Image from EEG signals

Discussions
1

Hi!
I’m newbie here
I would like to visualize 2D/3D brain image form EEG signals
It is possible using brainstorm?
I hope an image could be save in x.jpg or x.tiff format because I want to do image processing technique after 2D/3D image produced.

Thanks in advance!

2

This is the main purpose of the software. All you need is your EEG recordings in binary or ASCII format, and the 3D positions of the electrodes (if you use a standard 10-20 or 10-10 cap, or an EGI net, we also provide default positions for them).

Have a look to the following page, you will have a better idea of the kind of images you can produce with Brainstorm. You can export any figure in all the main images formats.
http://neuroimage.usc.edu/brainstorm/Screenshots

3

Hi!
Thanks for the answer
I try all the tutorial.
I tried to import my EEG data in X.mat and I found it possible.
My data from 10-20 electrode but I only taken from Fp1 and Fp2.
I tried to edit channel 19 to 2 channel but an error ocur
’No. of channel file and data file do not match’
What should I do?

4

You recordings file and your channel file must have the same number of electrodes defined, and electrodes must be in the same order.

[B]How to check this ?
[/B]
For the data file: Right click > File > View .mat file.

  • the “F” size is [nChannels x nTime],
  • the “ChannelFlag” size must be [nChannels x 1]
    For your channel file: Right click > Edit channel file.
  • the number of lines displayed in the table must be the same as in the data file.

[B]How to fix if it if the sizes do not match ?
[/B]
Edit the recordings file:

  1. Right click on recordings > File > Export to Matlab (export to a variable)
  2. Add/remove rows to “F” and “ChannelFlag” fields
  3. Right click on recordings > File > Import from Matlab (use the same variable name)

Edit the channel file:

  • Right click > Align > Align EEG sensors (add/remove electrodes easily with the toolbar buttons)
  • You can also delete a channel with the Channel Editor (Right click > Edit channel file): select “(Delete)” in the sensor’s “Type” drop-down list.
5

Hi!
Thanks for the answers

Using my EEG data I able to display in time series and display on cortex.
I tried to understand color map on the cortex and the value pAm.I go through the tutorial, unfortunately i don’t understand what is pAm unit? please explain to me.

I expect the color map shows in the frequency band(eg:0Hz to40Hz). Can brainstorm do it?

I also tried to display in 2D layout, 2D Disc, 2D sensor cap and 3D sensor cap but this error occur

2D layout - Line 111
Bad value for axes property 'Clim’
Values must be increasing and non-Nan
2D disc - Line 47
2D sensor cap Not enough unique points specified
3D sensor cap

Thanks again!

6

[QUOTE=patungling;2234]I tried to understand color map on the cortex and the value pAm.I go through the tutorial, unfortunately i don’t understand what is pAm unit? please explain to me.[/QUOTE]

1 pA.m = 1 picoAmpere * meter = 10^-12 Am
This unit is used to represent the strength of the current sources located in the cortex.

[QUOTE=patungling;2234]I expect the color map shows in the frequency band(eg:0Hz to40Hz). Can brainstorm do it?[/QUOTE]

You can either use the “Filters” tab in the main window to apply a bandpass filter on your recordings (display purpose only), or filter your files through the “Processes” panel. Drag’n’drop your files in it, then click on “Run”.
A tutorial will be available soon to explain how to use this panel.

[QUOTE=patungling;2234]I also tried to display in 2D layout, 2D Disc, 2D sensor cap and 3D sensor cap but this error occur
2D layout - Line 111
Bad value for axes property 'Clim’
Values must be increasing and non-Nan
2D disc - Line 47
2D sensor cap Not enough unique points specified
3D sensor cap
[/QUOTE]

Can you describe a little more precisely what you did ?
Can you tell me what you get when you right-click > File > View .mat file, both on your recordings file and your channel file.

7

Thanks for your explanations and answers

My data were taken using 10-20 system and in my case I’m using 2 channnel only which is Fp1 and Fp2. Also I already edit 19 channel to 2 channel using ‘Edit channel file’. I include here my channel in .mat file

|- Channel: [1x2 struct]
| |- <struct #1>
| | |- Name: ‘Fp1’
| | |- Comment: ‘AVERAGE REF’
| | |- Type: ‘EEG’
| | |- Loc: [0.10233; 0.037026; 0.027562]
| | |- Orient: []
| | |- Weight: [0, 0]
| |- <struct #2>
| | |- Name: ‘Fp2’
| | |- Comment: ‘AVERAGE REF’
| | |- Type: ‘EEG’
| | |- Loc: [0.10278; -0.034804; 0.028911]
| | |- Orient: []
| | |- Weight: [0, 0]
|- SCS:
| |- System: ‘CTF’
| |- NAS: []
| |- LPA: []
| |- RPA: []
|- Comment: ‘10-20 (2)’

My recording in .mat format

|- Comment: ‘EEG/MAT’
|- ChannelFlag: [1; 1]
|- Time: [1x76801 single]
|- F: [2x76801 double]
|- Device: ‘Unknown’

I hope my explanation will get you clear picture of my problem. I am only able to visualize the data in time series, cortex, and MRI(3D)
Thanks

8

So you have recordings with 2 EEG channels, right? What is exactly you are trying to do?

With 2 channels only there is no hope to obtain a topographic plot of your data in 2D or 3D. All you can do at this point with BrainStorm is look at the time series and play with the filters, or am I misunderstanding the issue?

thanks!