EEGLAB 2011 Mallorca

Jump to: navigation, search


Fourteenth EEGLAB Workshop

Mallorca, Spain, Sept. 22-25, 2011 preceding ICON


The 14th EEGLAB Workshop took place from Thursday, September 22 through Sunday, September 25, 2011 on the Spanish island of Mallorca preceding ICON XI. Participants in the first two parts of the Workshop were expected to bring laptops with Matlab installed so as to be able to participate in the practical sessions. The workshops consisted of three parts:

1. On Thursday, Sept. 22 there was a full-day Novice EEGLAB Workshop for those interested in learning the basics of using EEGLAB and independent component analysis (ICA) to analyze their EEG (or related) data.
2. Friday, Sept. 23 through noon on Sunday, Sept. 25, the first Advanced EEGLAB Workshop introduced and demonstrate dthe use of EEGLAB-linked tools for performing advanced analyses of EEG and related data, with detailed method expositions and practical exercises.
3. Sunday afternoon, Sept. 25, there was a (free) Open Discussion Session of evolving directions in EEG/ECoG research and free / open source data analysis, collection, and archival software. All interested were welcome to attend or participate in this discussion.

Costs and Registration

To reimburse travel expenses of Workshop faculty and facilities rental, costs for the workshop were as follows:

Part 1 - Novice EEGLAB Workshop (Sept. 22) 120€
Part 2 - Advanced EEGLAB Workshop (Sept. 23-25) 240€
Parts 1 & 2 - Novice & Advanced workshops (Sept. 22-25) 300€
Part 3 -- Open Discussion Session: Current directions in EEG research and open source software (no charge)

Registration is currently closed.

Relevant preparation in Matlab

The EEGLAB graphic interface is built on and provides easy ways to use the powerful Matlab scripting language. Exploiting the capabilities of EEGLAB for building macro commands and performing custom and automated processing requires the ability to manipulate EEGLAB data structures in Matlab. Because of time constrains, we will NOT provide an introduction to the Matlab language. Instead, users need to familiarize themselves with Matlab prior to the workshop.

Users of Matlab 7: we recommend running the following demos and reading the following help sections. After opening the Matlab desktop, select menu item "Help Demos" and run the following demos. Note that while the demo is running, you can retype the text (or copy it) to the main Matlab window:

Mathematics - Basic Matrix Operations
Mathematics - Matrix manipulations
Graphics - 2-D Plots
Programming - Manipulating Multidimentional arrays
Programming - Structures

In the Help Content, read and practice at least the following sections:

Getting Started - Matrices and Arrays - Matrices and Magic squares
Getting Started - Matrices and Arrays - Expressions
Getting Started - Matrices and Arrays - Working with Matrices
Getting Started - Graphics - Basic plotting functions
Getting Started - Programming - Flow Control
Getting Started - Programming - Other data structures
Getting Started - Programming - Scripts and Functions

Each section or demo (if read thoroughly) should take about 10 minutes. We encourage you to watch these demos and read these sections over several days. If you do not have access to the Matlab demos, here is a short online introduction to Matlab (recommended pages, 1 to 12)

IMPORTANT NOTE: Advanced tools and methods require script writing to perform analysis customized to your data and analysis goals. Script writing in Matlab is simple; the workshops will assume that you know at least the basics.

EEGLAB WIKI: refer to the EEGLAB wiki for additional help.

Some papers describing EEGLAB processing

Delorme, A., Makeig, S. EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. J Neurosci Methods. 2004; Mar 15; 134(1):9-21.

Makeig, S., Debener, S., Onton, J., Delorme, A. Mining event-related brain dynamics. Trends Cogn Sci. 2004; May; 8(5):204-10.

Jung, TP, Makeig, S, Westerfield, M, Townsend, J, Courchesne, E, Sejnowski, TJ. Analysis and visualizaion of single-trial event-related potentials. Human Brain Mapping. 2001; 14(3), 166-185.

