Re: [Freesurfer] Robust PET signal projection

[Greve, Douglas N.,Ph.D.]

No. You can use mri_concat with --mtx option, where the argument to --mtx is an 
asci matrix that is square and diagonal with the weighting factors on the 
diagonal. You could also use mri_glmfix with the --X option, where X is an 
ascii column matrix of weighting factors (make sure that the the sum.^2 = 1)

On 12/21/18 2:31 PM, Matthieu Vanhoutte wrote:

External Email - Use Caution

Thanks Douglas.

Is there a way to compute a weighted average in one command with fscalc ?

Best,
Matthieu


Le mar. 18 déc. 2018 à 18:03, Greve, Douglas N.,Ph.D. 
mailto:dgr...@mgh.harvard.edu>> a écrit :
you can use mri_vol2surf to sample at different depths. If you want a simple 
average across cortex, then you can use --projfrac-avg .35 .65 .05
If you want to do more sophisticated weighting, then you will have to sample at 
each layer and then combine the files (eg, fscalc). Unless you have very small 
voxels, it probably will not make a difference.

On 12/18/18 4:50 AM, Matthieu Vanhoutte wrote:

External Email - Use Caution

Dear experts,

Is there any advice on this point ?

Best,
Matthieu


Le mer. 12 déc. 2018 à 10:47, Matthieu Vanhoutte 
mailto:matthieuvanhou...@gmail.com>> a écrit :
Dear Freesurfer's experts,

In order to have a robust PET signal projection, is it possible compute the 
projected PET signal as a weighted average of the PET signal intersecting with 
the surfaces ranging from 35 to 65% of the cortical tickness with a step t=5% ?

More weight would be given to the surfaces located near the mid distance 
between the pial and white surfaces as they have a higher probability to be 
well located within the cortex (using a normal distribution for example).

Thank you for helping.

Best,
Matthieu



___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer



___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

Re: [Freesurfer] Robust PET signal projection

[Matthieu Vanhoutte]

External Email - Use Caution

Hi Douglas,

Could you answer to my previous request ?

Best,
Matt

> Le 21 déc. 2018 à 20:31, Matthieu Vanhoutte  a 
> écrit :
> 
> Thanks Douglas.
> 
> Is there a way to compute a weighted average in one command with fscalc ?
> 
> Best,
> Matthieu
> 
> 
> Le mar. 18 déc. 2018 à 18:03, Greve, Douglas N.,Ph.D.  > a écrit :
> you can use mri_vol2surf to sample at different depths. If you want a simple 
> average across cortex, then you can use --projfrac-avg .35 .65 .05
> If you want to do more sophisticated weighting, then you will have to sample 
> at each layer and then combine the files (eg, fscalc). Unless you have very 
> small voxels, it probably will not make a difference. 
> 
> On 12/18/18 4:50 AM, Matthieu Vanhoutte wrote:
>> External Email - Use Caution
>> 
>> Dear experts,
>> 
>> Is there any advice on this point ?
>> 
>> Best,
>> Matthieu
>> 
>> 
>> Le mer. 12 déc. 2018 à 10:47, Matthieu Vanhoutte 
>> mailto:matthieuvanhou...@gmail.com>> a écrit :
>> Dear Freesurfer's experts,
>> 
>> In order to have a robust PET signal projection, is it possible compute the 
>> projected PET signal as a weighted average of the PET signal intersecting 
>> with the surfaces ranging from 35 to 65% of the cortical tickness with a 
>> step t=5% ?
>> 
>> More weight would be given to the surfaces located near the mid distance 
>> between the pial and white surfaces as they have a higher probability to be 
>> well located within the cortex (using a normal distribution for example).
>> 
>> Thank you for helping.
>> 
>> Best,
>> Matthieu
>> 
>> 
>> ___
>> Freesurfer mailing list
>> Freesurfer@nmr.mgh.harvard.edu 
>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer 
>> 
> ___
> Freesurfer mailing list
> Freesurfer@nmr.mgh.harvard.edu 
> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer 
> 
___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

Re: [Freesurfer] Robust PET signal projection

[Matthieu Vanhoutte]

External Email - Use Caution

Thanks Douglas.

Is there a way to compute a weighted average in one command with fscalc ?

