-MNE version: 1.3.0
-Operating system: macOS-12.2.1
Hi there!
I am trying use the MNE BIDS pipeline to preprocess my data, but I have some questions/concerns about the output that I am getting.
For context:
My experiment involves a language paradigm where a context sentence is presented aloud and is then completed by a target word at the very end. The target word is either a plausible completion or is surprising (i.e., a mismatch) given that particular context. I am trying to look at the N400 ERP in response to that target word, and am expecting a higher amplitude in the mismatch condition relative to the unsurprising one.
The following screenshots are from the HTML report file (averaged across all subjects, N = 53) that results when I run the pipeline for conditions with which I should definitely see an N400. The contrast here is between the “el” (low entropy, unsurprising) and “elmm” (low entropy mismatch, surprising) conditions, and each includes about 1000 trials across all subjects.
And this screenshot comes from a script that I’ve created based on the tutorial here.
My main concern is that something is going wrong in the preprocessing stage, because the results that I get do not resemble the typical N400, neither in the median graph nor the GFP one. There is clearly a difference between conditions, but it looks like something is wrong with the shape, so I am thinking there could be some filtering issue(?). The time course plots also look strange, as there is an odd divergence as time goes on.
Moreover, I am trying to look only at centro-parietal electrodes, and have defined the following in my config file
analyze_channels = ['Cz', 'CPz', 'CP1', 'CP2', 'P1', 'P2', 'Pz']
but, it appears that still all of the electrodes are being processed. I have compared the outputs that I get when I use “all” the electrodes versus when I specify only these seven electrodes, but the results are exactly the same, so it appears that something is malfunctioning in that selection process or in my code.
Other possibly useful information:
I have defined Fp1 and Fp2 as EOG channels.
The EEG reference is the average across all channels.
The ECG channel is being dropped.
The frequency filters are 0.1 (low), 40 (high), and 50 (notch).
ICA is being applied for artifact removal.
Most other parameters have been left as they are in the default config file.
I was wondering if anyone can please help me figure out what is going wrong with my output, and/or how it can be fixed. Thank you in advance for any and all help! It is so much appreciated! 