Peak-to-peak bad-channel detection has stopped working

Discussions
,
1

Hi,

I'm not sure when this started happening, but the peak-to-peak bad-channel detection function is no longer working. I've tried testing different amplitude thresholds, but when I run the command, no bad channels are detected, even though there are clearly several.

What could be the problem? How can I fix it?

Thank you so much!

2

Hi @danto034, can you provide more information?

  • Brainstorm version
  • Is it applied on raw or imported data?
3

Hi,

I’m using Brainstorm Version 23-Jul-2025. The correction was applied on the raw file, after I ran filters (band pass filter) and DC-offset.

4

When the process is run on raw data, it creates bad events (for specific channel or all channels), it does not label the entire channel as bad.

image

Once the raw data is imported:

  1. If there were bad events for a channel, that channel labelled as bad (for all the trials that include the bad event)

  2. If there were bad events for all channels, the entire trial is set as bad.

Is this what is happening on your side?

5

On my side, I usually select “reject the entire segments/trials”, but I’ve tried both options and it still detects zero bad events after import, so no channels are labeled bad and no trials are rejected. I tested multiple amplitude thresholds on several datasets with the same result.

6

Can you share some screenshots?

  1. Time series of parts that should be detected as artifactual, and
  2. The parameters that are being used in the process