# Channel interpolation and frequency filtering in MNE-BIDS-Pipeline

**URL:** <https://mne.discourse.group/t/channel-interpolation-and-frequency-filtering-in-mne-bids-pipeline/6583>\
**Category:** Support & Discussions\
**Tags:** preprocessing, eeg, frequency-filter, mne-bids-pipeline\
**Created:** [March 23, 2023, 8:52am UTC](https://mne.discourse.group/t/channel-interpolation-and-frequency-filtering-in-mne-bids-pipeline/6583 "2023-03-23T08:52:00Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![fraenni](https://avatars.discourse-cdn.com/v4/letter/f/e56c9b/32.png) [@fraenni](https://mne.discourse.group/u/fraenni)\
**Post date:** [March 23, 2023, 8:52am UTC](https://mne.discourse.group/t/channel-interpolation-and-frequency-filtering-in-mne-bids-pipeline/6583/1 "2023-03-23T08:52:00Z")

</div>

Hello,  
thanks for the new features in the pipeline. I am using the pipeline for ONLY eeg-data

I have two general questions:

1. Is it possible to define bad EEG channels and interpolate them in a later state? Is it possible to use the meg-functions for that?

2. I am interested in the gamma frequency range (high+low gamma). but we are dealing strong line frequency in our data. Would you have any recommendations? As only high and low filter is possible, we seem to lose everything \>45Hz.

Thank you, Franziska

---

<div class="post-metadata">

**Author:** ![richard](https://yyz2.discourse-cdn.com/free1/user_avatar/mne.discourse.group/richard/32/15_2.png) [@richard](https://mne.discourse.group/u/richard)\
**Post date:** [March 23, 2023, 9:54am UTC](https://mne.discourse.group/t/channel-interpolation-and-frequency-filtering-in-mne-bids-pipeline/6583/2 "2023-03-23T09:54:49Z")

</div>

Hello,

> [@fraenni](#):
>
> - Is it possible to define bad EEG channels and interpolate them in a later state? Is it possible to use the meg-functions for that?

the bads should be marked as such in the input BIDS dataset. Interpolation happens by default [when calculating the grand average](https://mne.tools/mne-bids-pipeline/1.1/settings/sensor/group_level.html#mne_bids_pipeline._config.interpolate_bads_grand_average).

> [@fraenni](#):
>
> As only high and low filter is possible, we seem to lose everything \>45Hz.

I’m not sure I understand your question, but we also support a [notch filter](https://mne.tools/mne-bids-pipeline/1.1/settings/preprocessing/filter.html#mne_bids_pipeline._config.notch_freq) now, maybe this could be of help?

Best wishes,  
Richard

---

<div class="post-metadata">

**Author:** ![fraenni](https://avatars.discourse-cdn.com/v4/letter/f/e56c9b/32.png) [@fraenni](https://mne.discourse.group/u/fraenni)\
**Post date:** [March 24, 2023, 9:57am UTC](https://mne.discourse.group/t/channel-interpolation-and-frequency-filtering-in-mne-bids-pipeline/6583/3 "2023-03-24T09:57:11Z")

</div>

Perfect, thanks, that sorted it!

---

<div class="post-metadata">

**Author:** ![lightwind](https://avatars.discourse-cdn.com/v4/letter/l/e19adc/32.png) [@lightwind](https://mne.discourse.group/u/lightwind)\
**Post date:** [June 10, 2026, 2:32pm UTC](https://mne.discourse.group/t/channel-interpolation-and-frequency-filtering-in-mne-bids-pipeline/6583/4 "2026-06-10T14:32:48Z")

</div>

Hello,

I’ve been marking channels as ‘bad’ in the BIDS dataset, but I want those channels to be interpolated in the MNE-BIDS-Pipeline, rather than excluded. Is there a way to specifiy this behavior?

Thank you!