Missing triggers and data

[galit.shaham1]

Hi,
We have a strange problem with an experiment we ran. We recorded EMG using 4 flat electrodes + CMS-DRL. The experiment took ~50 min. and had 192 trials. However, when analyzing the data in Analyzer, after ~150 trials, I stop seeing my regular triggers and data. Instead, we get a lot of weird triggers: Start Epoch, Speed Mode, Response triggeres etc. and some kind of signal that seems to be recorded in electrodes different than the ones we use. This happened in all the subjects and always in the last ~40 trials.
Attached is an example of how the raw data looks like in Analyzer at the time it happens. Notice how the regular data in Exg1-4 changes and the many many triggers starting to appear. Also note the blue part in the right side of the navigation bar.
I would mention that the trigger codes we sent were 254 and 255 as start and pause codes and stimuli codes 30, 40, 50, 101-116 and 201-216. We sent zeros after each code. The experiment was ran with E-Prime and from there everything seems to be normal and all the trials were run as expected.
Any idea what this is about? Is there any way I can retrieve my missing data??
Thanks for your help!

Galit.

[Coen]

Do you see the same data corruption in ActiView during recording ?

Do you see the same result in alternative BDF file readers such as https://www.biosemi.com/download/ToolsF ... er-808.zip and http://www.teuniz.net/edfbrowser/ ?

Best regards, Coen (BioSemi)

[galit.shaham1]

Dear Coen,
Thanks for your reply.
During recording, the data seemed normal the whole time. I did not pay attention to the triggers.
I checked in EDFbrowser and the data and the timing of the triggers look normal (though I'm not sure how to see the trigger numbers I used. I just see Trigger Input 1/2/3... etc.).
Any idea what might cause the problem in Analyzer or how could I fix it?
Thanks,
Galit.

[Coen]

All I know about Analyzer is that it was updated with a total BDF blockade as far back as 2003 (after working perfectly between 2000 and 2003). Since then, all relevant publications from the vendor list the statement "BioSemi not supported". Not surprisingly, most of the BioSemi customers were quite frustrated by this unexpected change in policy, and choose to move on to other review software that does offer reliable BDF support (e.g. BESA, EMSE, EEGlab, Fieldtrip, etc.). Consequently, I do not have any recent feedback about Analyzer/BDF compatibility problems (actually, I am surprised that you can read BDF files at all in Analyzer). Because of this chequered support history, BioSemi does not recommend to use Analyzer software.

Regarding you specific problem, I would first test whether it is related to file length. Use the file cropper (https://www.biosemi.com/download/ToolsF ... pper86.zip) to split the files is several smaller parts. Then test whether the sections after 150 trials are displayed without errors.

Best regards, Coen (BioSemi)

[galit.shaham1]

Dear Coen,
What I hear from you about Analyzer is very disturbing. However, I am using Biosemi and Analyzer since 2010 and never experienced such a problem. Also, according to their web page they do support Biosemi: http://www.brainproducts.com/productdet ... d=17&tab=2 .

Regarding my specific problem, I contacted Analyzer support and they acknowledged that this was some problem related to the file size. Indeed, when I reduce the number of channels (in EDFBrowser) the file is read correctly. However, they also brought to my attention that even though I am using only 4 EMG electrodes, the file does contain 256 EEG channels and additional channels. These channels are obviously flat, but "recorded" in 2048 Hz, they make the file extremely big. My questions are:
1. Why were these channels recorded in the first place? What did I need to do for them not be recorded?
2. I would like to reduce the unwanted channels to all my files. Doing it one by one as I did with EDFBrowser would be very time consuming. Is there a way to reduce the unwanted channels for all my files at once?

Thanks for your help,

Galit.

[Coen]

> 1. Why were these channels recorded in the first place? What did I need to do for them not be recorded?

Use the proper setting in the ActiView file dialog. I assume you use 4 EX channels for EMG.

1) Select to only display the 4 EMG channels in the monopolar display (use Channel selection option "free choice")
2) Start a new BDF file, and in the file dialog, select for "Save Subset" the option "Monopolar Displyed Channels", uncheck all 4 extra save options.
3) During saving, check that in the right sidebar the "Nr Chan Saved" displays the numbers "4+1" (4 EMG channels and the Trigger/Status channel
4) You can save all the above settings in a new CFG file (use "About ActiView" tabpage), load this CFG file for your future 4 channel recordings.

> 2. I would like to reduce the unwanted channels to all my files. Doing it one by one as I did with EDFBrowser would be very time consuming. Is there a way to reduce the unwanted channels for all my files at once?

An alternative is to use the "Reducer" tool (https://www.biosemi.com/download/ToolsF ... ucer86.zip), but you still have to process each file separately. However, the tool is open-source LabVIEW, so you can in principle reprogram it to handle multiple files.

Best regards, Coen (BioSemi)

[galit.shaham1]

Thanks!