P1‐117: Development of an eeg‐based neurometric battery for the assessment of cognitive decline in patients at risk for Alzheimer's disease
Year: 2015
Type: article
Study design: Observational Study
Abstract: Electroencephalogram (EEG) recordings and event-related potentials (ERPs) are beneficial for identification of early indicators of Alzheimer's disease (AD). Attempting to identify the early changes of brain activity related to AD, our study used multiple EEG and ERP measures to detect subtle changes in sensory and perceptual function. Participants’ responses to computerized stimuli were recorded to create neurometric profiles used to discriminate individuals with mild cognitive impairment (MCI) and mild Alzheimer's dementia. After a comprehensive clinical assessment, 30 participants 52-90 years old with MCI or probable mild Alzheimer's dementia were selected from our memory clinic. Participants were presented with a 20 minute computerized auditory and visual stimuli designed to elicit a variety of ERPs, while performing a task for specific events, during which their electrophysiological response was measured. The study was performed in an outpatient setting, and data was analyzed using a multivariate analysis of variance (MANOVA) to determine if the neurometric profiles were discriminatory, and cluster analysis to identify sub-groups. Preliminary data analysis used MoCA scores (median MoCA=18.5) to categorize two groups: a low-scoring, mean MoCA score of 16.00 (SD=2.24), and a high-scoring group, mean MoCA score of 21.14 (SD=1.46). Component measurements were entered into a discriminant analysis and ERP responses marginally predicted participant group, Wilk's λ=.160, χ(8)=14.652, p=.066, Canonical Correlation=.916. Groups differentiated primarily by the N2pc, C1, and vMMN responses, which loaded most strongly onto the discriminant function. When using the derived function, 100% of cases were correctly reclassified into original MoCA groups. Using cross-validated classification, 78.6% of participants were correctly classified (85.7% low-scoring group, 71.4% high-scoring group). A 2 (MoCA: High, Low) x 8 (ERP: P50, Frequency MMN, Duration MMN, P3, N2pc, C1, vMMN, ERN) MANOVA with age included as a covariate, was statistically significant for differences in the profiles of low and high-scoring groups, Wilk's λ=.034, F(8,4) =14.296, p<.011, Partial η=.966. Preliminary results support a 20 minute battery of EEG and ERP tests can be used to discriminate MCI and probable mild Alzheimer's dementia. The EEG battery proved to be a complimentary diagnostic marker that was easy to use in a clinical setting.
Source: Alzheimer s & Dementia
Institutions: Eastern Virginia Medical School, William & Mary
Language: English
Subfield: Psychiatry and Mental health
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W4254646100