Need feedback on epochs and averaging

💬 Support & Discussions
, ,
1

Hi everyone!

I am super new studying and using EEG, I’m not sure if I’m doing this right so I went to this forum to have some feedbacks. I am using a publicly available dataset from Kaggle (M3CV dataset) according to their paper it is preprocessed and the files available are already cropped ranging from 1s - 4s with varying tasks.

I chose visual evoked tasks (VEP) as I want to further understand ERP. I came to this forum for feedback on the things I did regarding the dataset.

I did low-pass filtering (30Hz) on the already preprocessed as I think the dataset are still noisy, I cannot do high-pass because the files are in short duration

> (filter_length (8251) is longer than the signal (1000), distortion is likely. Reduce filter length or filter a longer signal.)

I did epoching as the files are still long in terms of ERP with this parameters:

    tmin = -0.3
    tmax = 0.5
    baseline = (None, 0)

image_2024-09-07_234403960
image_2024-09-07_234403960690×704 197 KB

This is one of the epochs and I am not sure if this undergo correct filtering and epoching and need your feedback.

image
image904×316 97 KB

Image above shows the VEP computed PSD.

image
I want to see the ERP components of all the channels combined and it output this.

I’m sorry if this is so long, I just need a feedback from someone regarding if this is correct or not. Thank you for your time!

Note: I’m using Google Colab and MNE version is 1.8.0.

2

The PSD looks very flat and smooth, it would be interesting to see the PSD of the raw data (which presumably is already preprocessed).

I don’t really understand what you mean by “I want to see the ERP components of all the channels combined” – do you mean you want to see the ERP for each channel separately? If so, this is a standard procedure that is described in several tutorials and examples, e.g. here: Estimating evoked responses — MNE 1.8.0 documentation. If you have more specific questions, please show what you did, what you expected, and what the actual outcome is.