External Email - Use Caution
Is there a way to resample a vertex-by-vertex matrix of data from one subject to another or to ico? For example, I have calculated the Euclidean distance matrix for all the vertices of lh.white for my subjects and wish to find the average distance matrix. Another example might be functional connectivity values.
If the data were univariate (i.e. vectors not matrices), I could save my measure as a .w file for each subject and use mri_surf2surf to morph each vector to ico and then average across the resulting ico vectors. One idea is to use make_average_subject to morph each subject’s surface to ico and then calculate bivariate measures on these surfaces. This might be ok for the functional connectivity case. However, this morphing distorts the absolute size of the surface, so the Euclidean distance matrix would be untenably distorted.
In other words, is there an extension of mri_surf2surf (or maybe mris_convert) that operates on v(i,j) measures rather than v(i) measures?
Thank you,
-Burke
No, there is not (at least that I know of). Bruce might have something along these lines.
On 4/26/2022 3:32 PM, Burke Rosen wrote:
External Email - Use CautionIs there a way to resample a vertex-by-vertex matrix of data from one subject to another or to ico? For example, I have calculated the Euclidean distance matrix for all the vertices of lh.white for my subjects and wish to find the average distance matrix. Another example might be functional connectivity values.
If the data were univariate (i.e. vectors not matrices), I could save my measure as a .w file for each subject and use mri_surf2surf to morph each vector to ico and then average across the resulting ico vectors. One idea is to use make_average_subject to morph each subject’s surface to ico and then calculate bivariate measures on these surfaces. This might be ok for the functional connectivity case. However, this morphing distorts the absolute size of the surface, so the Euclidean distance matrix would be untenably distorted.
In other words, is there an extension of mri_surf2surf (or maybe mris_convert) that operates on v(i,j) measures rather than v(i) measures?
Thank you,
-Burke
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Why not just flatten it and save it as a .mgz? Then I think it should be straightforward
-----Original Message----- From: Greve, Douglas N.,Ph.D. DGREVE@mgh.harvard.edu Sent: Wednesday, April 27, 2022 10:42 AM To: freesurfer@nmr.mgh.harvard.edu; Fischl, Bruce BFISCHL@mgh.harvard.edu Subject: Re: [Freesurfer] Resampling of surface-based bivariate data
No, there is not (at least that I know of). Bruce might have something along these lines.
On 4/26/2022 3:32 PM, Burke Rosen wrote:
External Email - Use CautionIs there a way to resample a vertex-by-vertex matrix of data from one subject to another or to ico? For example, I have calculated the Euclidean distance matrix for all the vertices of lh.white for my subjects and wish to find the average distance matrix. Another example might be functional connectivity values.
If the data were univariate (i.e. vectors not matrices), I could save my measure as a .w file for each subject and use mri_surf2surf to morph each vector to ico and then average across the resulting ico vectors. One idea is to use make_average_subject to morph each subject’s surface to ico and then calculate bivariate measures on these surfaces. This might be ok for the functional connectivity case. However, this morphing distorts the absolute size of the surface, so the Euclidean distance matrix would be untenably distorted.
In other words, is there an extension of mri_surf2surf (or maybe mris_convert) that operates on v(i,j) measures rather than v(i) measures?
Thank you,
-Burke
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
You'd have do both the rows and the cols.
On 4/27/2022 11:44 AM, Fischl, Bruce wrote:
Why not just flatten it and save it as a .mgz? Then I think it should be straightforward
-----Original Message----- From: Greve, Douglas N.,Ph.D. DGREVE@mgh.harvard.edu Sent: Wednesday, April 27, 2022 10:42 AM To: freesurfer@nmr.mgh.harvard.edu; Fischl, Bruce BFISCHL@mgh.harvard.edu Subject: Re: [Freesurfer] Resampling of surface-based bivariate data
No, there is not (at least that I know of). Bruce might have something along these lines.
On 4/26/2022 3:32 PM, Burke Rosen wrote:
External Email - Use CautionIs there a way to resample a vertex-by-vertex matrix of data from one subject to another or to ico? For example, I have calculated the Euclidean distance matrix for all the vertices of lh.white for my subjects and wish to find the average distance matrix. Another example might be functional connectivity values.
If the data were univariate (i.e. vectors not matrices), I could save my measure as a .w file for each subject and use mri_surf2surf to morph each vector to ico and then average across the resulting ico vectors. One idea is to use make_average_subject to morph each subject’s surface to ico and then calculate bivariate measures on these surfaces. This might be ok for the functional connectivity case. However, this morphing distorts the absolute size of the surface, so the Euclidean distance matrix would be untenably distorted.
In other words, is there an extension of mri_surf2surf (or maybe mris_convert) that operates on v(i,j) measures rather than v(i) measures?
Thank you,
-Burke
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Right, I meant flatten the matrix into a vector/vertex, do the cross-subject mapping, then reshape it back to a matrix
-----Original Message----- From: Greve, Douglas N.,Ph.D. DGREVE@mgh.harvard.edu Sent: Wednesday, April 27, 2022 12:24 PM To: Fischl, Bruce BFISCHL@mgh.harvard.edu; freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] Resampling of surface-based bivariate data
You'd have do both the rows and the cols.
On 4/27/2022 11:44 AM, Fischl, Bruce wrote:
Why not just flatten it and save it as a .mgz? Then I think it should be straightforward
-----Original Message----- From: Greve, Douglas N.,Ph.D. DGREVE@mgh.harvard.edu Sent: Wednesday, April 27, 2022 10:42 AM To: freesurfer@nmr.mgh.harvard.edu; Fischl, Bruce BFISCHL@mgh.harvard.edu Subject: Re: [Freesurfer] Resampling of surface-based bivariate data
No, there is not (at least that I know of). Bruce might have something along these lines.
On 4/26/2022 3:32 PM, Burke Rosen wrote:
External Email - Use CautionIs there a way to resample a vertex-by-vertex matrix of data from one subject to another or to ico? For example, I have calculated the Euclidean distance matrix for all the vertices of lh.white for my subjects and wish to find the average distance matrix. Another example might be functional connectivity values.
If the data were univariate (i.e. vectors not matrices), I could save my measure as a .w file for each subject and use mri_surf2surf to morph each vector to ico and then average across the resulting ico vectors. One idea is to use make_average_subject to morph each subject’s surface to ico and then calculate bivariate measures on these surfaces. This might be ok for the functional connectivity case. However, this morphing distorts the absolute size of the surface, so the Euclidean distance matrix would be untenably distorted.
In other words, is there an extension of mri_surf2surf (or maybe mris_convert) that operates on v(i,j) measures rather than v(i) measures?
Thank you,
-Burke
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
Burke Rosen -
Douglas N. Greve -
Fischl, Bruce