International Journal of Psychophysiology
Volume 26, Issues 1–3,
, Pages 299-317
Author links open overlay panel
The literature on the relationship between background EEG and event-related brain potentials (ERPs) is reviewed, with the conclusion that variation in the former can contribute to individual variability in the latter. The effects of background EEG activity on the P300 component are then described using the results of three experiments. Study 1 assayed the association between EEG spectral power/mean frequency and P300 amplitude/latency measures in young adults. For the slower delta, theta, and alpha bands generally strong correlations were obtained for both types of measures. Study 2 employed similar techniques to assess a large sample of adults who varied in age from 20–80+ years. EEG power in the slower bands was correlated positively with P300 amplitude across the age range, but few effects for mean frequency/component latency were observed. Study 3 measured a group of young adults ten times very 20 min to assess for temporal changes in the relationship between EEG and ERPs. The correlations between spectral power and P300 amplitude measures were found to vary in a manner that suggested the influence of ultradian rhythms on neuroelectric activity. Taken together, the findings from all three study indicate that background EEG variation contributes significantly to the individual variability of the P300 ERP. Theoretical and applied implications of the findings are discussed.© 1997 Elsevier Science B.V.
The P3(00) event-related brain potential (ERP) has demonstrated significant promise as an electrophysiological measure of human cognition (Donchin and Coles, 1988; Donchin et al., 1986). Variation in its amplitude (Pfefferbaum et al., 1989; Smith and Halgren, 1989; Fabiani et al., 1990; Noldy et al., 1990) and latency in normals (Howard and Polich, 1985; Emmerson et al., 1990; Johnson et al., 1985; O'Donnell et al., 1990; Polich et al., 1990b) as well as in clinical populations (Polich et al., 1986, Polich et al., 1990a; O'Donnell et al., 1987; Goodin et al., 1992) suggests that the P300 reflects individual differences in cognitive capability. However, despite the discovery of variables that contribute to between-subject variation in P300 measures (e.g. Geisler and Polich, 1990, Geisler and Polich, 1992a, Geisler and Polich, 1992b; Polich and Geisler, 1991; Polich and Martin, 1992; see Polich and Kok, 1995for a review), the utility of the P300 and ERPs in general has been limited by the variability observed for component measures (Polich, 1986; Polich, 1992; Alexander et al., 1994).
Even though the exact origins of ERP variability are not known, it is reasonable to suppose that background electroencephalographic (EEG) activity may contribute to variation in ERP values. The paper will review the literature context of this hypothesis and summarize the results of several initial investigations on the relationship between EEG and P300 variability. The focus of these studies was: (1) to determine if EEG and P300 measures are related in normal, young adult subjects; (2) to ascertain how such a relationship might change with normal aging; and (3) to explore the possible influence of ultradian rhythms on the EEG/P300 association. If variation in EEG is associated with individual differences in the P300 ERP, quantification of the connections between these neuroelectric phenomena should help account for some of the variability when ERPs are used to assess cognitive function.
The effects of EEG activity on sensory evoked potentials (EPs) have been investigated in various studies. In general, comparisons between background EEG and EP values have been made either by: (1) recording EEG in groups of individuals — sometimes selected for special characteristics (e.g., low vs. high alpha power) — and assessing how EEG variability affects EP amplitude and latency values between subjects; or (2) recording pre- and post-stimulus EEG and assessing how pre-stimulus EEG variability affects the post-stimulus EP measures within subjects. Each of these approaches has produced a disparate literature, although some generalizations about the findings can be made.
