Mother uses ChatGPT to pinpoint ultra-rare genetic disorder in infant
After doctors could not link her baby Lily's heart, bladder and eye problems, Rosie Griffiths consulted ChatGPT, which suggested Multisystemic Smooth Muscle Dysfunction Syndrome, leading to a genetic diagnosis.
During pregnancy and after birth, Lily displayed a range of issues—including an enlarged bladder, a patent ductus arteriosus that required open-heart surgery at four months, and permanently dilated pupils. Unsatisfied with fragmented explanations, her mother Rosie Griffiths typed the combined symptoms into ChatGPT, which instantly identified Multisystemic Smooth Muscle Dysfunction Syndrome, a disorder affecting only a handful of people worldwide.
Armed with this information, Rosie urged Lily's doctors to order genetic testing, which in January 2026 revealed the ACTA2 R179H mutation confirming MSMDS. The diagnosis allows specialists in cardiology, neurology, urology, gastroenterology and genetics to closely monitor Lily and intervene before severe events occur. Rosie and her family have since founded ACTA2 Alliance UK to fund research, raising over £40,000 toward potential treatments. She credits the AI tool for connecting disparate symptoms that individual clinicians had not linked.
Why it matters
The case shows how AI can help patients and doctors identify ultra-rare diseases that might otherwise be missed.
In this story
