Studying the Mechanisms of Alzheimer's Disease Progression with Deep Learning
Eran Dayan
May 31, 2023, Wednesday, 3:00 PM - 4:00 PM EDT
Cognitive aging is associated with substantial interindividual variability. While some individuals exhibit poor and accelerated cognitive aging, others appear to be more protected from normal and pathological age-associated cognitive decline. Marked variability is particularly notable in the progression of cognitive decline along the Alzheimer's disease (AD) continuum. Constraining this variability is instrumental for the future success of preventive and disease-modifying interventions for AD. In my talk I will describe recent work from my laboratory, where we developed methods based on explainable deep learning to examine progression along the AD continuum, focusing on the contribution of regional brain atrophy and biomarker synergies to progression rates. Altogether, our research demonstrates the utility of using deep learning methods to better understand variation in progression rates observed in normal and pathological cognitive aging.
Eran is an Associate Professor in the Department of Radiology, and a faculty member at the Biomedical Research Imaging Center. He completed his Ph.D. in Neuroscience, working in the departments of Computer Science and Applied Mathematics and Brain Sciences at the Weizmann Institute of Science, and was then a Postdoctoral Fellow at the National Institute of Neurological Disorders and Stroke. He joined UNC-Chapel Hill in 2016.

His lab studies neurodegeneration in the human brain using data science and informatics approaches. His lab seeks to identify the fundamental properties of brain organization and reorganization occurring due to neurodegenerative diseases, and develop methods and tools that can eventually be used for diagnostic and prognostic purposes.