The first procedure was employed in many early studies and generally found that large magnitude EEG was associated with large EP amplitudes despite considerable differences in methodology (Barlow, 1960; Kooi and Bagchi, 1964; Rodin et al., 1965; Spilker et al., 1969). These results were supported by investigations demonstrating that experimentally induced changes in the EEG, especially in the alpha band, also affected EPs (Cigánek, 1961; Garcı́a-Austt, 1963; Rémond and Lesèvre, 1967), perceptual phenomena (Donchin and Lindsley, 1965; Nunn and Osselton, 1974), and reaction time (Callaway and Yeager, 1960; Surwillo, 1963; Dustman and Beck, 1965; Surwillo, 1971). Hence, variation of background EEG appears to influence sensory EP measures across individuals, although a clear consensus of the effects is not apparent primarily because of wide variation in recording techniques and experimental methodology.
The second approach has been adopted by more recent reports, which was first introduced by Basar and his colleagues using animals (Basar et al., 1976a; Basar et al., 1979) and later applied to humans (Basar et al., 1976b, Basar et al., 1984; Basar, 1980; Basar and Stampfer, 1985; Stampfer and Basar, 1985). These studies have found reasonably strong associations between pre-stimulus EEG and the post-stimulus EP measures primarily in the slower theta as well as alpha bands, with large-magnitude pre-stimulus EEG again related to large amplitude EPs. However, the results were affected by stimulus parameters and whether the subject was engaged in an explicit task (Grillon and Buchsbaum, 1986; Romani et al., 1988; Brandt et al., 1991; Jansen and Brandt, 1991). Thus, variation in background EEG appears to be associated with characteristics of the averaged sensory EP within and between subjects across variegated studies (Boddy, 1971; Galin and Ellis, 1975; Rogers, 1980).
A positive relationship between pre-stimulus EEG spectral power, primarily in the theta and alpha bands, and amplitude of the P300 ERP has been reported (Basar et al., 1984; Pritchard et al., 1985; Jasiukaitis and Hakerem, 1988; Basar et al., 1989; Jasiukaitis et al., 1990). Indeed, a recent investigation has found that EEG and the P300 component interactively reflect the mental events underlying information processing during complex memory tasks (Mecklinger et al., 1992). These studies suggest that within-subject EEG variability is associated with variation in P300 ERP measures. However, the relationship between background EEG and ERP variability needs to be characterized more precisely to articulate exactly how the EEG is related to ERP phenomena across individuals.
Overview and methodology
Toward this end, the general method employed in the series of studies described below was to obtain both EEG and ERPs from the same subjects, quantify the electrophysiological data using standardized techniques, and correlate the measures from each procedure with one another across subjects. This approach was used in Study 1 to assess the individual variability in a homogeneous sample of young adults. Study 2 extended this method to characterize the effects of normal aging. Study 3 addressed
Given this conclusion, one possible mechanism that could underlie the observed interaction between EEG and the P300 component may be 'event-related desynchronization' (ERD) of the EEG alpha band during information processing (Pfurtscheller, 1977, Pfurtscheller, 1992; Pfurtscheller and Aranibar, 1977; Klimesch et al., 1992). ERDs are thought to originate from decreases in alpha band EEG power when attentional resources are allocated for cognitive operations (van Winsum et al., 1984; Sergeant et
Summary and caveat
The present findings support the hypothesis that individual variation in background EEG activity is associated with P300 amplitude and latency variability. Given that P300 measures can reflect cognitive capability, investigation of how the EEG is related to ERP generation should help to delimit the individual variability observed in ERPs from both normal and clinical populations. It should be emphasized, however, that the results reported here are correlational and say little about possible
This work was supported by NIA grant RO1-A610604 and is publication number NP8856 from the The Scripps Research Institute. Portions of this paper were reported at the meetings of the Society for Psychophysiological Research, Chicago, Illinois (1991) and the Evoked Potential International Congress X, Eger, Hungary (1992).
- Alexander, J.E., Polich, J., Bloom, F.E., Bauer, L.O., Kuperman, S., Rohrbaugh, J., Morzorati, S., O'Connor, S.,...
- Barlow, J.S. (1960) Rhythmic activity induced by photic stimulation in relation to intrinsic alpha activity of the...
