AI-driven diagnostic algorithm enhances early detection of paroxysmal nocturnal hemoglobinuria in real-world settings.
AI screening caught rare blood disorder cases that had gone undiagnosed for years, shortening the path to lifesaving treatment.
An AI algorithm deployed across 14 Polish healthcare organizations screened 1.3 million patients and identified individuals at high risk for paroxysmal nocturnal haemoglobinuria (PNH), a rare life-threatening blood disorder where diagnosis is delayed over 5 years in nearly a quarter of cases. The system improved positive predictive value from 6.9% to 10.9% and flagged patients with atypical presentations that had gone unrecognised for up to 3.7 years.
What the study was
- Study design
- real_world_prospective_deployment
- Category
- hematologic_malignancies
- Maturity
- Validated
- Journal
- NPJ Digit Med
Why it surfaced
Real-world prospective deployment of AI for ultra-rare hematologic disease across 1.3M patients; directly demonstrates scalable EHR-based screening that reduces diagnostic odyssey; aligns with hematology and AI diagnostics topics simultaneously; AstraZeneca-sponsored but prospective deployment strengthens validity.
A plain-language summary of published research — not medical advice. Talk to a clinician about your own care.