7 Computational EEG Analysis for the Diagnosis of Psychiatric …
167
Fig. 7.7 Effect-size of differences of nodal clustering coefficients between post-traumatic stress
disorder (PTSD) and healthy controls in three frequency bands. The threshold value was set as 0.06
(medium effect). In the brain model, the density of colors and size of circles represent the difference
direction and effect-size, respectively. *η240.06. Cited from the article of Shim et al. [121]
7.4 Diagnosis for Cognitive Impairment: Dementia
and Minimal Cognitive Impairment
Neurophysiological markers, including EEG and ERP components, have been proposed in the literature to serve as indicators of cognitive decline [105]. This can further
assist identification of current disease stage [3], and even treatment assessment [132].
Slowing of the EEG rhythm has been reported in dementia of the Alzheimer’s (AD)
type [83]. More specifically, theta band power has been found to increase and the
alpha power to decrease in the left hemisphere for patients with AD. Simultaneously,
all band powers showed decreased activity in the right hemisphere [1]. D’Onofrio and
colleagues reported that the increase of delta and theta activity, and decrease of alpha
distinguish normal aging, minimal cognitive impairment (MCI) and AD [17]. In a
more recent study, Kim and colleagues reported that patients with AD with moderate
severity had significantly increased theta and decreased beta power compared with
healthy controls [61]. In patients with AD, theta power was significantly correlated
with a poor performance for global cognition. However, beta power was positively
correlated with a good performance for global cognition, attention, memory, visuospatial function, and executive function. Source activity analysis identified superior
temporal gyrus, transverse temporal gyrus, insula, postcentral gyrus, cuneus, and
lingual gyrus regions to be significantly affected in patients with AD.
It has been shown that MMN amplitude decreases with aging when short ISIs
are used, while memory trace decays faster in patients with AD when long ISIs
are applied [110]. The MMN has also been used as a feature of prognosis in MCI
progression to AD [105].
As previously mentioned, P300 is one of the most studied ERP components with
regards to attention, working memory, and cognitive decline. Therefore, it has been
often studied in respect to dementia and cognitive impairment. The P3a has been
associated with attention mechanisms and novel stimulus processing, while P3b, the
second subcomponent, has been more related to stimulus evaluation and decision
making [111]. The latency and amplitude of P300 have been shown to aptly reflect
neurodegeneration progress [67, 89], and significant differences in its characteristics
have been found between controls and MCI or AD patients [8, 27, 44, 105, 109].
However, several studies have reported minor variability in P300 features as well [2,
70], which raises questions to whether the P300 can act as a reliable biomarker of
cognitive impairment.
Park et al. [107] computed the global field synchronization (GFS), a measure
of functional synchronization, which were lower in beta1, beta2, beta3, and full
bands in AD patients than in healthy controls. These values positively correlated
with the mini mental state examination (MMSE) and clinical dementia rating (CDR)
167
Fig. 7.7 Effect-size of differences of nodal clustering coefficients between post-traumatic stress
disorder (PTSD) and healthy controls in three frequency bands. The threshold value was set as 0.06
(medium effect). In the brain model, the density of colors and size of circles represent the difference
direction and effect-size, respectively. *η240.06. Cited from the article of Shim et al. [121]
7.4 Diagnosis for Cognitive Impairment: Dementia
and Minimal Cognitive Impairment
Neurophysiological markers, including EEG and ERP components, have been proposed in the literature to serve as indicators of cognitive decline [105]. This can further
assist identification of current disease stage [3], and even treatment assessment [132].
Slowing of the EEG rhythm has been reported in dementia of the Alzheimer’s (AD)
type [83]. More specifically, theta band power has been found to increase and the
alpha power to decrease in the left hemisphere for patients with AD. Simultaneously,
all band powers showed decreased activity in the right hemisphere [1]. D’Onofrio and
colleagues reported that the increase of delta and theta activity, and decrease of alpha
distinguish normal aging, minimal cognitive impairment (MCI) and AD [17]. In a
more recent study, Kim and colleagues reported that patients with AD with moderate
severity had significantly increased theta and decreased beta power compared with
healthy controls [61]. In patients with AD, theta power was significantly correlated
with a poor performance for global cognition. However, beta power was positively
correlated with a good performance for global cognition, attention, memory, visuospatial function, and executive function. Source activity analysis identified superior
temporal gyrus, transverse temporal gyrus, insula, postcentral gyrus, cuneus, and
lingual gyrus regions to be significantly affected in patients with AD.
It has been shown that MMN amplitude decreases with aging when short ISIs
are used, while memory trace decays faster in patients with AD when long ISIs
are applied [110]. The MMN has also been used as a feature of prognosis in MCI
progression to AD [105].
As previously mentioned, P300 is one of the most studied ERP components with
regards to attention, working memory, and cognitive decline. Therefore, it has been
often studied in respect to dementia and cognitive impairment. The P3a has been
associated with attention mechanisms and novel stimulus processing, while P3b, the
second subcomponent, has been more related to stimulus evaluation and decision
making [111]. The latency and amplitude of P300 have been shown to aptly reflect
neurodegeneration progress [67, 89], and significant differences in its characteristics
have been found between controls and MCI or AD patients [8, 27, 44, 105, 109].
However, several studies have reported minor variability in P300 features as well [2,
70], which raises questions to whether the P300 can act as a reliable biomarker of
cognitive impairment.
Park et al. [107] computed the global field synchronization (GFS), a measure
of functional synchronization, which were lower in beta1, beta2, beta3, and full
bands in AD patients than in healthy controls. These values positively correlated
with the mini mental state examination (MMSE) and clinical dementia rating (CDR)