External Email - Use Caution
Dear Doug Dear Bruce
Martin Reuter suggested me to write you since you are familiar with true inner workings of free surfer. In my project I try to create a family specific template that is based on three or more family members. We try to avoid the bias that may arise by choosing one family member as a target. There is procedure that tries to this described in:
https://secure-web.cisco.com/1Blsa_5fBePXLqTD4MyXB0MVZ8BfrqXwEgfuiFRtv-42VQP...
We found out that family specific templates result in better alignment among family members when compared to the standard fs average template. The creation of a template depends on the mris_register command that has several options that can be manipulated. However, we do not know what those options are doing in detail
What is the initial sulcal alignment? Is this a rigid alignment procedure? Maybe a hard spherical rotation based on least squares algorithms?
What is the final curve alignment? Is this a non-linear alignment procedure? If sulcal alignment is rigid, why does it take three times as long as non-linear curv alignment?
What is normalization in context of mris_register which can be disabled trough the -nonorm step?
Does the spring factor affect how much the Individual scul/curv maps are aligned with the family template? And if the latter is the case what is a reasonable spring value if you want to obtain individuals that are as closely matched to the family template as possible?
Thank you in Advance Many greetings from Austria
Jan
External Email - Use Caution
Hi Jan
If you create a family template you will only be able to compare within family, unless you also align the templates I guess?
-nonorm will disable z-normalization of curv and sulc. We typically start with a single subject atlas, do a nonlinear (but highly regularized) warp to it, then rebuild the atlas and re-register. We do have a prototype of a more principled procedure but it won't be ready for a few months
Cheers Bruce
From: freesurfer-bounces@nmr.mgh.harvard.edu freesurfer-bounces@nmr.mgh.harvard.edu On Behalf Of Koten, Jan Willem (jan.koten@uni-graz.at) Sent: Monday, May 8, 2023 8:01 AM To: freesurfer@nmr.mgh.harvard.edu Subject: [Freesurfer] mris_register options
External Email - Use Caution Dear Doug Dear Bruce
Martin Reuter suggested me to write you since you are familiar with true inner workings of free surfer. In my project I try to create a family specific template that is based on three or more family members. We try to avoid the bias that may arise by choosing one family member as a target. There is procedure that tries to this described in:
MailScanner has detected a possible fraud attempt from "secure-web.cisco.com" claiming to be https://secure-web.cisco.com/1IEewrYydOjuYqfmmr7slT8MnQVuTxf5Fupk3DeKFnIOTz1...https://secure-web.cisco.com/1Blsa_5fBePXLqTD4MyXB0MVZ8BfrqXwEgfuiFRtv-42VQP34OZ6UZqdfXxQi0tp2V1EoRTro-xOcgb7ClgzFVab407J4Z41cRtu_iJQ7EKKFe0dAvW9s60J_gwqS1U2oxlZRIad20a0m-SEzzq_ngB-95XOOsQpMhOfq6ISgUdxzB8ogeIbf9x0lfc2JgBeXKvf1G9CU7-7m6bSre4_KnhmyXHwl-5_eSx-dnrYxMX7JAnXnHbhmjXzfh_iBpIw7lCaHPbppGijCuVEpwx63sC4TUffSW5dEK1Gxyynbnu8l0ZQdcwiR0kzFZjQvXTtAQikt0tdUkeVQhN8lPE8JlA/https%3A%2F%2Ffreesurfer.net%2Ffswiki%2FSurfaceRegAndTemplates
We found out that family specific templates result in better alignment among family members when compared to the standard fs average template. The creation of a template depends on the mris_register command that has several options that can be manipulated. However, we do not know what those options are doing in detail
What is the initial sulcal alignment? Is this a rigid alignment procedure? Maybe a hard spherical rotation based on least squares algorithms?
What is the final curve alignment? Is this a non-linear alignment procedure?
If sulcal alignment is rigid, why does it take three times as long as non-linear curv alignment?
