What should I do if my raw data without projects information?

đź’¬ Support & Discussions
,
1

If the raw data without projects info, does that mean I cannot apply the SSP correction to reduce the system noise?

When I write the following code in Python, the output was an empty value.

>>> print(raw.info["projs"])
[]

How could I write a projs info in the raw data?

2

You can create SSP projectors as described in this tutorial:

https://mne.tools/stable/auto_tutorials/preprocessing/50_artifact_correction_ssp.html#example-eog-and-ecg-artifact-repair

Note that for EEG data, ICA may be better-suited than SSP.

3

Thanks a lot. I want to have a try to decrease noise using SSP as the tutorial said.

I cannot find the tutorials about how to create SSP projectors on this link. It just described how to create an empty_room projs. But when I tried to compare the system projs with the empty_room projs as the tutorials described, I still have no idea about creating a system projs.

4

The “system projs” are projectors that are specific to the MEG machine. Not all MEG machines have them. Their main purpose is to make the raw data look nicer during data acquisition, on the acquisition computer. If you have an empty-room recording then you don’t need them; for offline analysis, it is better to compute your own environmental noise projectors anyway.

Ideally the empty-room projectors should be based on an empty-room recording taken on the same day as the data were collected (ideally just before or just after the data recording session). That will give you the most accurate picture of the environmental noise, and thus will do the best job of removing that noise.

If you don’t have system projectors and you didn’t make an empty-room recording, there isn’t really a way to add environmental-noise SSP projectors I don’t think. Try SSS or ICA instead.

5

Thank you so much for the detailed explanation.