Delorme, A., Sejnowski, T., Makeig, S. Improved rejection of artifacts from EEG data using high-order statistics and independent component analysis. Neuroimage. 2007; 34, 1443-1449.

Delorme, A., Palmer, J. Oostenveld, R., Onton, J., Makeig, S. Comparing results of algorithms implementing blind source separation of EEG data. unpublished manuscript.

Onton J, Delorme, A., Makeig, S. Frontal midline EEG dynamics during working memory. NeuroImage. 2005;27, 341-356

Workshop Agendas

Key: Lecture, Practicum, Break

I. Novice EEGLAB Workshop -- Thursday, Sept. 22

This workshop is designed for researchers who would like to learn how to process their EEG or related datasets using the ICA, time/frequency, and other tools provided in the EEGLAB software environment for Matlab ( The workshop instructor will be Julie Onton, Ph.D., long-time EEGLAB user and SCCN laboratory member. Scott Makeig, Director of the Swartz Center for Computational Neuroscience, UCSD, and originator of EEGLAB methods will give an introductory talk on evolving methods for analyzing EEG dynamics.

Novice workshop topics will include:

• Data import and preprocessing options
• Basic independent component analysis (ICA) theory and application
• Methods for imaging IC activations (ERPs, time/frequency, coherence)
• Equivalent dipole source localization of independent components
• Introduction to Matlab scripting using EEGLAB structure

Note: Because of time limitations, the Novice workshop will NOT include:

• New, more advanced toolboxes (NFT, SIFT, BCILAB, MPT)
• Mathematical derivations of the algorithms discussed

These topics will be covered in the Advanced EEGLAB workshop (described below).


08:30 -- 09:15am -- Mining event-related brain dynamics (Scott Makeig)(PDF)
09:15 -- 10:15am -- Introduction and getting started with EEGLAB (Julie Onton) (PDF)
Data import - Preprocessing tools and pipeline - Running ICA decomposition
10:15 -- 10:30 BREAK
10:30 -- 11:15 -- Evaluating ICA components (ICs) (PDF)
Apply ICA weights - IC scalp map interpretation - Basic IC evaluation - Identify artifact ICs
11:15 -- 12:15 -- IC analysis tools (PDF)
Removing ICs and back-projection - IC ERP envelope - IC ERP images - Time-frequency analysis - IC Event-related spectral perturbations (ERSPs) - IC Cross coherence
12:15 -- 13:30 LUNCH
13:30 -- 14:30 -- EEGLAB 'EEG' structure and basic Matlab scripting (PDF)
‘EEG’ structure overview - ‘EEG’ structure overview - Search EEG.event structure - Matlab functions - Converting from ‘pop’ functions to output functions
14:30 -- 15:15 – Equivalent dipole modeling (PDF)
Co-registration of electrodes with head model - Dipole fitting using Fieldtrip's dipfit function - Co-registration for 3D headplots
15:15 -- 15:30 BREAK
15:30 -- 16:30 -- Introduction to EEGLAB STUDY structure (PDF)
Build a STUDY - Precompute, precluster, and cluster ICs across subjects - Plot and edit STUDY clusters
15:30 -- 17:30 -- Advanced STUDY scripting (PDF)
Build a STUDY from the command line - STUDY structure overview - Cluster ERP image - Accessing raw STUDY data measures

II. Advanced EEGLAB Workshop -- Friday-Sunday, Sept. 23-25

This 2.5-day workshop will focus on emerging computational methods for EEG/ECoG analysis that have recently been made available within the EEGLAB environment as plug-in toolboxes. The lectures and practica will be more technically advanced than previous EEGLAB workshops. Participants will be expected to have at least passing familiarity with concepts such as linear regression, matrix inversion and other basic linear algebraic operations, Fourier transforms, and basic probability theory.

