Predictive PET imaging of Alzheimer’s Disease

Alzheimer's Disease (AD) is a debilitating brain disorder that progressively destroys memory, thinking, speech, and the ability to perform simple tasks. Affecting around 44 million people globally, it imposes significant burdens on patients, their families, and society. AD is characterized by the accumulation of misfolded proteins, such as amyloid beta plaques and tau protein, which damage brain cells.
Recent treatments, like the anti-amyloid beta antibody Lecanemab, have shown promise in slowing amyloid plaque formation, gaining fast-track FDA approval. However, patient selection for these therapies requires amyloid beta load quantification, best achieved through PET imaging. Despite its effectiveness, PET imaging is limited by the need for radioactive tracers and high costs.
MRI, a more accessible imaging method, lacks PET's specificity but can be enhanced using AI. By training neural networks to map MRI data to PET biomarkers, researchers aim to predict amyloid plaques and other relevant tracers. This approach could provide early diagnosis and treatment assessment without the drawbacks of PET, reducing costs and expanding access to advanced diagnostics, especially in underserved areas.