What is the advantage of using Matlab?

Matlab allows incremental design of functions, so adding new features to a function can be easily accomplished while preserving backward compatibility.

What is the way to test significance across conditions?

To test significance across conditions or subjects, the authors either use parametrical tests or accumulated significance results from each subject.

Why is the density of the cortical connections mainly compact?

Because the density of cortical connections is weighted towards local connections (<<1 cm), particularly in the network of inhibitory cells that sustain cortical oscillations (Pauluis et al., 1999), the partially synchronous domains giving rise to EEG activity recorded on the scalp should be mainly compact – though the extent and density of these partially synchronous activities are not known.

What is the initial normalization step for linear inter-trial coherence?

For linear inter-trial coherence (ITLC), the initial normalization step is omitted: the vector sum is computed and then normalized by RMS power in the single-trial estimates.

What is the way to reduce memory usage in EEGLAB?

Though hopefully some future Matlab versions may allow the option of processing data in 32-bit floating-point format, the authors have taken care to address this issue in EEGLAB by including various options to minimize memory usage, such as constraining EEGLAB to work on a single dataset, or computing the ‘activation’ time courses of independent components only as needed.

What is the rationale for focusing data analysis on response averages alone?

as the authors have demonstrated recently, focusing data analysis on response averages alone ignores, first, event-related dynamics that do not appear in, or are poorly represented in response averages, and second, ignores ongoing EEG processes that may be partially time and phase-locked by experimental events, thereby contributing portions of response averages (Delorme et al., 2002; Makeig et al., 2002).

Why do ICA components account for more equal portions of the total signals than PCA components?

Because ICA does not attempt to maximize the variance of each component, ICA components may account for more equal portions of the total signals than PCA components.

What is the function that pops up when the user enters a parameter?

Called without (or in some cases with few) arguments, an interactive data-entry window pops up to allow input of additional parameters.

Why is it not reasonable to compute an unlimited amount of surrogate data?

Since it is not reasonable to compute an unlimited amount of surrogate data to estimate very low probability thresholds heuristically, the authors have implemented a method to fit the observed surrogate data distribution using a fourth order distribution fit (Ramberg et al., 1979).

Open AccessJournal Article10.1016/J.JNEUMETH.2003.10.009

EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis.

Q: What are the contributions mentioned in the paper "Eeglab: an open source toolbox for analysis of single-trial eeg dynamics including independent component analysis" ?

For example, EEGLAB this paper is a MATLAB tool for processing EEG data of any number of channels, including EEG data, channel and event information importing, data visualization ( scrolling, scalp map and dipole model plotting, plus multi-trial ERP-image plots ), preprocessing ( including artifact rejection, filtering, epoch selection, and averaging ), Independent Component Analysis ( ICA ) and time/frequency decomposition including channel and component cross-coherence supported by bootstrap statistical methods based on data resampling.

Q: Why are the projections of smaller principal components to the scalp resemble checkerboard maps?

because of the spatial orthogonality constraint, projections of smaller principal components to the scalp typically resemble checkerboard maps that could not represent coherent activity within a connected patch of cortex.

Q: How many time points are needed to obtain reliable decompositions?

The size of the weight matrix being the square of the number of channels, a number of time points at least a few times the square of the number of channels is usually needed to obtain reliable decompositions.

Q: What is the drawback of using linear filters?

One of drawback of using linear filters is that the signal roll-off at the cut-off frequency is weaker than what it would be using nonlinear filters.

Q: What is the initial normalization step for linear inter-trial coherence?

For linear inter-trial coherence (ITLC), the initial normalization step is omitted: the vector sum is computed and then normalized by RMS power in the single-trial estimates.

Q: What is the way to reduce memory usage in EEGLAB?

Though hopefully some future Matlab versions may allow the option of processing data in 32-bit floating-point format, the authors have taken care to address this issue in EEGLAB by including various options to minimize memory usage, such as constraining EEGLAB to work on a single dataset, or computing the ‘activation’ time courses of independent components only as needed.

Q: What is the rationale for focusing data analysis on response averages alone?

as the authors have demonstrated recently, focusing data analysis on response averages alone ignores, first, event-related dynamics that do not appear in, or are poorly represented in response averages, and second, ignores ongoing EEG processes that may be partially time and phase-locked by experimental events, thereby contributing portions of response averages (Delorme et al., 2002; Makeig et al., 2002).

