The Knight Alzheimer Research Imaging dataset: a comprehensive multimodal resource for exploring aging, preclinical, and symptomatic Alzheimer disease pathology

Abstract

Alzheimer disease (AD) remains a significant global public health challenge, requiring robust multimodal datasets to elucidate its prolonged preclinical phase, improve early detection, advance understanding of disease trajectories, and guide intervention strategies. To address this, the Charles F. and Joanne Knight Alzheimer Disease Research Center (Knight ADRC) at Washington University in St. Louis established the Knight Alzheimer Research Imaging (KARI) dataset. This paper characterizes this dataset emphasizing its phenotypical depth and longitudinal scope, detailing comprehensive multimodal neuroimaging from 1,645 participants (aged 42-97) across 6,217 acquisitions. Spanning the AD spectrum from healthy aging to symptomatic disease, the cohort undergoes extensive longitudinal imaging using structural and functional magnetic resonance imaging (MRI) alongside positron emission tomography (PET) tracers for amyloid and tau pathology. This is complemented by rich clinical, cognitive, genetic, and biomarker data. In addition to raw imaging, the dataset provides quality-controlled processed outputs, including anatomical segmentations and biomarker quantification. By making this data accessible to researchers, the Knight ADRC aims to accelerate discoveries in pathophysiology, biomarker identification, and therapeutic development.