If you want to recover meaningful physical units I don't think it's
possible because the data undergo several normalisation steps as part
of the inversion. We might have a separate routine in SPM12 that will
preserve scaling but we don't have that at the moment.
Vladimir
On Mon, Oct 29, 2012 at 4:12 PM, Sophie Galer <[log in to unmask]> wrote:
> Thank you very much Stephan and Vladimir for your rapid answers.
>
> I have looked at the code 'spm_eeg_inv_extract' and i was wondering
> if you know how is the scale factor calculated?
>
> Best,
> Sophie
>
>
> 2012/10/29 Vladimir Litvak <[log in to unmask]>
>>
>> Dear Sophie,
>>
>> Time series extraction is implemented in SPM in spm_eeg_inv_extract
>> and the corresponding batch tool. You can look at the code if you need
>> something more specific.
>>
>> Best,
>>
>> Vladimir
>>
>> On Mon, Oct 29, 2012 at 9:32 AM, Sophie Galer <[log in to unmask]>
>> wrote:
>> > Dear SPM users,
>> > I have extracted the structure of a source reconstruction of interest:
>> >
>> > D.inv{29}.inverse
>> >
>> > ans =
>> >
>> > type: 'IID'
>> >
>> > woi: [-200 800]
>> >
>> > Han: 1
>> >
>> > lpf: 0
>> >
>> > hpf: 128
>> >
>> > pQ: {}
>> >
>> > trials: {'Congruent' 'Non Congruent'}
>> >
>> > modality: {'MEGPLANAR'}
>> >
>> > smooth: 0.6000
>> >
>> > xyz: [0 0 0]
>> >
>> > rad: 128
>> >
>> > scale: 9.0321e+09
>> >
>> > M: [8196x117 double]
>> >
>> > J: {[8196x13 double] [8196x13 double]}
>> >
>> > Y: {[117x13 double] [117x13 double]}
>> >
>> > L: [117x8196 double]
>> >
>> > qC: [8196x1 double]
>> >
>> > qV: [1001x1001 double]
>> >
>> > T: [1001x13 double]
>> >
>> > U: {[117x204 double]}
>> >
>> > Is: [8196x1 double]
>> >
>> > It: [1x1001 double]
>> >
>> > Ic: {[1x204 double]}
>> >
>> > Nd: 8196
>> >
>> > pst: [1x1001 double]
>> >
>> > dct: [1x257 double]
>> >
>> > F: 3.5993e+03
>> >
>> > ID: 246.3651
>> >
>> > R2: 99.1468
>> >
>> > VE: 0.9958
>> >
>> >
>> > I would like to applicate the inverse operator to my data. Do you know
>> > how
>> > this
>> > inverse operator is calculated from this structure ?
>> >
>> > Thank you very much for your help,
>> >
>> > --
>> > Sophie Galer
>> >
>
>
>
>
> --
> Sophie Galer - Ph.D. student
> National Fund for Scientific Research (F.R.S - FNRS)
> Université Libre de Bruxelles
> LCFC-UR2NF
> +322/555.42.98
> +322/650.42.62
> Hôpital Erasme
> 808 route de Lennik
> B-1070 Brussels, Belgium
>
|