Researchers from Boston University have made significant strides in addressing the global burden of Alzheimer’s disease (AD). With the projected cost of caring for millions with AD exceeding $1 trillion soon, they developed a deep learning framework to identify individuals at high risk of progression to AD in the early stages of the disease.
By studying data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) and National Alzheimer’s Coordinating Center (NACC), they categorized individuals with mild cognitive impairment (MCI) based on their brain fluid amyloid-β levels. The researchers then used neural networks and survival analysis to predict MCI progression to AD, linking their model predictions with biological evidence.
Their innovative approach, combining neurology and computer science, identified crucial brain regions, like the medial temporal lobe, associated with AD progression. This model utilized routinely collected information like structural MRI, making it accessible and non-invasive. The team’s efforts represent a promising step towards early intervention and improved patient care, offering hope to patients and caregivers alike. The study has been published in the journal iScience.