- Basar, E. (1980) EEG-brain Dynamics: Relation Between EEG and Brain Evoked Potentials. Elsevier/North Holland...
- Basar, E. and Stampfer, H.G. (1985) Important associations among EEG-dynamics, event-related potentials, short-term...
- Basar, E., Gönder, A. and Ungan, P. (1976a) Important relation between EEG and brain evoked potentials. I. Resonance...
- Basar, E., Gönder, A. and Ungan, P. (1976b) Important relation between EEG and brain evoked potentials. II. A systems...
- Basar, E., Durusan, R., Gönder, A. and Ungan, P. (1979) Combined dynamics of EEG and evoked potentials II. Studies of...
- Basar, E., Basar-Eroglu, C., Rosen, R. and Schütt, A. (1984) A new approach to endogenous event-related potentials in...
- Basar, E., Basar-Eroglu, C., Röschke, J. and Schütt, A. (1989) The EEG is a quasi-deterministic signal anticipating...
- Boddy, J. (1971) The relationship of reaction time to brain wave period: a re-evaluation. Electroenceph. Clin....
Prestimulus alpha and beta contributions to equiprobable Go/NoGo processing in healthy ageing
2020, International Journal of Psychophysiology
The immediately-prestimulus electroencephalographic (EEG) brain state influences subsequent event-related processing, dynamically impacting event-related potential (ERP) and behavioural outcomes. Both EEG and ERPs are known to undergo age-related change, yet few have investigated the consistency in their dynamic interrelations in the context of ageing. The present investigation assessed the impact of prestimulus alpha and beta brain states in 20 young (18–26years) and 20 gender-matched healthy older (59–75years) adults who completed an equiprobable auditory Go/NoGo paradigm. Prestimulus alpha and beta band amplitudes in their prominent band topographies were separately used to derive Go and NoGo ERPs at 10 ascending levels of prestimulus activity, and ERP components were derived for these levels using Principal Components Analysis. Prestimulus alpha directly modulated Go/NoGo P3a amplitudes across the groups, while beta inversely modulated the young (cf. older) adult NoGo N1–1, each supporting and extending limited prior research. Several novel effects were also uncovered, most notably an inverse relationship between prestimulus alpha and reaction time. Prestimulus alpha and beta were confirmed as significant determinants of the processing outcomes in this task, and the complex pattern of results provides a normative map in healthy ageing.
Abnormal cortical neural synchronization mechanisms in quiet wakefulness are related to motor deficits, cognitive symptoms, and visual hallucinations in Parkinson's disease patients: an electroencephalographic study
2020, Neurobiology of Aging
Compared with Alzheimer's disease (AD), Parkinson's disease (PD) shows peculiar clinical manifestations related to vigilance (i.e., executive cognitive deficits and visual hallucinations) that may be reflected in resting-state electroencephalographic rhythms. To test this hypothesis, clinical and resting-state electroencephalographic rhythms in age-, sex-, and education-matched PD patients (N= 136) and Alzheimer's disease patients (AD, N= 85), and healthy older participants (Nold, N= 65), were available from an international archive. Electroencephalographic sources were estimated by eLORETA software. The results are as follows: (1) compared to the Nold participants, the AD and PD patients showed higher widespread delta source activities (PD > AD) and lower posterior alpha source activities (AD > PD); (2) the PD patients with the most pronounced motor deficits exhibited very low alpha source activities in widespread cortical regions; (3) the PD patients with the strongest cognitive deficits showed higher alpha source activities in widespread cortical regions; and (4) compared to the PD patients without visual hallucinations, those with visual hallucinations were characterized by higher posterior alpha sources activities. These results suggest that in PD patients resting in quiet wakefulness, abnormalities in cortical neural synchronization at alpha frequencies are differently related to cognitive, motor, and visual hallucinations. Interestingly, parallel PD neuropathological processes may have opposite effects on cortical neural synchronization mechanisms generating cortical alpha rhythms in quiet wakefulness.
