mne_make_movie cannot find data set

✉️ Mailing List Archive (read-only)
1

I sent this to Matti, sharing it to the list now. I am using a single epoch of raw data, not an averaged evoked response, but apparently mne-make-movie is expecting the later?. mne_analysze will not work for this type of data, so I am trying to use mne-make-movie. The error I am getting below seems like the type of error mne-analyze would give with this EEG data.

-Jeff

2

hi jeff,

I see 2 options:

option 1: make an event file with only 1 event and generate a fake evoked data
option 2: use matlab or python [1]

modifying mne_make_movie would be much more difficult.

Alex

[1] http://martinos.org/mne/auto_examples/inverse/plot_compute_mne_inverse_raw_in_label.html

3

Alex,

I will explore these two suggestions. I do find it odd that apparently MNE
has no straightforward path to analyze unaveraged data. I thought some
people used it for epilepsy analysis?

I am perfectly happy to use Matlab at this point (no experience with
Python - is there a version for Macintosh?). I tried initially to use
Matlab to import my EEG data but that was very difficult, even with help,
and I was encouraged to use mne_edf2fiff instead.

Thanks,
-Jeff

4

I will explore these two suggestions. I do find it odd that apparently MNE
has no straightforward path to analyze unaveraged data.

you can generate stc files from raw data with mne_compute_raw_inverse
but you will need to provide labels.

I thought some
people used it for epilepsy analysis?

People do at the martinos but I don't know how they work.

I am perfectly happy to use Matlab at this point (no experience with
Python - is there a version for Macintosh?). I tried initially to use
Matlab to import my EEG data but that was very difficult, even with help,
and I was encouraged to use mne_edf2fiff instead.

http://martinos.org/mne/python_tutorial.html

on a mac I recommend you install EPD (free with an academic license)
or EPD free

then you can you should be able to install MNE following the instructions at:

then you can adapt the example for your own data.

hope this helps

Alex

5

Please see below
-Jeff

I will explore these two suggestions. I do find it odd that apparently
MNE
has no straightforward path to analyze unaveraged data.

you can generate stc files from raw data with mne_compute_raw_inverse
but you will need to provide labels.

If I supplied labels, which I understand are somewhat like ROIs, are you
sure I would not run into the same problem I did with mne-make-movie? I
thought mne-compute-raw-inverse was very similar to mne_make_movie?

I thought some
people used it for epilepsy analysis?

People do at the martinos but I don't know how they work.

Can anyone out there put me in touch with these people please? My current
project involves autism and ADHD, but epilepsy application is another one
of my interests. All of these involve unaveraged EEG.

6

Jeff,

Briefly,

1. mne_make_movie can only produce movies from averaged evoked response data

2. mne_compute_raw_inverse can produce continuous raw data source estimates from a set of labels.

3. If you have trigger points defined, you can also load raw data to mne_analyze and view source estimates at individual epochs.

4. If you load the raw data to mne_browse_raw, you can start mne_analyze from there and interactively pick time points to show source estimates corresponding to, e.g., epileptic spikes.

The options have been hardly used in practice because most of the time researchers are interested in calculating, e.g., time-frequency representations, which leads averaging of the transformed epochs in the source space.

- Matti

7

really nice !

I just gave it a try on the mne-sample data and it worked like a charm.

Thanks matti for continuing to show me the hidden gems of MNE :slight_smile:

Alex

8

really nice !

I just gave it a try on the mne-sample data and it worked like a charm.

Thanks matti for continuing to show me the hidden gems of MNE :slight_smile:

Number 4 or number 3?

- Matti

9

really nice !

I just gave it a try on the mne-sample data and it worked like a charm.

Thanks matti for continuing to show me the hidden gems of MNE :slight_smile:

Number 4 or number 3?

both !

for 3 you will need to click on ?File > switch to data set" to go to
the next epoch

Alex

10

Matti,

Looks like option 3 will give me the most bang for my buck, so I will try to accomplish that next, by manually adding trigger points. I am hoping I can add them to the original NetStation data, then export to EDF format, then to FIFF.

Thanks,
-Jeff