Synergetic analysis of spatio-temporal EEG patterns: Alzheimer's disease

Abstract

A spatio-temporal analysis has been employed on EEG signals recorded in groups of patients with Alzheimer's disease (AD). The so-called Karhunen-Loeve analysis method was applied to four groups of subjects: 12 patients who were diagnosed as having severe AD, 8 mild AD patients, 10 vascular dementia patients, and 10 normal aged controls, to obtain the spatio-temporal eigenpatterns. The local difference of the global EEG pattern in mild AD patients from that of normal controls was expanded to the frontal regions in the case of severe AD. The analysis showed significant bilateral temporal functioning differences and inter-hemispheric connection difficulty, along with remarkable intra-hemispheric correlation in severe AD patients. Compared to the normal controls, severe AD patients' brains exhibit only weak local connections and correlations, resulting into limited harmonious inter-hemispheric information processing. The results from the spatio-temporal EEG study of AD patients can be considered to be due to a change in the relative activity of the brain corresponding to the pathologic variation in AD, and the results are in accordance with reported clinical studies.

Understanding brain function in vascular cognitive impairment and dementia with EEG and MEG: A systematic review

“…However, VCI patients presented lower Wackermann’s Ω complexity when compared with AD patients with a similar degree of deterioration, assessed using MMSE. The pattern of differences in the distribution of the most dominant components appeared similar when comparing VCI or severe AD to HC, while the differences were milder when comparing moderate AD to HC ( Kim et al, 2001 ).…”

“…All of them used EEG recordings. Two were recorded during visual tasks ( Lou et al, 2011 , Xu et al, 2012 ), 1 during an auditory task ( Al-Qazzaz et al, 2017a ), 2 during resting-state alone ( Jeong et al, 2001 , Kim et al, 2001 ), and 1 during resting-state and photic stimulation ( Lin et al, 2015 ). Similar to connectivity analysis, several possible methods can be used to calculate entropy and complexity.…”

“…However, in other resting state studies, VCI patients exhibited a lower Wackermann’s Ω complexity, which indicates that results could be explained with a lower number of independent components. The most dominant components, which accounted for more variability of the total signal in VCI than in HC, also varied in their topological distribution, being more significant in postero-temporal areas and less in central and anterior areas ( Kim et al, 2001 ).…”

EEG measures for clinical research in major vascular cognitive impairment: recommendations by an expert panel

“…Complexity measures of the rsEEG rhythms were derived by information theory or nonlinear dynamical theory to account for the brain as a system characterized by complex spatiotemporal dynamics of neural networks ( Gao et al, 2011 ). Kim et al (2001) employed a test of Wackermann's global complexity and found decreased complexity in major VCI patients compared with agematched CU persons. Using the nonlinear dynamical measures including dimensional complexity and Lyapunov exponents, Jeong et al (2001) found that the values of the complexity measures in rsEEGs were pathologically increased in several electrodes of major VCI patients, whereas the complexity measure values in most electrodes were decreased in ADD patients.…”

Alzheimer's Disease: Models of Computation and Analysis of EEGs

“…In patients with AD, studies report reduced values for the global correlation dimension and the global largest Lyapunov exponent but no significant differences were found in the values of the global Kolmogorov-Sinai entropy. 125 –127 Kim et al 128 propose the use of eigenvalue distribution using Karhunen-Loeve decomposition to distinguish between AD patients and normal controls. They report that in severe AD the three largest eigenvalues are much larger than the rest and consequently, the eigenvalue distribution decays rapidly, whereas in normal controls the decay is smoother.…”