Inter-subject P300 variability relates to the efficiency of brain networks reconfigured from resting- to task-state: Evidence from a simultaneous event-related EEG-fMRI study
The P300 event-related potential (ERP) varies across individuals, and exploring this variability deepens our knowledge of the event, and scope for its potential applications. Previous studies exploring the P300 have relied on either electroencephalography (EEG) or functional magnetic resonance imaging (fMRI). We applied simultaneous event-related EEG-fMRI to investigate how the network structure is updated from rest to the P300 task so as to guarantee information processing in the oddball task. We first identified 14 widely distributed regions of interest (ROIs) that were task-associated, including the inferior frontal gyrus and the middle frontal gyrus, etc. The task-activated network was found to closely relate to the concurrent P300 amplitude, and moreover, the individuals with optimized resting-state brain architectures experienced the pruning of network architecture, i.e. decreasing connectivity, when the brain switched from rest to P300 task. Our present simultaneous EEG-fMRI study explored the brain reconfigurations governing the variability in P300 across individuals, which provided the possibility to uncover new biomarkers to predict the potential for personalized control of brain-computer interfaces.
P3b amplitude as a signature of cognitive decline in the older population: An EEG study enhanced by Functional Source Separation
Citation Excerpt :
The potential drawback is that FSS cannot be used to extract unexpected or unknown brain activities, however previous studies have found that this semi-blind approach often outperforms fully blind source separation approaches like ICA (Barbati et al., 2008; Porcaro et al., 2011, 2010). Consistent with previous studies we showed that both peak latency and peak amplitude of the P3a and P3b ERPs changed with age (Fjell et al., 2007; O’Connell et al., 2012a; Polich, 1997; Walhovd and Fjell, 2001). Specifically, the peak latency of both the P3a and P3b responses increased in the elderly group, whereas the peak amplitude of both the P3a and P3b decreased with age.
With the greying population, it is increasingly necessary to establish robust and individualized markers of cognitive decline. This requires the combination of well-established neural mechanisms, and the development of increasingly sensitive methodologies. The P300 event-related potential (ERP) has been one of the most heavily investigated neural markers of attention and cognition, and studies have reliably shown that changes in the amplitude and latency of the P300 ERP index the process of aging. However, it is still not clear whether either the P3a or P3b sub-components additionally index levels of cognitive impairment. Here, we used a traditional visual three-stimulus oddball paradigm to investigate both the P3a and P3b ERP components in sixteen young and thirty-four healthy elderly individuals with varying degrees of cognitive ability. EEG data extraction was enhanced through the use of a novel signal processing method called Functional Source Separation (FSS) that increases signal-to-noise ratio by using a weighted sum of all electrodes rather than relying on a single, or a small sub-set, of EEG channels. Whilst clear differences in both the P3a and P3b ERPs were seen between young and elderly groups, only P3b amplitude differentiated older people with low memory performance relative to IQ from those with consistent memory and IQ. A machine learning analysis showed that P3b amplitude (derived from FSS analysis) could accurately categorise high and low performing elderly individuals (78% accuracy). A comparison of Bayes Factors found that differences in cognitive decline within the elderly group were 87 times more likely to be detected using FSS compared to the best performing single electrode (Cz). In conclusion, we propose that P3b amplitude could be a sensitive marker of early, age-independent, episodic memory dysfunction within a healthy older population. In addition, we advocate for the use of more advanced signal processing methods, such as FSS, for detecting subtle neural changes in clinical populations.