What is normalization in context of mris_register which can be disabled trough the -nonorm step?
Does the spring factor affect how much the Individual scul/curv maps are aligned with the family template? And if the latter is the case what is a reasonable spring value if you want to obtain individuals that are as closely matched to the family template as possible?
Thank you in Advance Many greetings from Austria
Jan
External Email - Use Caution
FYI, there is also a make_folding_atlas script, though you won't be able to fine tune the parameters when it runs mris_register
On 5/8/2023 10:12 AM, Fischl, Bruce R.,PHD wrote:
External Email - Use Caution
Hi Jan
If you create a family template you will only be able to compare within family, unless you also align the templates I guess?
-nonorm will disable z-normalization of curv and sulc. We typically start with a single subject atlas, do a nonlinear (but highly regularized) warp to it, then rebuild the atlas and re-register. We do have a prototype of a more principled procedure but it won’t be ready for a few months
Cheers
Bruce
*From:* freesurfer-bounces@nmr.mgh.harvard.edu freesurfer-bounces@nmr.mgh.harvard.edu *On Behalf Of *Koten, Jan Willem (jan.koten@uni-graz.at) *Sent:* Monday, May 8, 2023 8:01 AM *To:* freesurfer@nmr.mgh.harvard.edu *Subject:* [Freesurfer] mris_register options
* External Email - Use Caution *
Dear Doug
Dear Bruce
Martin Reuter suggested me to write you since you are familiar with true inner workings of free surfer. In my project I try to create a family specific template that is based on three or more family members. We try to avoid the bias that may arise by choosing one family member as a target. There is procedure that tries to this described in:
*MailScanner has detected a possible fraud attempt from "secure-web.cisco.com" claiming to be* https://secure-web.cisco.com/1DDLYdS73CiClH0FB--pPmqer1JYvU2AbOANdxXt1q3AnTU... https://secure-web.cisco.com/1Blsa_5fBePXLqTD4MyXB0MVZ8BfrqXwEgfuiFRtv-42VQP34OZ6UZqdfXxQi0tp2V1EoRTro-xOcgb7ClgzFVab407J4Z41cRtu_iJQ7EKKFe0dAvW9s60J_gwqS1U2oxlZRIad20a0m-SEzzq_ngB-95XOOsQpMhOfq6ISgUdxzB8ogeIbf9x0lfc2JgBeXKvf1G9CU7-7m6bSre4_KnhmyXHwl-5_eSx-dnrYxMX7JAnXnHbhmjXzfh_iBpIw7lCaHPbppGijCuVEpwx63sC4TUffSW5dEK1Gxyynbnu8l0ZQdcwiR0kzFZjQvXTtAQikt0tdUkeVQhN8lPE8JlA/https%3A%2F%2Ffreesurfer.net%2Ffswiki%2FSurfaceRegAndTemplates
We found out that family specific templates result in better alignment among family members when compared to the standard fs average template. The creation of a template depends on the mris_register command that has several options that can be manipulated. However, we do not know what those options are doing in detail
What is the initial sulcal alignment? Is this a rigid alignment procedure? Maybe a hard spherical rotation based on least squares algorithms?
What is the final curve alignment? Is this a non-linear alignment procedure?
If sulcal alignment is rigid, why does it take three times as long as non-linear curv alignment?
What is normalization in context of mris_register which can be disabled trough the -nonorm step?
Does the spring factor affect how much the Individual scul/curv maps are aligned with the family template? And if the latter is the case what is a reasonable spring value if you want to obtain individuals that are as closely matched to the family template as possible?
Thank you in Advance
Many greetings from Austria
Jan
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://secure-web.cisco.com/1t37ZzxtbzFxfLdM5NWTvwyGz2WFkSdSgaKTJveEN47_bI7...
freesurfer@nmr.mgh.harvard.edu
-
Douglas N. Greve -
Fischl, Bruce R.,PHD -
Koten, Jan Willem (jan.koten＠uni-graz.at)