MathWorks is sponsoring a session at SfN that will highlight the Neural Decoding Toolbox. The session will be on Sunday afternoon from 2-4pm at Booth #170. Please come by any time during that window to ask questions about the toolbox, give me some feedback, or just to say hello.
Version 1.0.4 of the Neural Decoding Toolbox has been released. This version has a few small enhancements including:
1) The ability to combine data from all time bins when creating a null distribution by setting the field collapse_all_times_when_estimating_pvals = 1 in the plot_standard_results_object (or in the pvalue_object). This allows you to calculate p-values much more efficiently (i.e. by running a decoding analysis 5 to 10 times instead of running it 200 times).
2) The ability to stop the decoding procedure if the results have converged to a stable value by setting the field stop_resample_runs_only_when_specfic_results_have_converged = 1 in the standard_resample_CV object. This can reduce the number of resample runs used by stopping as soon as the results converge (e.g., from 50 resample runs as the default to 20 resample runs).
3) The toolbox is now compatible with Octave which is a free open source language that is very similar to MATLAB.
A new tutorial has also been added to the www.readout.info website that describes how to apply decoding analyses to MEG/EEG data and an additional package of functions have been added to help with MEG/EEG decoding.
Please let me know if you have any feedback about the toolbox!
Version 1.0 of the Neural Decoding Toolbox has been released. The website is more or less complete, and an article about the toolbox has been published in Frontiers in Neuroinformatics.
This website is still under construction, check back in a couple of months and hopefully it will be completed by then!