Objective.Deep learning (DL) networks are increasingly attracting attention across various fields, including electroencephalography (EEG) signal processing. These models provide comparable performance to that of traditional techniques. At present, however, there is a lack of well-structured and standardized datasets with specific benchmark limit the development of DL solutions for EEG denoising.Approach.Here, we present EEGdenoiseNet, a benchmark EEG dataset that is suited for training and testing DL-based denoising models, as well as for performance comparisons across models. EEGdenoiseNet contains 4514 clean EEG segments, 3400 ocular artifact segments and 5598 muscular artifact segments, allowing users to synthesize contaminated EEG segments with the ground-truth clean EEG.Main results.We used EEGdenoiseNet to evaluate denoising performance of four classical networks (a fully-connected network, a simple and a complex convolution network, and a recurrent neural network). Our results suggested that DL methods have great potential for EEG denoising even under high noise contamination.Significance.Through EEGdenoiseNet, we hope to accelerate the development of the emerging field of DL-based EEG denoising. The dataset and code are available athttps://github.com/ncclabsustech/EEGdenoiseNet.

High-density electroencephalography (hdEEG) is an emerging brain imaging technique that can be used to investigate fast dynamics of electrical activity in the healthy and the diseased human brain. Its applications are however currently limited by a number of methodological issues, among which the difficulty in obtaining accurate source localizations. In particular, these issues have so far prevented EEG studies from reporting brain networks similar to those previously detected by functional magnetic resonance imaging (fMRI). Here, we report for the first time a robust detection of brain networks from resting state (256-channel) hdEEG recordings. Specifically, we obtained 14 networks previously described in fMRI studies by means of realistic 12-layer head models and exact low-resolution brain electromagnetic tomography (eLORETA) source localization, together with independent component analysis (ICA) for functional connectivity analysis. Our analyses revealed three important methodological aspects. First, brain network reconstruction can be improved by performing source localization using the gray matter as source space, instead of the whole brain. Second, conducting EEG connectivity analyses in individual space rather than on concatenated datasets may be preferable, as it permits to incorporate realistic information on head modeling and electrode positioning. Third, the use of a wide frequency band leads to an unbiased and generally accurate reconstruction of several network maps, whereas filtering data in a narrow frequency band may enhance the detection of specific networks and penalize that of others. We hope that our methodological work will contribute to rise of hdEEG as a powerful tool for brain research. Hum Brain Mapp, 2017.

Resting state networks (RSNs) in the human brain were recently detected using high-density electroencephalography (hdEEG). This was done by using an advanced analysis workflow to estimate neural signals in the cortex and to assess functional connectivity (FC) between distant cortical regions. FC analyses were conducted either using temporal (tICA) or spatial independent component analysis (sICA). Notably, EEG-RSNs obtained with sICA were very similar to RSNs retrieved with sICA from functional magnetic resonance imaging data. It still remains to be clarified, however, what technological aspects of hdEEG acquisition and analysis primarily influence this correspondence. Here we examined to what extent the detection of EEG-RSN maps by sICA depends on the electrode density, the accuracy of the head model, and the source localization algorithm employed. Our analyses revealed that the collection of EEG data using a high-density montage is crucial for RSN detection by sICA, but also the use of appropriate methods for head modeling and source localization have a substantial effect on RSN reconstruction. Overall, our results confirm the potential of hdEEG for mapping the functional architecture of the human brain, and highlight at the same time the interplay between acquisition technology and innovative solutions in data analysis.

In electroencephalography (EEG) measurements, the signal of each recording electrode is contrasted with a reference electrode or a combination of electrodes. The estimation of a neutral reference is a long-standing issue in EEG data analysis, which has motivated the proposal of different re-referencing methods, among which linked-mastoid re-referencing (LMR), average re-referencing (AR) and reference electrode standardization technique (REST). In this study we quantitatively assessed the extent to which the use of a high-density montage and a realistic head model can impact on the optimal estimation of a neutral reference for EEG recordings. Using simulated recordings generated by projecting specific source activity over the sensors, we assessed to what extent AR, REST and LMR may distort the scalp topography. We examined the impact electrode coverage has on AR and REST, and how accurate the REST reconstruction is for realistic and less realistic (three-layer and single-layer spherical) head models, and with possible uncertainty in the electrode positions. We assessed LMR, AR and REST also in the presence of typical EEG artifacts that are mixed in the recordings. Finally, we applied them to real EEG data collected in a target detection experiment to corroborate our findings on simulated data. Both AR and REST have relatively low reconstruction errors compared to LMR, and that REST is less sensitive than AR and LMR to artifacts mixed in the EEG data. For both AR and REST, high electrode density yields low re-referencing reconstruction errors. A realistic head model is critical for REST, leading to a more accurate estimate of a neutral reference compared to spherical head models. With a low-density montage, REST shows a more reliable reconstruction than AR either with a realistic or a three-layer spherical head model. Conversely, with a high-density montage AR yields better results unless precise information on electrode positions is available. Our study is the first to quantitatively assess the performance of EEG re-referencing techniques in relation to the use of a high-density montage and a realistic head model. We hope our study will help researchers in the choice of the most effective re-referencing approach for their EEG studies.

Resting state fMRI (rs-fMRI) is commonly used to study the brain's intrinsic neural coupling, which reveals specific spatiotemporal patterns in the form of resting state networks (RSNs). It has been hypothesized that slow rs-fMRI oscillations (5 Hz); however, causal evidence for this relationship is currently lacking. Here we measured rs-fMRI in humans while applying transcranial alternating current stimulation (tACS) to entrain brain rhythms in left and right sensorimotor cortices. The two driving tACS signals were tailored to the individual's α rhythm (8–12 Hz) and fluctuated in amplitude according to a 1 Hz power envelope. We entrained the left versus right hemisphere in accordance to two different coupling modes where either α oscillations were synchronized between hemispheres (phase-synchronized tACS) or the slower oscillating power envelopes (power-synchronized tACS). Power-synchronized tACS significantly increased rs-fMRI connectivity within the stimulated RSN compared with phase-synchronized or no tACS. This effect outlasted the stimulation period and tended to be more effective in individuals who exhibited a naturally weak interhemispheric coupling. Using this novel approach, our data provide causal evidence that synchronized power fluctuations contribute to the formation of fMRI-based RSNs. Moreover, our findings demonstrate that the brain's intrinsic coupling at rest can be selectively modulated by choosing appropriate tACS signals, which could lead to new interventions for patients with altered rs-fMRI connectivity.SIGNIFICANCE STATEMENTResting state fMRI (rs-fMRI) has become an important tool to estimate brain connectivity. However, relatively little is known about how slow hemodynamic oscillations measured with fMRI relate to electrophysiological processes. It was suggested that slowly fluctuating power envelopes of electrophysiological signals synchronize across brain areas and that the topography of this activity is spatially correlated to resting state networks derived from rs-fMRI. Here we take a novel approach to address this problem and establish a causal link between the power fluctuations of electrophysiological signals and rs-fMRI via a new neuromodulation paradigm, which exploits these power synchronization mechanisms. These novel mechanistic insights bridge different scientific domains and are of broad interest to researchers in the fields of Medical Imaging, Neuroscience, Physiology, and Psychology.