Improving Alzheimer’s blood tests by accounting for personal differences

The new NORMAL: Normalizing the within-person variability in the Alzheimer’s blood test

This project aims to improve the accuracy of Alzheimer’s disease blood tests by accounting for personal differences in protein levels. It is conducted through a newly established public-private partnership that brings together academic researchers in neurochemistry and AI and industry partners. By combining scientific methods and expertise with cutting-edge technology, the project seeks to develop a more reliable, personalized blood test that can support earlier and more accurate diagnosis of Alzheimer’s disease.

Alzheimer’s disease affects more than 55 million people worldwide, a number expected to rise rapidly in the coming decades. In Europe alone, the economic cost of dementia is estimated at over €300 billion per year. Early and accurate diagnosis is essential for effective treatment and care, yet current blood tests show significant variability between individuals, making its diagnostic potential less precise. Improving the reliability of these tests could enhance patient care, reduce healthcare costs, and support broader access to emerging treatments.

By applying AI methods to rich biological datasets, the projects aims to find new protein markers in the blood that explain variability in current Alzheimer’s blood test results. By measuring these markers and incorporating them into the test, the test can be adjusted for individual differences that would otherwise lead to false results. This approach would allow for a more personalized and precise diagnostic tool, reducing overlap in test results between healthy individuals and patients, and enhancing our ability to low-invasively identify Alzheimer’s disease at an early stage.

The project will deliver a validated set of protein markers that account for individual variability and protocols for integrating these markers into the existing Alzheimer's blood tests. Ultimately, the project will demonstrate the feasibility of a more robust and personalized but still cost-effective and scalable blood test for Alzheimer’s Disease.

Summary
Alzheimer’s blood tests are highly promising, but their results can be affected by person-specific variability. In this project we apply AI methods to biological datasets to develop a more robust version of the Alzheimer's blood test.
Technology Readiness Level (TRL)
5 - 7
Time period
36 months
Partners