AI Breakthrough Provides Answers for Patients With Rare Illnesses

WHO:

ABC News - Researchers from OpenAI and Boston Children's Hospital conducted the study.

ABC News - Kyra Benton, a 28-year-old patient, was diagnosed with Myofibrillar Myopathy (MFM) after nearly two decades of uncertainty.

Yahoo News Canada - Catherine Brownstein, a lead researcher from the Manton Center for Orphan Disease Research at Boston Children's Hospital, described the AI model as “a total game changer.”

Yahoo News Canada - Suyash Shringarpure, another author of the study and a researcher at OpenAI, contributed insights on the research process.

WHAT:

ABC News - An AI model is helping patients receive diagnoses for rare diseases after years of uncertainty.

ABC News - The study analyzed genetic data from 18 pediatric patients, many of whom are now adults.

Yahoo News Canada - The AI model identified new diagnoses for 18 children, including 10 with rare neurodevelopmental diseases, four with neuromuscular disorders, two who had died suddenly, and two with early psychosis.

Yahoo News Canada - The AI model identified nearly five percent of new diagnoses from 376 genomes analyzed.

WHERE:

ABC News - The study was conducted at Boston Children's Hospital.

WHEN:

ABC News - The study was published in the New England Journal of Medicine on a Thursday.

Yahoo News Canada - The AI model was released in April 2025.

WHY:

ABC News - The researchers aim to help thousands of American children impacted by rare diseases, as one in 10 Americans has a rare disease.

Yahoo News Canada - The study highlights the need for revisiting old genetic test results due to rapid advancements in genetic research.

HOW:

ABC News - The AI model served as an additional tool for specialists, helping them analyze complex information in about six to ten minutes per case.

Yahoo News Canada - Researchers used the AI model by feeding it information such as doctors’ notes, patients’ symptoms, and potential genetic links, which were then reviewed by humans for final diagnoses.

(UPDATED June 21, 2026 01:47 UTC)

Researchers from OpenAI and Boston Children's Hospital have developed an AI model that is revolutionizing the diagnosis of rare diseases, helping patients like 28-year-old Kyra Benton, who spent nearly two decades uncertain about her condition. The study analyzed genetic data from 18 pediatric patients, leading to new diagnoses for many, including ten with rare neurodevelopmental diseases and four with neuromuscular disorders, all while highlighting the importance of revisiting old genetic test results. This innovative tool allows specialists to analyze complex information in just six to ten minutes per case, significantly improving the diagnostic process for thousands of American children affected by rare diseases.