Boston Children’s Hospital has discovered various rare conditions by using AI.
According to an article by OpenAI, ChatGPT for Healthcare was integrated as an enterprise AI layer across the hospital’s clinical and operational infrastructure. The institution, which has nearly 1 million outpatient visits annually, deployed more than 50 automations for operational workflows, benefiting the hospital’s supply chain, billing, and operations teams.
“The problem isn’t effort. It’s human cognitive limits,” John Brownstein, chief innovation officer at Boston Children’s, said in the article.
AI was used to support research in data analysis and cohort building, and administrative teams have found it useful for drafting documents, coding, and other tasks, according to OpenAI. Additionally, it has been used to streamline administrative tasks, including processing invoices and scheduling surgeries.
“If you can use AI tools to help predict when a patient might need a surgery, you can create better optimization of schedules and utilize our ORs in a more efficient way,” Brownstein said in a video shared by OpenAI. “That is leading to more uptake and use of surgical rooms and much more significant revenue back to the hospital. But even more importantly, kids can get the care that they need faster.”
More significantly, AI helped improve diagnostic outcomes, particularly for individuals with rare conditions that had previously gone undiagnosed or unresolved, per the article. Even highly trained and experienced physicians can face challenges when evaluating cases that require synthesizing evidence from the global medical literature alongside incomplete genetic data and clinical histories, it noted. So the hospital developed a “co-pilot geneticist.”
“We combine genetic information, phenotypic information, literature search, and the reasoning of AI to deliver diagnoses to families that were once left without any answers,” Brownstein said in the article. “This was unthinkable before, but is now providing hope to so many families.”
The “co-pilot geneticist” has led to more than 40 rare conditions being diagnosed, he said.
“The incredible partnership that we have with OpenAI has allowed us to go well beyond the expertise that exists in the hospital. We’ve now identified at least 40 families who have now gotten a diagnosis of a rare condition as a result,” Brownstein explained in the video. “This was unthinkable before, and now it’s providing incredible hope to so many families, and this is about discovery. It’s identifying potential genes and new therapies that might treat these conditions. AI is serving as this important co-pilot now to our genetic service that ultimately is gonna be for every service line in the hospital.”
