Theta/Alpha band induced power

✉️ Mailing List Archive (read-only)
1

Hi all,
A quick question, in the whole brain induced power computation on the
example dataset (below), I wodered if the onset of the main peak is before
0ms.
http://martinos.org/mne/stable/auto_examples/time_frequency/plot_source_space_time_frequency.html#sphx-glr-auto-examples-time-frequency-plot-source-space-time-frequency-py

So, tested lower frequency bands (theta 4-8Hz and alpha 8-12Hz) the result
of which is attached. It seems as if the onset of the main peak of theta
starts around -50ms and alpha -30ms. Was wondering if someone might have
thoughts on this (as looks a bit odd to me)?

Thanks
Rezvan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.nmr.mgh.harvard.edu/pipermail/mne_analysis/attachments/20161122/02f8b13c/attachment-0001.html
-------------- next part --------------
A non-text attachment was scrubbed...
Name: sample_theta_alpha_TF.jpg
Type: image/jpeg
Size: 60795 bytes
Desc: not available
Url : http://mail.nmr.mgh.harvard.edu/pipermail/mne_analysis/attachments/20161122/02f8b13c/attachment-0001.jpg

2

hi,

2 things:
- you loose temporal resolution by looking at frequency content. In
other words, the peak will spread
in time (backward and forward)
- the mne code does not fix edge artifacts. So you need to crop the
output to get rid of it, depending
on the number of cycles you used.

HTH
Alex

3

Thanks Alex,
Yes, hadn't cropped this example.
But my point concerns the first reason you mentioned. However, I would have
thought that the peak spreading won't include prestimulus times this
obviously and made me think that perhaps this is due to the implementation.
Thanks a lot for reassuring that this looks normal to you.

Best
Rezvan