Arnaud Delorme, +1 more

- 15 Mar 2004

- Journal of Neuroscience Methods

- Vol. 134, Iss: 1, pp 9-21

22.1K

TL;DR: EELAB as mentioned in this paper is a toolbox and graphic user interface for processing collections of single-trial and/or averaged EEG data of any number of channels, including EEG data, channel and event information importing, data visualization (scrolling, scalp map and dipole model plotting, plus multi-trial ERP-image plots), preprocessing (including artifact rejection, filtering, epoch selection, and averaging), Independent Component Analysis (ICA) and time/frequency decomposition including channel and component cross-coherence supported by bootstrap statistical methods based on data resampling.

About: This article is published in Journal of Neuroscience Methods. The article was published on 15 Mar 2004. and is currently open access. The article focuses on the topics: EEGLAB & Data visualization.

AI Agents for this Paper

Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps

Most frequently asked questions

1. What are the contributions mentioned in the paper "Eeglab: an open source toolbox for analysis of single-trial eeg dynamics including independent component analysis" ?

For example, EEGLAB this paper is a MATLAB tool for processing EEG data of any number of channels, including EEG data, channel and event information importing, data visualization ( scrolling, scalp map and dipole model plotting, plus multi-trial ERP-image plots ), preprocessing ( including artifact rejection, filtering, epoch selection, and averaging ), Independent Component Analysis ( ICA ) and time/frequency decomposition including channel and component cross-coherence supported by bootstrap statistical methods based on data resampling.

2. Why are the projections of smaller principal components to the scalp resemble checkerboard maps?

because of the spatial orthogonality constraint, projections of smaller principal components to the scalp typically resemble checkerboard maps that could not represent coherent activity within a connected patch of cortex.

3. How many time points are needed to obtain reliable decompositions?

The size of the weight matrix being the square of the number of channels, a number of time points at least a few times the square of the number of channels is usually needed to obtain reliable decompositions.

4. What is the drawback of using linear filters?

One of drawback of using linear filters is that the signal roll-off at the cut-off frequency is weaker than what it would be using nonlinear filters.

FIG. 4. Visualizing independent components. (A) Topographical 2-D scalp maps of the nine independent components (ICs) accounting for the most EEG variance of the 32 components returned by the ICA algorithm for the sample dataset. The component scalp map values returned by ICA are proportional to µV (scaling is distributed between the component maps and activity time courses). From its far-frontal scalp map, IC3 appears to account for eye movement artifacts. (B) The ‘Component Properties’ display for IC3 verifies that it accounts for eye artifacts since its activity spectrum is smoothly decreasing (bottom panel), and prominent eye movement artifacts appear in its activity ERP image (top right panel). By removing this and other eye movement components (not shown) from the dataset, the user can remove most evidence of eye movements from the data without removing other activity of interest (Jung et al., 2000).

FIG. 1. Sample EEGLAB session. Screen capture of an EEGLAB user session running under Linux. Users call EEGLAB functions from the main window menus (center) via ‘pop-up’ parameter selection windows (upper left). Warnings and data processing messages are shown in the Matlab command line window (lower left), which can also be used to call EEGLAB or other data processing functions directly.

FIG. 5. Evaluating independent component contributions. (A) An EEGLAB spectopo() plot showing the components accounting for the largest portions of 10-Hz activity at electrode POz (middle scalp map). The figure shows the power spectrum of the selected channel (top black trace), the activity spectra of the projection to that channel of each of the 32 components (lower traces), and the scalp power maps of the four largest-contributing components (4, 5, 7, 10). (B) An envtopo() plot showing the envelopes (i.e., the min and max values, over all channels, at each time point) of the five independent components making the largest potential contributions to the ERP. The black thick traces show the envelope of the (all channel) ERP data and the thin traces, the envelopes of the depicted component contributions to the ERP.

FIG. 2. Data scrolling. The EEGLAB scrolling data review function, eegplot(), allows the user to review and reject data by visual inspection. Here, five data epochs (separated by dashed lines) are plotted at 31 electrode sites (channel names on the left). Other channels in the dataset can be accessed using the vertical slider on the left. The arrow buttons (lower left) scroll horizontally through the data. The user may zoom in on a selected time range and/or electrode group and may change the plotting parameters using menu options (upper left). Values of the data point closest to the cursor are continuously displayed at the bottom of the display. In this example, two (central) data epochs have been automatically marked for rejection for out-of-bounds values set by the user in the EEGLAB data-rejection pop-up window (not shown). The rejection routine here highlights (in white) the channels containing the outlier values. The user can further mark (or unmark) data epochs for rejection by clicking on them. Pressing ‘Update Marks’ (lower right) saves the accumulated rejection markings.

