INTRODUCTION: The number of patients living with Alzheimer’s and other neurodegenerative diseases is rapidly growing. In the absence of any cure, there is a need for accurate, scalable, and non-invasive methods to detect these diseases in the earliest stages and to quantify treatment outcomes. The objective of this work is to derive EEG-based biomarkers from data collected during both sleep and waking in order to provide methods for the early detection of neurodegenerative diseases and to acquire insights into the neuropathology associated with these diseases.
METHODS Daytime EEG was collected from a total of AD (n=19), MCI (n=31), and age matched Healthy Controls (HC; n=41) using a 24-channel wireless EEG system (Advanced Brain Monitoring, Inc.) synchronized with administration of the neurocognitive tasks:
• Resting state eyes open and eyes closed
• 3-choice vigilance task for sustained attention (3CVT) • Image recognition visual memory task (SIR)
• Sternberg verbal memory task (VMS)
Power spectral densities (PSDs) were calculated during resting state and event-related potentials (ERPs) were computed for memory and attention tasks. Two nights of in-home sleep data were acquired with Sleep Profiler™ and sleep architecture, power spectral densities, and other sleep variables were computed after artifact rejection.