Amplitude of sensor level average

Hi,

I have looked in the manual, cookbook and discussion archives but can't find out how to measure the amplitude of an evoked average in mne_browse_raw. I know the latency comes up when you select a time point of interest, however I am unsure of how to measure the amplitude of the peak of the response, if it is indeed possible.

Any help would be greatly appreciated.

Thanks
Lauren

Dear Lauren,

I don't think you can do this with mne_browse_raw.
I'd give a try to mne_analyze but I personally do this with mne-python
or matlab.

Best,
Alex

You cannot measure the amplitude with mne_browse_raw. It is easy to do in either Matlab or Python, just a SMOP. It can be challenging to find a definition of a peak which can be turned into automatic detection, though.

However, it is possible to measure amplitudes from raw data in mne_browse_raw. How about turning the ERF/ERP file into a short raw file in Matlab or Python and then measuring amplitudes from this file with mne_browse_raw?

- Matti

Hi both,

Thanks for your suggestions, I will let you know how I get on.

Cheers
Lauren

Dear Lauren,

for your interest, we just added simple peak finding methods in
MNE-Python for evoked and stc objects.

https://github.com/mne-tools/mne-python/pull/984#issuecomment-30583928

Those allow you to get the channel name or the vertex ID associated
with the maximum response as well as the latency, either in seconds or
as time index pointing at the related position in the times vector.

Here two examples.

https://github.com/mne-tools/mne-python/blob/c8633989fc3429d7dc1a4635fe05d09848a99e43/examples/inverse/plot_compute_mne_inverse.py#L57

https://github.com/mne-tools/mne-python/blob/c8633989fc3429d7dc1a4635fe05d09848a99e43/examples/plot_from_raw_to_epochs_to_evoked.py#L59

For evoked data with a clear peak this is expected to work reliably.
For single trial data it might take more sophisticated handling.
However, with the new functions you can specify tmin and tmax
paramters as well as handle the sign (only consider negative / positve
or absolute signals).

Best,
Denis