on May 27, featured 12 competing teams of clinicians and technologists from Connecticut pitching their clinically focused AI models to a panel of six judges from venture capital firms and health systems. Each team had five minutes to pitch and two minutes to answer questions from the judges.

The winner, a team led by Yale adult and pediatric abdominal transplant surgeon Dr. Ramesh Batra, received an award of $100,000 for its deep learning model that can predict time-to-death following withdrawal of life support in organ donors within intensive care.

As AI becomes more prominent, physicians see it as a way to fix areas of healthcare that they say are broken. The current organ procurement system is riddled with wasted efforts and logistical challenges, Batra said. By predicting when someone in the ICU is going to die after withdrawal of life support, the model aims to help clinicians precisely know which organs will be available and suitable for transplantation from an organ donor and therefore efficiently plan organ transplant to ultimately save someone else’s life, which otherwise is unpredictable and unknown.

Batra, who has been working on the model since 2018 and originally received a Yale Innovation Grant in 2022 for it, said he entered the Yale New Haven competition to get in front of investors and the business community.

“I was very interested in the pitch competition, not from the dollar perspective, but to get exposure to the business community,” Batra said. “I wanted to see if the [business] community and investors were interested in an idea like mine. And I got my answer.”

Tamar Rudnick, head of partnerships at venture capital firm Define Ventures and one of the judges, said Batra’s model won in part because it was unique in the marketplace. It also had a good evidence base and fills a huge clinical need, Rudnick said.

Getting the perspective of AI from clinician-led teams doing on-the-ground work at health systems is what attracted Rudnick to the event.

“Healthcare is not just like AI for the sake of AI, but AI applied to a really important use case,” Rudnick said. “These guys are living and working as clinicians. They feel the pain points.”
Looking at their AI future

The prize money — awarded to three third-place winners, two second-place winners, and sole first-place winner Batra — came from funds allocated by the Connecticut legislature. For this reason, participants were limited to Connecticut-based organizations, including Yale New Haven, Hartford HealthCare, Connecticut Children’s Hospital, UConn Health, and Gaylord HealthCare.

The one-day event also featured prominent AI leaders at vendors, health systems and nonprofit organizations. Dr. Shiv Rao, CEO of AI vendor Abridge and a practicing cardiologist at the University of Pittsburgh Medical Center, was one of the speakers, partly to serve as an aspirational guide for potential entrepreneurs, Schwamm said.

“It’s helpful for people to see you don’t start a company and take it to this level without paying your dues for a long time,” Schwamm said.

Dr. David Chen, a physiatrist at Wallingford, Connecticut-based rehabilitation center Gaylord Specialty Healthcare, said he has no doubt that his team’s prior authorization tool could eventually turn into a full-fledged business enterprise. Chen took part in the Yale Health AI Championship in part because he wanted to network with potential investors and find strategic partners.

Chen embraced the spirit of the event by giving a five-minute pitch in the style of a contestant on the TV show Shark Tank. While his team didn’t win a cash prize, Chen said he was pleased with how the event turned out, especially because he made important connections. He said if the opportunity presents itself, he would love to turn this hobby into a full-time profession.

“This was more like a hobby for many years,” Chen said. “I need to pay back my loans and pay the bills, so I do my patient-care day job. I go home, and then I write computer code, download models, read forums and technical papers, and spend my weekends writing more code and doing demos. I love the technical side of it because it’s fun.”

Not every physician is looking to turn their AI side project into a full-time business. Dr. Steven Zweibel, medical director of innovation at the Heart and Vascular Institute at Hartford HealthCare, developed an AI model that helps diagnose and predict a patient’s risk of heart failure and sudden death.

Zweibel, who collaborated with the Massachusetts Institute of Technology to develop the tool, was one of the third-place winners. He said the model is a passion project and that he saw the competition as a good forum to showcase it. As the model evolves, his goal isn’t to become an AI entrepreneur.

“I’m a little bit older. I don’t have the runway to just quit my job,” Zweibel said. “That’s not in the cards for me.”

Access this article at its original source.