FIG. 6. Time/frequency decompositions of independent component activities. Time/frequency decomposition was applied to the activities of two independent EEG components using sinusoidal wavelet transforms, 3 cycles in length at the lowest frequency (6 Hz), increasing linearly with frequency up to 9 cycles at the highest plotted frequency (35 Hz). Using this approach, it is possible to obtain reasonable time and frequency stability at all frequencies. (A-B) Event-related spectral perturbation (ERSP) plots showing mean changes in spectral power during the epoch, relative to a 1-s pre-stimulus baseline (plotted vertically on the left). Component IC4 shows a transient increase near 12 Hz centered at 500 ms, while component IC9 shows a power decrease in this range following 500 ms. (C-D) Phase cross-coherence (ERPCOH) magnitude and phase delay between the two components shown in panels A-B, zero-masked in regions in which cross-coherence magnitude was not significant (p>0.01). The components appear to become partially synchronized above 10 Hz (coherence ≤ 0.53) during the period 400 to 1000 ms with a phase offset near -120 degrees. Under the minimum phase assumption, this implies that high-alpha activity of IC9 tends to lead that of IC4 during this period by about 30 ms.

FIG. 3. ERP image construction. (A) ERP-image plots are constructed by color-coding (grey bars) potential variations occurring in single-trial epochs (black traces). (B) Vertically stacking thin color-coded horizontal bars, each representing a single trial in an event-related dataset, produces an ERP image. Here, trials were sorted vertically according to the subject reaction-time (right curving black trace) before applying a 10- epoch vertical moving average. The trace below the ERP image shows the ERP average of the imaged data epochs. The dot on the scalp map (top) indicates the scalp position of the channel whose data are imaged. (C) The erpimage() function automates several methods of sorting trials. Here, EEG phase in a given time/frequency window was used as the sorting variable. For each trial, a 10-Hz wavelet was applied to measure oscillatory activity in a 3-cycle window centered at time 0. Trials were then sorted (top to bottom) in order of their alpha band frequency phase values (-π to π) relative to stimulus onset and were displayed as an ERP image, again smoothed by a 10-trial moving average. The data were not otherwise filtered. The partial inter-trial phase coherence of the data following the stimulus onset is then visible as a change in the slope of the imaged activity wave fronts to near-vertical after 200 ms. Inter-trial phase coherence (bottom trace) shows that the distribution of alpha activity phase across trials is non-random (i.e., is partially phase-reset) between 200 and 450 ms (dotted line in lower trace shows p=0.01), resulting in same alpha activity appearing in the ERP average trace (top panel). The middle trace shows that mean changes in alpha power (in ‘dB’) did not change significantly (dotted lines) during the epochs. The baseline power level at the analysis frequency (25.9 dB, relative units) is indicated for possible comparison with other conditions.

Citations

•Journal Article•10.1162/JOCN.2009.21125

Theta phase synchrony and conscious target perception: Impact of intensive mental training

Heleen A. Slagter, +4 more

- 01 Aug 2009

- Journal of Cognitive Neuroscience

TL;DR: Elect EEG spectral analyses implicate theta phase locking in conscious target perception, and suggest that after mental training the cognitive system is more rapidly available to process new target information.

...read moreread less

148

Journal Article•10.1016/J.BIOPSYCH.2008.07.017

Independent component analysis reveals atypical electroencephalographic activity during visual perception in individuals with autism.

Elizabeth Milne, +5 more

- 01 Jan 2009

- Biological Psychiatry

TL;DR: Atypical processing is evident in individuals with ASD during perception of simple visual stimuli, and several processes that contributed to the average event related potential recorded at the scalp are identified.

...read moreread less

148

•Journal Article•10.1371/JOURNAL.PONE.0020422

A Collaborative Brain-Computer Interface for Improving Human Performance

Yijun Wang, +1 more

- 31 May 2011

- PLOS ONE

TL;DR: Results suggest that a collaborative BCI can effectively fuse brain activities of a group of people to improve the overall performance of natural human behavior.

...read moreread less

147

•Journal Article•10.3389/FPSYG.2011.00137

Modeling Single-Trial ERP Reveals Modulation of Bottom-Up Face Visual Processing by Top-Down Task Constraints (in Some Subjects).

