Atropos Health and Novartis Launch AI-Driven Collaboration to Accelerate Rare Disease Diagnosis for Paroxysmal Nocturnal Hemoglobinuria

Atropos Health has announced a strategic collaboration with Novartis to accelerate the diagnosis of paroxysmal nocturnal hemoglobinuria (PNH), a rare and serious blood disorder. The initiative leverages artificial intelligence and real-world clinical data to identify undiagnosed patients earlier, significantly reducing the time between symptom onset and treatment. The collaboration centers on developing and evaluating AI models trained on high-quality real-world data from the Atropos Evidence Network, including the GENEVA OS® platform. These models are designed to detect potential PNH cases by analyzing longitudinal patient records and clinical patterns across diverse health systems. The initial AI model, now available for integration into healthcare systems, aims to help clinicians identify at-risk patients earlier, enabling faster testing and diagnosis. PNH affects approximately 10 to 20 people per million globally and is often diagnosed years after symptom onset due to its varied and non-specific presentation, which can involve multiple organ systems. Many patients experience delays exceeding one year, with some waiting over five years. Early diagnosis is critical, as timely treatment can prevent severe complications and improve long-term outcomes. Dr. Brigham Hyde, CEO and co-founder of Atropos Health, emphasized the transformative potential of AI in precision medicine. “Building AI models on real-world data reduces diagnostic uncertainty and empowers providers to act sooner,” he said. “For patients, this means faster access to care—potentially life-changing. For health systems, it means better outcomes and higher satisfaction.” Rodney Gillespie, Head of Oncology at Novartis US, highlighted the partnership’s alignment with Novartis’s mission to deliver meaningful patient impact. “Using AI and machine learning to accelerate diagnosis reflects our commitment to precision health,” he said. “By identifying PNH earlier, we can help patients access appropriate care sooner, reducing delays that can severely affect quality of life.” The AI models developed through this collaboration will be deployed across health system members of the Atropos Evidence Network and integrated directly into clinical workflows at the point of care. This enables real-time, actionable insights that support faster, data-driven decisions. Atropos Health’s work builds on its earlier advancements in AI model training and its partnerships with organizations like Arcadia to scale precision medicine. The company continues to expand its capabilities in developing advanced models—including MOTOR, CLMBR, and Foundation models—on its real-world data network. Atropos Health, creator of GENEVA OS™, is dedicated to transforming healthcare by turning real-world data into timely, actionable evidence. The company helps healthcare and life science organizations close evidence gaps, improve patient outcomes, and accelerate medical innovation.