BEst - Brain Entropy in space and time

This toolbox implements the "maximum-entropy on the mean" (MEM) technique for solving MEG, EEG and MEG/EEG source localizations. Along with the classical MEM (Amblard et al. 2004), we exploit wavelet decompositions of the data in order to target specifically oscillatory patterns (wMEM) or synchronous cortical sources (rMEM). Explanation of cMEM, rMEM, wMEM in 2-3 sentences here.

BEst, for "Brain Entropy in space and time", is the result of a collaborative work between École de Technologie Supérieure and the department of biomedical enginnering at McGill university (Montreal, Canada). The present tutorial shows how to use MEM on MEG data from

1) Simulations: 2 examples of realistic simulations of spikes

2) Inter-ictal epileptic discharges (IEDs): Average of 171 spikes

3) Median Nerve Stimulation (MNS) data: Average of 98 trials from Right hand responses

4) Sleep spindles: 5 spontaneous events

These datasets can be downloaded from here. For each dataset the anatomy and gain matrices are provided. Below we will illustrate the application of the three MEM methods on MNS dataset (BEst_tutorial_median_nerve).

See below for more info on the BEst license and the associated publications.

MEM: preliminary steps

We need one gain matrix for each recording modality considered for source localization. These matrices can be computed within Brainstorm using either, spherical or realistic (recommended) head models. See the tutorial on Head models for details. Note that unlike the other methods proposed in Brainstorm, the MEM does not require prior calculation of a noise covariance matrix. This matrix can be estimated using a physiological baselinesignal that can either be part of the data to be localized or a different piece of data. We can also use empty room noise recordings if available to estimate this matrix.

1. Uncompress downloaded dataset in a folder

2. Import the data in brainstorm by clicking: File > Load protocol > Load from folder > browse the protocol (BEst_tutorial_median_nerve)

3. Go to the data view mode (second button on top of the database explorer)

4. Right-click on the data average > Compute sources > BEst. Hit 'Ok' Note: The MEM option only appears in expert mode. Make sure it is toggled from the buttom at the bottom left. It is recommended that you customize the comment field so as to avoid confusion when performing multiple localizations for a given dataset.

5. Select the MEM method (cMEM: time-series representation, wMEM: time-scale representation, or rMEM: wavelet representation). Define the data time window and the baseline data while keeping other parameters at default values. Then hit 'Ok'.

cMEM

1. Selection of a specific time window may be useful for long data epochs (default: localize whole data window). Suggested time window for this dataset: 0.010s-0.030s.
2. Three choices are available for baseline definition.

   • Default: the panel looks for a dataset called 'baseline' within the same study as the data to be localized. When not found, this option is disabled.

   • Within data: the baseline is a time window within the recordings.

   • Import: the user can import baseline data from a file using brainstorm import tools In all cases, the user can define a time window for the baseline

3. Hitting 'Ok' will launch the cMEM method. At the end, a new results file appears in the database under the appropriate dataset. Simply double-click it to visualize.

Parameters of cMEM:

• Clustering: Dynamic (Link), Stable in time (Link), MSP scores threshold (Link - Arbitrary/FDR), neighborhood order (Link), spatial smoothing (Link).
• Data definition: Time window (Link), baseline (Link).
• Expert: (Link to a page explaining the expert parameters).

wMEM

1. Selection of a specific time window may be useful to localize specific time-frequency components of the recordings (default: localize whole data window). Suggested time window for this dataset: -1s to +1s.
2. Three choices are available for baseline definition.

o Default: the panel looks for a dataset called 'baseline' within the same study as the data to be localized. When not found, this option is disabled.

o Within data: the baseline is a time window within the recordings.

o Import: the user can import baseline data from a file using brainstorm import tools In all cases, the user can define a time window for the baseline

1. Selection of scales to be localized….. Suggested scales

2. Hitting 'Ok' will launch the wMEM method. At the end, a new results file appears in the database under the appropriate dataset. Simply double-click it to visualize.

Parameters of wMEM:

• Clustering: Dynamic (Link), wavelet-adaptive (Link), MSP scores threshold (Link - FDR), neighborhood order (Link), spatial smoothing (Link).
• Data definition: Time window (Link), baseline (Link).
• Expert: (Link to a page explaining the expert parameters).

The most innovative aspect of the BEst toolbox is the application of the entropic source localization framework (MEM) on wavelet coefficients. The aim of this approach is to localize neural sources responsible for a specific activity such as oscillations or synchrony. We consider two kinds of wavelet transforms, continuous (rMEM, Zerouali et al. 2011) and discrete (wMEM, Lina et al. 2012), which make explicit the synchronous and oscillatory components, respectively.

• wMEM: The data is decomposed following a discrete multiresolution scheme using real Daubeachies wavelets. The time-scale plane can then be thresholded to keep only significant coefficients, up to a fraction of total signal power. The MEM then localizes the sources of each coefficient, thus yielding wavelet coefficients for each source. As we keep only significant oscillatory modes from the time-scale planes, this method efficiently targets the oscillating neural sources. Interestingly, real Daubechies wavelets have a stable inverse transform, which allows us to retrieve real time course for each source.

• rMEM: The signal is decomposed into continuous time-frequency plane using analytic Morse wavelets. We then extract multivariate ridges from these planes, and extract the complex coefficients along the ridge lines. The sources of these complex signals are then localized using the MEM. As the ridges reflect synchronous signal components, the rMEM targets synchronous sources. The following figure illustrates the results obtained by the MEM solver and the wavelet-based MEM on Brainstorm tutorial data.

rMEM

1. Selection of a specific time window may be useful for long data epochs (default: localize whole data window). Suggested time window for this dataset: 0.010s-0.030s.
2. Three choices are available for baseline definition.

   • Default: the panel looks for a dataset called 'baseline' within the same study as the data to be localized. When not found, this option is disabled.

   • Within data: the baseline is a time window within the recordings.

   • Import: the user can import baseline data from a file using brainstorm import tools In all cases, the user can define a time window for the baseline

3. Specify the frequency band for ridge computation. Suggested values for this dataset: 15-30Hz
4. Definition of the minimum duration of the synchrony. Suggested value for this dataset: 20ms
5. Hitting 'Ok' will launch the rMEM method. At the end, a new results file appears in the database under the appropriate dataset. Simply double-click it to visualize.

Parameters of rMEM:

• Clustering: Dynamic (Link), wavelet-adaptive (Link), MSP scores threshold (Link - Arbitrary/FDR), neighborhood order (Link), spatial smoothing (Link).
• Data definition: Time window (Link), baseline (Link).
• Expert: (Link to a page explaining the expert parameters).

Results from the other datasets

• cMEM

o Simulations

o Epileptic spikes

• wMEM

o Simulations

o Sleep spindles

• rMEM

o Simulations