Best,
Matthieu


Le mar. 18 déc. 2018 à 18:03, Greve, Douglas N.,Ph.D. <
dgr...@mgh.harvard.edu> a écrit :

> you can use mri_vol2surf to sample at different depths. If you want a
> simple average across cortex, then you can use --projfrac-avg .35 .65 .05
> If you want to do more sophisticated weighting, then you will have to
> sample at each layer and then combine the files (eg, fscalc). Unless you
> have very small voxels, it probably will not make a difference.
>
> On 12/18/18 4:50 AM, Matthieu Vanhoutte wrote:
>
> External Email - Use Caution
> Dear experts,
>
> Is there any advice on this point ?
>
> Best,
> Matthieu
>
>
> Le mer. 12 déc. 2018 à 10:47, Matthieu Vanhoutte <
> matthieuvanhou...@gmail.com> a écrit :
>
>> Dear Freesurfer's experts,
>>
>> In order to have a robust PET signal projection, is it possible compute
>> the projected PET signal as a weighted average of the PET signal
>> intersecting with the surfaces ranging from 35 to 65% of the cortical
>> tickness with a step t=5% ?
>>
>> More weight would be given to the surfaces located near the mid distance
>> between the pial and white surfaces as they have a higher probability to be
>> well located within the cortex (using a normal distribution for example).
>>
>> Thank you for helping.
>>
>> Best,
>> Matthieu
>>
>
>
> ___
> Freesurfer mailing 
> listfreesur...@nmr.mgh.harvard.eduhttps://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
>
>
> ___
> Freesurfer mailing list
> Freesurfer@nmr.mgh.harvard.edu
> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

Re: [Freesurfer] Robust PET signal projection

[Greve, Douglas N.,Ph.D.]

you can use mri_vol2surf to sample at different depths. If you want a simple 
average across cortex, then you can use --projfrac-avg .35 .65 .05
If you want to do more sophisticated weighting, then you will have to sample at 
each layer and then combine the files (eg, fscalc). Unless you have very small 
voxels, it probably will not make a difference.

On 12/18/18 4:50 AM, Matthieu Vanhoutte wrote:

External Email - Use Caution

Dear experts,

Is there any advice on this point ?

Best,
Matthieu


Le mer. 12 déc. 2018 à 10:47, Matthieu Vanhoutte 
mailto:matthieuvanhou...@gmail.com>> a écrit :
Dear Freesurfer's experts,

In order to have a robust PET signal projection, is it possible compute the 
projected PET signal as a weighted average of the PET signal intersecting with 
the surfaces ranging from 35 to 65% of the cortical tickness with a step t=5% ?

More weight would be given to the surfaces located near the mid distance 
between the pial and white surfaces as they have a higher probability to be 
well located within the cortex (using a normal distribution for example).

Thank you for helping.

Best,
Matthieu



___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

Re: [Freesurfer] Robust PET signal projection

[Matthieu Vanhoutte]

External Email - Use Caution

Dear experts,

Is there any advice on this point ?

Best,
Matthieu


Le mer. 12 déc. 2018 à 10:47, Matthieu Vanhoutte <
matthieuvanhou...@gmail.com> a écrit :

> Dear Freesurfer's experts,
>
> In order to have a robust PET signal projection, is it possible compute
> the projected PET signal as a weighted average of the PET signal
> intersecting with the surfaces ranging from 35 to 65% of the cortical
> tickness with a step t=5% ?
>
> More weight would be given to the surfaces located near the mid distance
> between the pial and white surfaces as they have a higher probability to be
> well located within the cortex (using a normal distribution for example).
>
> Thank you for helping.
>
> Best,
> Matthieu
>
___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

[Freesurfer] Robust PET signal projection

[Matthieu Vanhoutte]

External Email - Use Caution

Dear Freesurfer's experts,

In order to have a robust PET signal projection, is it possible compute the
projected PET signal as a weighted average of the PET signal intersecting
with the surfaces ranging from 35 to 65% of the cortical tickness with a
step t=5% ?

More weight would be given to the surfaces located near the mid distance
between the pial and white surfaces as they have a higher probability to be
well located within the cortex (using a normal distribution for example).

Thank you for helping.

Best,
Matthieu
___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer