# Recommended preprocessing pipeline for 32-channel eyes-closed resting EEG before PSD analysis

**URL:** <https://mne.discourse.group/t/recommended-preprocessing-pipeline-for-32-channel-eyes-closed-resting-eeg-before-psd-analysis/11866>\
**Category:** Support & Discussions\
**Tags:** preprocessing, ica\
**Created:** [June 17, 2026, 11:13am UTC](https://mne.discourse.group/t/recommended-preprocessing-pipeline-for-32-channel-eyes-closed-resting-eeg-before-psd-analysis/11866 "2026-06-17T11:13:32Z")\
**Posts on this page:** 1\
**Showing post:** 4

<div class="post-metadata">

**Author:** ![Vicky\_Kachera](https://yyz2.discourse-cdn.com/free1/user_avatar/mne.discourse.group/vicky_kachera/32/4790_2.png) [@Vicky\_Kachera](https://mne.discourse.group/u/Vicky_Kachera)\
**Post date:** [June 18, 2026, 8:18am UTC](https://mne.discourse.group/t/recommended-preprocessing-pipeline-for-32-channel-eyes-closed-resting-eeg-before-psd-analysis/11866/4 "2026-06-18T08:18:43Z")

</div>

Thank you for the clarification.

Just to confirm, is the following ordering what you would recommend?

1. Mark/remove bad channels
2. Apply average reference
3. Run and apply ICA
4. Interpolate bad channels

I noticed that another MNE discussion @richard suggested that average referencing should be done after ICA (See [here](https://mne.discourse.group/t/order-to-pre-process-data/4761/2)), since doing it before ICA can smear artifacts across channels. I wanted to confirm which ordering you recommend for this case.

Thank you.

---

_[View the full topic](https://mne.discourse.group/t/recommended-preprocessing-pipeline-for-32-channel-eyes-closed-resting-eeg-before-psd-analysis/11866)._