Prestimulus delta and theta contributions to equiprobable Go/NoGo processing in healthy ageing
2018, International Journal of Psychophysiology
Citation Excerpt :
Low frequency delta and theta band EEG activity have been proposed as possible neurophysiological markers of subclinical cognitive dysfunction or decline (Babiloni et al., 2006; Cummins and Finnigan, 2007; Güntekin and Başar, 2016; Yener et al., 2016), hence this investigation focused on activity in these bands. Healthy ageing is typically associated with diminished (e.g., Babiloni et al., 2006; Barry and De Blasio, 2017; Breslau et al., 1989; Cummins and Finnigan, 2007; Duffy et al., 1984; Emek-Savaş et al., 2016; Polich, 1997a, 1997b; Vysata et al., 2012), and more temporally distributed (e.g., Breslau et al., 1989; Klass and Brenner, 1995) activity in these low frequency bands, but their dominant midline topography is reported to be preserved (e.g., Barry and De Blasio, 2017; Breslau et al., 1989; Emek-Savaş et al., 2016). Although ERP component measures (latency, amplitude) are each sensitive to age-related change, component amplitudes are the focus of the current investigation.
Ongoing EEG activity contributes to ERP outcomes of stimulus processing, and each of these measures is known to undergo (sometimes significant) age-related change. Variation in their relationship across the life-span may thus elucidate mechanisms of normal and pathological ageing. This study assessed the relationships between low-frequency EEG prestimulus brain states, the ERP, and behavioural outcomes in a simple equiprobable auditory Go/NoGo paradigm, comparing these for 20 young (Mage = 20.4 years) and 20 healthy older (Mage = 68.2 years) adults. Prestimulus delta and theta amplitudes were separately assessed; these were each dominant across the midline region, and reduced in the older adults. For each band, (within-subjects) trials were sorted into ten increasing prestimulus EEG levels for which separate ERPs were derived. The set of ten ERPs for each band-sort was then quantified by PCA, independently for each group (young, older adults). Four components were primarily assessed (P1, N1-1, P2/N2b complex, and P3), with each showing age-related change. Mean RT was comparable, but intra-individual RT variability increased in older adults. Prestimulus delta and theta each generally modulated component positivity, indicating broad influence on task processing. Prestimulus delta was primarily associated with the early sensory processes, and theta more with the later stimulus-specific processes; prestimulus theta also inversely modulated intra-individual RT variability across the groups. These prestimulus EEG–ERP dynamics were consistent between the young and older adults in each band for all components except the P2/N2b, suggesting that across the lifespan, Go/NoGo categorisation is differentially affected by prestimulus delta and theta.
Laughter catches attention!
2017, Biological Psychology
In social interactions, emotionally salient and sudden changes in vocal expressions attract attention. However, only a few studies examined how emotion and attention interact in voice processing. We investigated neutral, happy (laughs) and angry (growls) vocalizations in a modified oddball task.
Participants silently counted the targets in each block and rated the valence and arousal of the vocalizations. A combined event-related potential and time-frequency analysis focused on the P3 and pre-stimulus alpha power to capture attention effects in response to unexpected events.
Whereas an early differentiation between emotionally salient and neutral vocalizations was reflected in the P3a response, the P3b was selectively enhanced for happy voices. The P3b modulation was predicted by pre-stimulus frontal alpha desynchronization, and by the perceived pleasantness of the targets.
These findings indicate that vocal emotions may be differently processed based on task relevance and valence. Increased anticipation and attention to positive vocal cues (laughter) may reflect their high social relevance.
Issues and considerations for using the scalp surface Laplacian in EEG/ERP research: A tutorial review
International Journal of Psychophysiology, Volume 97, Issue 3, 2015, pp. 189-209
Despite the recognition that the surface Laplacian may counteract adverse effects of volume conduction and recording reference for surface potential data, electrophysiology as a discipline has been reluctant to embrace this approach for data analysis. The reasons for such hesitation are manifold but often involve unfamiliarity with the nature of the underlying transformation, as well as intimidation by a perceived mathematical complexity, and concerns of signal loss, dense electrode array requirements, or susceptibility to noise. We revisit the pitfalls arising from volume conduction and the mandated arbitrary choice of EEG reference, describe the basic principle of the surface Laplacian transform in an intuitive fashion, and exemplify the differences between common reference schemes (nose, linked mastoids, average) and the surface Laplacian for frequently-measured EEG spectra (theta, alpha) and standard event-related potential (ERP) components, such as N1 or P3. We specifically review common reservations against the universal use of the surface Laplacian, which can be effectively addressed by employing spherical spline interpolations with an appropriate selection of the spline flexibility parameter and regularization constant. We argue from a pragmatic perspective that not only are these reservations unfounded but that the continued predominant use of surface potentials poses a considerable impediment on the progress of EEG and ERP research.
