Hello,
I am using a public dataset (https://doc.ml.tu-berlin.de/simultaneous_EEG_NIRS/) with HbO, HbR, and HbT extracted from NIRS. The total channel is 108 (36 x 3).
Each NIRS-Hb contains three events (0,2,3 back). Now, I delete all events except 0-back from the time-series data of NIRS-Hb->delete group (event tab). So, I expect 108 columns with data of the 0-back event. But when I export the data by right-clicking the NIRS-Hb file, I found many more columns than 108 (more than 1000). I am afraid how is it export data? I want only 0-back data for HbO, HbR Please guide me in this regard. I execute the custom script to generate the Hbo, HbR data from the public dataset.
Also, can I remove the HbT channels from the channel file while computing analysis?
Attachment: (custom script to generate the NIRS file)
1.
IMP_Brainstorm_NIRX_NIRS.m (8.9 KB)
2.
Hi @ddaschakladar,
By doing this, you are only removing the other events, the data is continuous NIRS_Hb time series. This step is not necessary.
When exporting (to Matlab) you get a variable wit filed F which contains the data in a 2D matrix [nChannels, nSamples]
You can epoch the time series by importing it to Brainstorm. Right-click > Import in database, then check Use events and select the 0-back event indicating the Epoch time (signal duration with respect to the event, and check Create a separate folder ....
Each of the imported epochs will have the desired section of the time series, export it to Matlab and see the F field.
More info in this tutorial: https://neuroimage.usc.edu/brainstorm/Tutorials/Epoching#Import_in_database).
Lastly, if you want to see HbO, HbR only, just select the first two channels every three
Well, in the shared script adds HbT channels. You could edit it to adding those so the created files will have only 72 signals (36x2)
Best,
Raymundo