Bringing Alzheimer's blood testing to people, wherever they are
Can blood samples collected and processed outside traditional clinical settings still accurately detect Alzheimer's disease pathology?
The problem
What's the challenge, and why is it so hard to solve?
Even as FDA-cleared blood tests for Alzheimer's reach specialty and primary care, measuring these biomarkers still requires a venous draw at a clinic. These samples require processing including centrifugation, and storage under controlled conditions within a limited timeframe. For older adults with limited mobility, people far from clinics, or without transportation, that trip is a real barrier. Collecting outside the clinic would open the door. However, time, temperature, and shipping conditions encountered during transport may affect biomarker stability and influence test results.
Our approach
How are we tackling the challenge?
Rather than commit to a single device, we treat remote sampling itself as the problem. We ask what a biomarker must survive to travel from home to the lab, then evaluate collection formats against that bar: how faithfully each preserves the signal through real-world handling, and how the values track the standard venous test. This establishes a sampling approach that holds up outside the clinic.
Why it matters
What difference could this make for patients and society?
Most early cognitive decline goes undetected, and people often wait a long time before raising memory concerns with a doctor. Sampling that works outside the clinic could give people biological answers earlier, ease the burden on specialty centers, and open the door to treatments that work best early. It can also reach rural and underserved communities that memory clinics rarely serve.
What's next
Once we've solved this, what do we want to explore next?
Can people collect their own samples reliably, at scale, and out in the field? And can a decentralized workflow serve the diverse populations most often left out of biomarker research?
Key publications
Where can I read more?
Team
Who's working on this?