Resting-state EEG power and connectivity are associated with alpha peak frequency slowing in healthy aging
Neurobiology of Aging, Volume 71, 2018, pp. 149-155
The individual alpha peak frequency (IAPF) of the human electroencephalography (EEG) typically experiences slowing with increasing age. Despite this hallmark change, studies that investigate modulations of conventional EEG alpha power and connectivity by aging and age-related neuropathology neglect to account for intergroup differences in IAPF. To investigate the relationship of age-related IAPF slowing with EEG power and connectivity, we recorded eyes-closed resting-state EEG in 37 young adults and 32 older adults. We replicated the finding of a slowed IAPF in older adults. IAPF values were significantly correlated with the frequency of maximum global connectivity and the means of their distributions did not differ, suggesting that connectivity was highest at the IAPF. Older adults expressed reduced global EEG power and connectivity at the conventional upper alpha band (10–12Hz) compared with young adults. By contrast, groups had equivalent power and connectivity at the IAPF. The results suggest that conventional spectral boundaries may be biased against older adults or any group with a slowed IAPF. We conclude that investigations of alpha activity in aging and age-related neuropathology should be adapted to the IAPF of the individual and that previous findings should be interpreted with caution. EEG in the dominant alpha range may be unsuitable for examining cortico-cortical connectivity due to its subcortical origins.
Male prairie voles display cardiovascular dipping associated with an ultradian activity cycle
Physiology & Behavior, Volume 156, 2016, pp. 106-116
Mammals typically display alternating active and resting phases and, in most species, these rhythms follow a circadian pattern. The active and resting phases often are accompanied by corresponding physiological changes. In humans, blood pressure decreases during the resting phase of the activity cycle, and the magnitude of that “nocturnal dipping” has been used to stratify patients according to the risk for cardiovascular disease. However, in contrast to most mammals, prairie voles (Microtus ochrogaster) have periods of activity and rest that follow an ultradian rhythm with period lengths significantly <24h. While rhythmic changes in blood pressure across a circadian activity cycle have been well-documented, blood pressure patterns in species that display ultradian rhythms in activity are less well-studied. In the current study, we implanted pressure-sensitive radiotelemetry devices in male prairie voles and recorded activity, mean arterial pressure (MAP), and heart rate (HR) continuously for 3days. Visualization of the ultradian rhythms was enhanced using a 1h running average to filter the dataset. Positive correlations were found between activity and MAP and between activity and HR. During the inactive period of the ultradian cycle, blood pressure decreased by about 15%, which parallels the nocturnal dipping pattern seen in healthy humans. Further, the duration of inactivity did not affect any of the cardiovascular measures, so the differences in blood pressure values between the active and inactive periods are likely driven by ultradian oscillations in hormones and autonomic function. Finally, specific behavioral patterns also were examined. Both the instrumented animal and his non-instrumented cagemate appeared to show synchronized activity patterns, with both animals displaying sleep-like behavior for more than 90% of the inactive period. We propose that the prairie vole ultradian rhythm in blood pressure is an analogue for circadian blood pressure variability and can be used to study the long-term effects of commonly prescribed drugs on blood pressure dipping.
