| ICA | Summary of the signal processing functions of the eeglab toolbox |
| TOOLBOX CREDIT: | %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % MATLAB functions for psychophysiological data analysis %% % including use of enhanced versions of the %% % Independent Component Analysis (ICA) algorithm %% % of Bell & Sejnowski (1995). %% % By Scott Makeig, Colin Humphries, Sigurd Enghof & Tzyy-Ping Jung, %% % with contributions from Tony Bell, Delorme Arnaud, Martin McKeown, %% % Alex Dimitrov, Te-Won Lee, J-F Cardoso, Benjamin Blankertz et al. %% % email: scott@salk.edu %% % CNL/Salk Institute, 2000, Version 3.61 %% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
| GENERAL ELECTROPHYSIOLOGICAL DATA PROCESSING TOOLS: | Make plot axes pop up into zoomable windows on mouse click axcopy Find abs peak frames and amplitudes: abspeak Change reference from common to average: averef Simple block average data epochs: blockave Plot custom colorbar cbar Make a 2-D scalp field movie: eegmovie Frequency band filter data: eegfilt View continuous data traces: eegplotg) Average data epochs (with windowing options): erpave Display raw or smoothed single data epochs: erpimage) Re-align event-related epochs to given events: eventlock Plot one or more field maps on 3-D head model(s): headplot) Construct a movie of a moving field on a 3-D head model: headmovie Compute and view log power spectra of single data epochs: logspec Select chans,frames,epochs of concatenated data epochs: matsel Perform moving averaging on data: movav Plot a multichannel data epoch on a single axis: ploterp Perform principal component analysis (PCA) via SVD pcasvd() Perform nonlinear (post-PCA) rotations: varimax, promax, qrtimax View concatenated multichannel data epochs: plotdata View concatenated data epochs in topographic arrangement: plottopo Plot a data epoch with topoplots at selected time points: timtopo Change the data sampling rate: resample() Remove baseline means from data epochs: rmbase Regress out EEG data artifacts: rmart View a 2-D or 3-D scalp-field movie: seemovie Time/frequency (ERSP, ITC) averages of single-trial data: timef Iter-channel coherence averages of single-trial data: crossf View data scalp topography(s): topoplot) View images using scalp topography info: imagetopo() Convert Cartesian (x,y,z) channel locs to topoplot format: cart2topo Convert 2-D topoplot channel locs to 3-D headplot format: topo2sph Convert 2-D headplot channel locs to 2-D topoplot format: sph2topo |
| SPECIFIC ICA TOOLS: | Perform ica analysis using logistic infomax or extended-infomax runica Fast, compact Matlab MEX-file implementation of runica mexica() Fastest, most compact: system-call of binary runica binica Perform ica analysis using 2nd & 4th-order cumulants (Cardoso) jader Test ica algorithm accuracy, varying data parameters: testica Plot data and component envelopes: envproj envtopo Compute component activations: icaact Compute component variances on scalp: icavar Make activations all rms-positive: posact Compute component projections: icaproj Plot the data decomposition: plotproj -> chanproj Plot the data decomposition using plotopo(): projtopo Sort ica components by max projected latency and variance: compsort Sort ica components by mean projected variance only: varsort Compare ica weight matrices: matcorr -> matperm Plot selected time periods of component activations: tree View a projected ica component (time course plus topo map): compplot Squash or expand data into a PCA-defined subspace: pcsquash> expproj() -> pcexpand TOOLBOX DEMO icademo TOOLBOX tutorial tutorial |
| REFERENCES: | http://www.sccn.ucsd.edu/eeglab/icabib.html |
| Further information: | http://www.sccn.ucsd.edu/eeglab/icafaq.html |
| Please send news/bugs/fixes/suggestions to: | scott@salk.edu |