Multi-site longitudinal Integrated Neurocognitive and Sleep-Behavior Profiler for the Endophenotypic Classification of Dementia Subtypes (INSPECDS)

It is estimated that Alzheimer’s and other neurodegenerative diseases causing dementia will
surpass cancer as the second leading cause of death by the year 2040. Alzheimer’s disease (AD) is the
leading cause of dementia, followed by synucleinopathies, including dementia with Lewy bodies (DLB),
Parkinson’s disease with dementia (PDD), and Fronto-temporal dementia. There is an urgent, unmet
need for effective tools to aid in the classification of dementia subtypes, in the earliest detectable stages
of the pathophysiological process. To address this, Advanced Brain Monitoring (ABM) is leveraging
day/night assessment technologies to create the Integrated Neurocognitive and Sleep-Behavior Profiler
for the Endophenotypic Classification of Dementia Subtypes (INSPECDS) to profile Alzheimer’s and other
dementias. The components of the platform are the Alertness and Memory Profiler (AMP), the Sleep
Profiler (SP), and integrated machine-learning, classification algorithms, hosted on a secure cloud-based,
infrastructure for automated data processing, analysis, & reporting. AMP is unique among
neurocognitive testing platforms in that it is the only one that integrates advanced electrophysiological
measures (e.g., 24-channel wireless EEG/ECG) during the performance of computerized neurocognitive
tasks and has proven effective in characterizing cognitive decline in Alzheimer’s disease. This capability
permits researchers to explore real-time relations between fluctuations in alertness, discrete cognitive
functions, and specific neural processes believed to subserve observed performance deficits in
Alzheimer’s disease and other dementias. The SP is FDA-cleared, easily applied, wireless-EEG device that
was developed and validated to measure sleep architecture for in-home sleep studies with submental
EMG and wireless accelerometers to monitor head and limb movements to quantify the characteristics
of REM-sleep behavior disorder, considered to be a prodromal expression of synucleinopathy. The
application of machine-learning, classification algorithms streamlines the processing and analyses of
these data to derive statistical probabilities of Alzheimer’s disease and other dementia subtypes. The
overarching goal of the current submission is to finalize implementation of a secure, cloud-based
infrastructure to compile the data obtained from the AMP and SP, train classification algorithms to
discriminate among Alzheimer’s disease and other dementia subtypes, validate diagnostic accuracy, and
integrate optimized classifiers within the cloud-based architecture. The INSPECDS system is the first
clinical research tool of its kind with application in both university-based research settings and
pharmaceutical clinical trials to aid in the endophenotypic stratification of Alzheimer’s disease and other
dementias.

Grant Number: 5R44AG050326-05
NIH Institute/Center: NIH
Principal Investigator: Chris Berka