Guillaume A. Rousselet, +3 more

- 23 Jun 2011

- Frontiers in Psychology

TL;DR: It is concluded that, during face processing, bottom-up sensitivity to phase noise can be modulated by top-down task constraints, in a broad window around the P2, at least in some subjects.

...read moreread less

147

...

Expand

References

•Journal Article•10.2307/4615733

A Simple Sequentially Rejective Multiple Test Procedure

Sture Holm

- 01 Jan 1979

- Scandinavian Journal of Statistics

TL;DR: In this paper, a simple and widely accepted multiple test procedure of the sequentially rejective type is presented, i.e. hypotheses are rejected one at a time until no further rejections can be done.

...read moreread less

23.4K

Journal Article•10.1162/NECO.1995.7.6.1129

An information-maximization approach to blind separation and blind deconvolution

Anthony J. Bell, +1 more

- 01 Nov 1995

- Neural Computation

TL;DR: It is suggested that information maximization provides a unifying framework for problems in "blind" signal processing and dependencies of information transfer on time delays are derived.

...read moreread less

9.9K

•Journal Article•10.1016/S0893-6080(00)00026-5

Independent component analysis: algorithms and applications

Aapo Hyvärinen, +1 more

- 01 May 2000

- Neural Networks

TL;DR: The basic theory and applications of ICA are presented, and the goal is to find a linear representation of non-Gaussian data so that the components are statistically independent, or as independent as possible.

...read moreread less

9.7K

•Journal Article•10.1109/PROC.1982.12433

Spectrum estimation and harmonic analysis

David J. Thomson

- 01 Sep 1982

TL;DR: In this article, a local eigenexpansion is proposed to estimate the spectrum of a stationary time series from a finite sample of the process, which is equivalent to using the weishted average of a series of direct-spectrum estimates based on orthogonal data windows to treat both bias and smoothing problems.

...read moreread less

4.4K

•Journal Article•10.1002/(SICI)1097-0193(1999)8:4<194::AID-HBM4>3.0.CO;2-C

Measuring phase synchrony in brain signals

Jean-Philippe Lachaux, +3 more

- 01 Jan 1999

- Human Brain Mapping

TL;DR: It is argued that whereas long‐scale effects do reflect cognitive processing, short‐scale synchronies are likely to be due to volume conduction, and ways to separate such conduction effects from true signal synchrony are discussed.

...read moreread less

3.9K

...

Expand

EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis.

Chat with Paper

AI Agents for this Paper

Most frequently asked questions

1. What are the contributions mentioned in the paper "Eeglab: an open source toolbox for analysis of single-trial eeg dynamics including independent component analysis" ?

2. Why are the projections of smaller principal components to the scalp resemble checkerboard maps?

3. How many time points are needed to obtain reliable decompositions?

4. What is the drawback of using linear filters?

5. What is the advantage of using Matlab?

6. What is the way to test significance across conditions?

7. Why is the density of the cortical connections mainly compact?

8. What is the initial normalization step for linear inter-trial coherence?

9. What is the way to reduce memory usage in EEGLAB?

10. What is the rationale for focusing data analysis on response averages alone?

11. Why do ICA components account for more equal portions of the total signals than PCA components?

12. What is the function that pops up when the user enters a parameter?

13. Why is it not reasonable to compute an unlimited amount of surrogate data?

Figures

Citations

Theta phase synchrony and conscious target perception: Impact of intensive mental training

Independent component analysis reveals atypical electroencephalographic activity during visual perception in individuals with autism.

A Collaborative Brain-Computer Interface for Improving Human Performance

Single-trial coupling of EEG and fMRI reveals the involvement of early anterior cingulate cortex activation in effortful decision making.

Modeling Single-Trial ERP Reveals Modulation of Bottom-Up Face Visual Processing by Top-Down Task Constraints (in Some Subjects).

References

A Simple Sequentially Rejective Multiple Test Procedure

An information-maximization approach to blind separation and blind deconvolution

Independent component analysis: algorithms and applications

Spectrum estimation and harmonic analysis

Measuring phase synchrony in brain signals

Related Papers (5)

FieldTrip: open source software for advanced analysis of MEG, EEG, and invasive electrophysiological data

Nonparametric statistical testing of EEG- and MEG-data

An information-maximization approach to blind separation and blind deconvolution

Updating P300: An Integrative Theory of P3a and P3b

The assessment and analysis of handedness: The Edinburgh inventory