Progressive gene dose-dependent disruption of the methamphetamine-sensitive circadian oscillator-driven rhythms in a knock-in mouse model of Huntington's disease
Experimental Neurology, Volume 286, 2016, pp. 69-82
Huntington's disease (HD) is a progressive genetic neurodegenerative disorder characterised by motor and cognitive deficits, as well as sleep and circadian abnormalities. In the R6/2 mouse, a fragment model of HD, rest-activity rhythms controlled by the suprachiasmatic nucleus disintegrate completely by 4months of age. Rhythms driven by a second circadian oscillator, the methamphetamine-sensitive circadian oscillator (MASCO), are disrupted even earlier, and cannot be induced after 2months of age. Here, we studied the effect of the HD mutation on the expression of MASCO-driven rhythms in a more slowly developing, genetically relevant mouse model of HD, the Q175 ‘knock-in’ mouse. We induced expression of MASCO output by administering low dose methamphetamine (0.005%) chronically via the drinking water. We measured locomotor activity in constant darkness in wild-type and Q175 mice at 2 (presymptomatic), 6 (early symptomatic), and 12 (symptomatic) months of age. At 2months, all mice expressed MASCO-driven rhythms, regardless of genotype. At older ages, however, there was a progressive gene dose-dependent deficit in MASCO output in Q175 mice. At 6months of age, these rhythms could be observed in only 45% of heterozygous and 15% of homozygous mice. By 1year of age, 90% of homozygous mice had an impaired MASCO output. There was also an age-dependent disruption of MASCO output seen in wild-type mice. The fact that the progressive deficit in MASCO-driven rhythms in Q175 mice is HD gene dose-dependent suggests that, whatever its role in humans, abnormalities in MASCO output may contribute to the HD circadian phenotype.
An end-to-end deep learning approach to MI-EEG signal classification for BCIs
Expert Systems with Applications, Volume 114, 2018, pp. 532-542
Goal: To develop and implement a Deep Learning (DL) approach for an electroencephalogram (EEG) based Motor Imagery (MI) Brain-Computer Interface (BCI) system that could potentially be used to improve the current stroke rehabilitation strategies.
Method: The DL model is using Convolutional Neural Network (CNN) layers for learning generalized features and dimension reduction, while a conventional Fully Connected (FC) layer is used for classification. Together they build a unified end-to-end model that can be applied to raw EEG signals. This previously proposed model was applied to a new set of data to validate its robustness against data variations. Furthermore, it was extended by subject-specific adaptation. Lastly, an analysis of the learned filters provides insights into how such a model derives a classification decision.
Results: The selected global classifier reached 80.38%, 69.82%, and 58.58% mean accuracies for datasets with two, three, and four classes, respectively, validated using 5-fold crossvalidation. As a novel approach in this context, transfer learning was used to adapt the global classifier to single individuals improving the overall mean accuracy to 86.49%, 79.25%, and 68.51%, respectively. The global models were trained on 3s segments of EEG data from different subjects than they were tested on, which proved the generalization performance of the model.
Conclusion: The results are comparable with the reported accuracy values in related studies and the presented model outperforms the results in the literature on the same underlying data. Given that the model can learn features from data without having to use specialized feature extraction methods, DL should be considered as an alternative to established EEG classification methods, if enough data is available.
Frontal theta as a mechanism for cognitive control
Trends in Cognitive Sciences, Volume 18, Issue 8, 2014, pp. 414-421
Recent advancements in cognitive neuroscience have afforded a description of neural responses in terms of latent algorithmic operations. However, the adoption of this approach to human scalp electroencephalography (EEG) has been more limited, despite the ability of this methodology to quantify canonical neuronal processes. Here, we provide evidence that theta band activities over the midfrontal cortex appear to reflect a common computation used for realizing the need for cognitive control. Moreover, by virtue of inherent properties of field oscillations, these theta band processes may be used to communicate this need and subsequently implement such control across disparate brain regions. Thus, frontal theta is a compelling candidate mechanism by which emergent processes, such as ‘cognitive control’, may be biophysically realized.
Copyright © 1997 Elsevier Science B.V. All rights reserved.