UDNF joins ARPA-H rare disease AI program as patient experience partner

The Undiagnosed Diseases Network Foundation has been selected to help shape ARPA-H’s RAPID program, a federal effort to speed rare disease diagnosis with AI and data tools. UDNF will bring patient and family experience into the design of the initiative, which targets the years-long diagnostic odyssey faced by millions of Americans.

Why it matters: - ARPA-H’s RAPID program is aimed at shortening the time from first symptoms to diagnosis for people with rare and undiagnosed diseases. - UDNF’s role is intended to keep patient and family needs at the center of AI-driven diagnostic tools. - The initiative could affect millions of people living with rare disease-related uncertainty, missed treatment and high medical costs.

What happened: - The Undiagnosed Diseases Network Foundation said it was selected as a Patient Experience Partner for ARPA-H’s Rare Disease AI/ML for Precision Integrated Diagnostics, or RAPID, program. - The announcement was made Aug. 31, 2026. - UDNF will work with Sage Bionetworks and collaborators on the Rare Disease Data Commons and other RAPID solutions. - Dr. Danielle Carnival, UDNF’s CEO, said the role recognizes that solving the diagnostic odyssey requires patient and family voice as well as science.

The details: - RAPID is built around the question of whether the rare disease diagnostic odyssey can be ended. - UDNF will help researchers, technologists and innovators better reflect the lived experience of undiagnosed patients and families. - The foundation will connect RAPID teams with patients, caregivers and families for input on research design, tool development and deployment. - UDNF will push for solutions that are accessible, equitable and patient-centered. - UDNF will also advocate for underserved and underrepresented communities to be included in RAPID datasets. - ARPA-H’s program is designed to use artificial intelligence and federal investment to improve diagnostic speed and accuracy. - The press release says the diagnostic odyssey averages six years, can last decades and costs about $500,000 per patient. - The release says rare diseases affect more than 10,000 unique conditions and more than 350 million people worldwide, including one in ten Americans. - It also says half of people with a rare disease remain undiagnosed or misdiagnosed.

Between the lines: - The selection signals that patient input is becoming a formal part of rare disease AI development, not an add-on. - The focus on rural and underserved communities suggests RAPID is trying to avoid building tools that work only for patients already close to specialty care. - UDNF’s role could help make the data commons and diagnostic tools more practical for real-world families, not just research settings. - Dr. F Sessions Cole, chair of UDNF’s board, said the program represents the kind of bold thinking needed to advance the foundation’s mission.

What's next: - UDNF is expected to begin working with RAPID teams on patient engagement and community feedback. - The Rare Disease Data Commons and RAPID’s AI-driven diagnostics will continue to be developed with input from patient experience partners. - The program’s impact will depend on whether the tools can be deployed broadly across communities in the United States.

The bottom line: - UDNF’s appointment gives rare disease patients a direct role in shaping a federal AI program built to speed diagnosis and reduce years of uncertainty.