Dear Johan,
You should simply enter your source-reconstructed nifti images (summaried in terms of a time-frequency contrast) into a second-level model in the stats module.s
Best wishes
Martin
> On 07 Nov 2015, at 14:56, Johan Baijot <[log in to unmask]> wrote:
>
> Dear SPM experts,
>
> Can the first level analysis be done on inverted EEG/MEG data? (on a similar way as fMRI)
>
> We imported our EEG data into SPM applied a high pass filter of 0.1Hz, a low pass filter of 30Hz and a band stop filter between 48 and 52 Hz, selected 3 different reoccurring epochs and implemented the 3D source reconstruction. Proceeding to the first level analysis we run into trouble, depending on the input for the conditions we get:
>
> - When we don't specify any conditions:
>
> Warning:"Imputs must be character arrays or cell arrays of strings" due to the empty result out of "out.D = spm_eeg_firstlevel(S);" = []
> This gives a warning on the next line when using fullfile(out.D)
>
> -When we define the conditions as done for the epoch part:
> Trial definition requires continuous dataset as input
> Error using spm_eeg_definetrial (line 49)
>
> Can someone elaborate on the correct way to specify the first level analysis conditions for inverted EEG data file with epoch's? As we couldn't find any additional documentation or example and don't understand the necessity of these conditions with our data set, as it is already necessary to cut into epochs before the 3D source reconstruction otherwise, source reconstruction is not possible on the complete continuous data due to lack of memory.
>
> Thank you in advance!
> Best regards
> Johan
|