In addition, participants should be comfortable with using Matlab including performing the following operations using EEGLAB:

• Performing ICA decompositions and evaluating ICA component
• Obtaining equivalent dipole models of independent components using DIPFIT
• Performing time-frequency transforms and coherence analysis in EEGLAB
• Building an EEGLAB data STUDY

Interested participants who do not have the above background are strongly encouraged to study the relevant parts of the extensive Online EEGLAB Tutorial(, to go through the relevant video/slide lectures and practica in the Online EEGLAB Workshop (, and to consider attending the Novice EEGLAB Workshop (described above).

Advanced workshop topics will include:

• Applying adaptive-mixture ICA (Amica) to non-stationary EEG source dynamics
• Using Measure Projection analysis of multi-subject ICA-resolved EEG dynamics
• Applying statistical machine learning to EEG data analysis and Brain-Computer Interface design
• Analysis of oscillatory source network dynamics including Granger causality
• Forward and inverse modeling for EEG/ECoG source localization

Participants are expected to bring a laptop with Matlab and EEGLAB installed to work on (detailed instructions will be sent out before the workshop). Pairs of participants may also choose to share a laptop.

Preliminary Program -- Friday, Sept. 23

08:30 – 09:00 -- Welcome, introductions and brief overview (Scott Makeig)
Session A – Adaptive mixture independent component analysis (AMICA) decomposition (Jason Palmer)
This session will motivate, derive simply, and demonstrate the Adaptive Mixture ICA (Amica) algorithm of (Palmer et al., 2007) that finds more physiologically interpretable component processes in high-density EEG (or related) data and, further, detects and models changes in the spatial EEG source structure. Use of a set of tools and measures for interpreting the results of Amica decomposition will be demonstrated.
09:00 -- 09:45 ICA methods overview, with motivation for and derivation of Amica (PDF)
09:45 -- 10:45 Amica toolbox practicum
Please see wiki pages:
10:45 -- 11:00 BREAK
Session B -- Improving EEG source estimation using the Neuroelectromagnetic Forward Head Modeling Toolbox (NFT) (Zeynep Akalin Acar)
This session will give an overview of the forward head modeling problem and its approaches, followed by a demonstration of using NFT tools ( to derive forward head models from subject MR images and/or recorded electrode positions, and using such models to estimate source locations.
Both PDFs, supplementary functions and data for NFT
Supplementary functions and data ONLY, for practicum
11:00 -- 12:00 -- Forward head modeling overview(PDF)
12:00 -- 13:30 LUNCH
13:30 -- 14:30 -- NFT head modeling toolbox practicum(PDF)
Session C -- Comparing EEG dynamics across subjects using the Measure Projection Analysis (MPA) toolbox (Nima Bigdely Shamlo)
This session will address multi-subject EEG source analysis and will introduce a novel probabilistic method, Measure Projection Analysis (MPA, wiki is located here), to study and visualize the consistency of independent component localization and activities across subjects, groups, and/or conditions.
14:30 – 15:30 – Measure projection analysis theory (PDF)
15:30 -- 15:45 BREAK
15:45 -- 17:30 -- Measure projection analysis practicum (PDF)

