Our blood is full of information about our health. The difficulty is to find the Alzheimer's pattern among billions of potential signals. Using Artificial Intelligence, our multiomics platform is designed to detect Alzheimer's at its earliest stages to help clinicians optimize the next generation of precision therapies.
Alzheimer's is an extremely complex disease. Spotting the right pattern would have required thousands of well-annotated blood samples. Instead, we pre-identified the most informative biomarkers on the first gene-transfer based animal model of the silent phase of Alzheimer's (Audrain et al., (2018) Cereb Cortex).
Then by training only on hundreds of Alzheimer's blood samples, our multiomics platform learned which biomarkers patterns are associated with each stage of the disease (Asymptomatic, Prodromal, Dementia).
Our algorithms combine proteins and metabolites which greatly enhance the sensitivity and specificity of the diagnosis by considering different biological pathways.
Our biomarkers are not produced by the brain: they are produced or regulated by peripheral organs. This is a major advantage. Using these peripheral biomarkers signals, our blood test can detect Alzheimer’s as soon as the amyloid pathway is engaged while being very specific.
Many diagnoses are based on threshold analyses. If a blood marker exceeds a threshold, you are sick (e.g. diabetes, high cholesterol, hypertension...). It is not relevant for progressive diseases as complex as Alzheimer's. By considering the dynamic aspect of Alzheimer's biomarkers, our algorithms can spot the disease long before the dementia symptoms.
Sign up to receive the latest news on clinical trials, publications, and more.