remove heat beats artifacts using SSP

Mailing List Archive (read-only)
1

Dear MNE-users,

I would like to use SSP to remove heart beats artifacts. After calculating
the projections separately for different runs within one subject, I tried
to epoch the data and then combine the epochs from different runs based on
their conditions. Then I got an error message showing that I have to use
the same projections when combining them together.

My first question is: how can I calculate the ERFs of each subject with all
the runs combined?

I'd also like to remove the epochs that are contaminated by eye blinks.
Then the second question is: Should I do the SSP first or should I remove
the EOG artifacts first?

Thanks for your help!

Best,
Lin
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.nmr.mgh.harvard.edu/pipermail/mne_analysis/attachments/20161115/d430ab18/attachment.html

2

hi Lin,

I would like to use SSP to remove heart beats artifacts. After calculating
the projections separately for different runs within one subject, I tried to
epoch the data and then combine the epochs from different runs based on
their conditions. Then I got an error message showing that I have to use the
same projections when combining them together.

My first question is: how can I calculate the ERFs of each subject with all
the runs combined?

http://martinos.org/mne/stable/generated/mne.concatenate_raws.html#mne.concatenate_raws

I'd also like to remove the epochs that are contaminated by eye blinks.

use reject parameter in Epochs

Then
the second question is: Should I do the SSP first or should I remove the EOG
artifacts first?

remove the EOG artifacts first.

HTH
Alex

3

Hi Alex,

Thanks for your reply.

Do you mean to treat the epochs as raw data and use 'mne.concatenate_raws'
to concatenate them? If I do that, I got this error: 'Epochs' object has no
attribute 'append'.

Or did I misunderstand anything here?

Thanks!
Lin

4

please read the doc...

to concatenate Epochs use:

http://martinos.org/mne/stable/generated/mne.concatenate_epochs.html#mne.concatenate_epochs

don't pass Epochs to a function that is called concatenate_raws...

A

5

Hi Alex,

I used 'all_epochs.append(epochs)' to concatenate all epochs of different
runs, as shown in the Biomag demo data (
http://mne-tools.github.io/mne-biomag-group-demo/auto_scripts/05-make_epochs.html#sphx-glr-auto-scripts-05-make-epochs-py
).

However, I got an error because I calculated SSP separately for different
runs, and the error is caused by the different projections obtained from
different runs.

Thanks for your help!

Lin

6

However, I got an error because I calculated SSP separately for different
runs, and the error is caused by the different projections obtained from
different runs.

you should then compute the SSP on the concatenated raw files
and apply them on your Epochs or better recompute the Epochs
from the concatenated raws.

HTH
Alex

7

Got it. Thanks! :slight_smile:
Lin