Preliminary Program -- Saturday, Sept. 24

Session D -- Analyzing oscillatory EEG/ECoG source dynamics and interactions using the Source Information Flow Toolbox (SIFT) (Tim Mullen)
This session will survey contemporary computational approaches for analyzing oscillatory dynamics and synchronization/information flow in electrophysiological data. Topics will include the basic theory and practical issues surrounding estimation of Granger causality, Partial Directed Coherence and related information flow measures, phase-locking value, and phase-amplitude coupling. Participants will gain hands-on expertise in modeling and visualizing effective connectivity and synchronization between quasi-independent sources of EEG activity using the Source Information Flow Toolbox (SIFT) (
8:30 -- 10:00 am -- Analyzing Oscillatory Dynamics and Effective Connectivity (PDF)
Phase-amplitude coupling
Phase-locking value
Adaptive vector autoregressive models
Granger-causality, Partial Directed Coherence
Directed Transfer Functions, and related effective connectivity measures
10:00 -- 10:30 BREAK
10:30 --12:00 -- Using the Source Information Flow Toolbox: practicum | (PDF)
Getting started with SIFT
Vector Autoregressive Model Fitting and Validation
Connectivity Analysis (Granger Causality, Partial Coherence, Directed Transfer)
Computing reliability of network activity
Visualizing source connectivity across time, frequency and space
Group analysis using causal projection and other techniques
12:00 -- 13:30 LUNCH
Session E -- Statistical Learning Theory and Brain-Machine Interface Design (Christian Kothe)
This session will introduce the research areas of Brain-Computer Interface (or brain-machine interface) design and cognitive monitoring. Central concepts and challenges will be reviewed, and a selection of existing and emerging computational approaches will be examined. The presentation will focus on the use of oscillatory processes, with less focus on ERP-like phenomena. An introduction to creating BCILAB scripts and extensions will be included. In the practicum, participants will gain hands-on experience in designing, constructing, evaluating, visualizing and running BCI models using the open-source BCILAB software (
Slides from Christian's lecture.
13:30 -- 15:00 -- Statistical machine learning and Brain-Computer Interface design
Construction, learning and testing of predictive models (signal processing, machine learning, cross-validation)
Spatial filters for spectral BCI measures (CSP, ICA, DAL)
Capturing spectral structure (DFT, time-frequency representations)
Complexity control and regularization (sparsity norms, priors, nested cross-val)
Imposing prior knowledge in space, time and frequency (Talairach, DIPFIT, etc.)
Scaling to large / heterogenous data sources (across sessions, subjects, and/or task conditions).
115:00 -- 15:30 BREAK
15:30 -- 17:30 -- Using BCILAB to design and run BCI, cognitive monitoring, and neurofeedback experiments / applications (Download the PDF slide instructions)
Getting started with BCILAB
Learning, evaluating and visualizing spectral models: from simple to advanced
Offline and online data processing
BCILAB scripting, customization, and plug-in authoring

Preliminary Program -- Sunday, Sept. 25

Session F - Workshop Review, Results, and Discussion
09:00 -- 10:00 -- EEG Research: Current and Future Directions (Scott Makeig)
10:00 -- 10:30 -- Group practica results preparation
10:30 -- 10:45 BREAK
10:45 -- 12:00 -- Group practica results presentations and general discussion

III. EEGLAB Workshop Open Discussion Session -- Sunday, Sept. 25, 13:30 - 16:30

This will be a topic-focused discussion bringing together EEGLAB tool developers, advanced users, and computationally-inclined members of the cognitive neuroscience community as well as others interested to discuss the current state and future development of open-source EEG/ECoG analysis software within or connecting to the EEGLAB framework. This discussion will be open to all ICON participants -- any and all are welcome to attend and participate.

This discussion will take place at the Hotel Palas Atenea (second floor hall, by the pool)

13:30-16:30 -- Evolving Methods and Tools for EEG/ECoG Research (preliminary agenda)

1. Introduction -- Framing remarks and EEGLAB overview (Scott Makeig)
Designed EEGLAB uses and architecture
Use of the Matlab platform
Extending the EEGLAB plug-in mechanism
Building an online user community and collaborative network
Bridging to other open source tools and data collection systems
Multimodal data collection software -- ERICA and other efforts
Data and tools archive development -- HeadIT and other efforts
2. Advanced EEGLAB-compatible tools and directions (Panel)
Spatial source decomposition tools (Jason Palmer & TBN)
Forward and inverse head modeling tools (TBN)
Subject group analysis tools (Nima Bigdeley Shamlo)
Oscillatory dynamics analysis tools (Tim Mullen & TBN)
Information flow, synchronization, and causal network tools (Tim Mullen)
Brain-computer interface tools and directions (Christian Kothe)


Please help us by answering workshop survey questions