How AI helped treat a newborn’s ultra rare disease. ‘It was almost like a light switch.’

Jorie Kraus spent her first 73 days in a neonatal intensive care unit, struggling with a rare genetic disorder that left her muscles, heart, and breathing severely impaired. Her parents, Joanie and her partner, faced repeated fears of losing their daughter until doctors at Mayo Clinic sequenced her genome and applied an artificial intelligence tool called Biomedical Data Translator. This AI analyzed vast biomedical databases to pinpoint Klonopin—a common muscle relaxer—as a potential treatment for her symptoms. The discovery transformed Jorie’s prognosis. Within weeks of starting Klonopin, she began moving freely, interacting with toys, and even speaking simple words, despite earlier developmental delays. Her mother, Joanie Kraus, described the change as ‘almost like a light switch’ during a presentation at STAT’s Breakthrough Summit West in San Francisco. Jorie, now nearly three, continues to require lifelong care but has made remarkable progress. The AI tool, funded by the National Institutes of Health and developed by a research consortium, integrates disparate biomedical data to identify treatment links humans might miss. Physician Whitney Thompson emphasized its critical role, stating the AI ‘made inferences across all biomedical literature’ that would have otherwise gone unnoticed. However, barriers remain, including limited access to rapid genomic sequencing and lab testing in diverse healthcare settings. Jorie’s parents launched the Jorie Effect, a nonprofit supporting families affected by similar genetic deletions. They’ve already seen impact: a neurologist recently contacted them about an older child with comparable symptoms, suggesting the AI-driven approach could help others. While scaling the technology remains challenging, Mayo Clinic’s work highlights AI’s potential to revolutionize rare disease treatment.