steps to clean noisy data and artifacts

If you have a question or issue with MNE-Python, please include the following info:

  • MNE-Python version: 0.23
  • operating system: Windows 10

Greeting everyone,

I hope all is well with you.

It’s been more than two months I’m having difficulties cleaning my dataset. I have done filtering and ICA but I’m not sure if I can do better cleaning. I’m planning to do the next steps, so correct me if you feel I’m wrong or if you have a better idea. I’m planning to do band-pass filtering and then ICA and then find a method that detects artefacts I couldn’t repair and remove them or I do it manually. The dataset I’m using contains lots of noise, so experts opinions are much appreciated.

Hello, what exactly is your question?

Hello @richard ,

Thank you for your response.

My question is when you look at the photos attached, do you think the steps and the order of the steps I’m planning to do is the best way to fix this type of EEG signal? or is there a better way and better methods?

Kind regards,

Hi @ibra332,
I think the scaling you use when plotting EEG is too high and will make it difficult to spot artifacts - most of the channels get trimmed because of that.
But as a general remark: before running ICA, I would reject nonstereotypical noisy data (data without clear topography, especialy in the lower frequencies), because their presence will very likely distort ICA results. Other than that I think that your plan is ok.

1 Like

Hello @mmagnuski,

thank you for the advice.