Re: [HCP-Users] Glasser nature atlas
We do need to get an MSM binary out with a revised MSMSulc config file. This will hopefully happen soon, we are just waiting for the author of a software library to e-mail us back about some licensing stuff. Peace, Matt. From: <hcp-users-boun...@humanconnectome.org<mailto:hcp-users-boun...@humanconnectome.org>> on behalf of Jared P Zimmerman <jar...@mail.med.upenn.edu<mailto:jar...@mail.med.upenn.edu>> Date: Wednesday, October 12, 2016 at 4:36 PM To: "Dierker, Donna" <do...@wustl.edu<mailto:do...@wustl.edu>> Cc: "hcp-users@humanconnectome.org<mailto:hcp-users@humanconnectome.org>" <hcp-users@humanconnectome.org<mailto:hcp-users@humanconnectome.org>> Subject: Re: [HCP-Users] Glasser nature atlas Thanks Donna, much obliged. On Oct 12, 2016, at 5:09 PM, Dierker, Donna <do...@wustl.edu<mailto:do...@wustl.edu>> wrote: Search for MSMSulc here: https://github.com/Washington-University/Pipelines/blob/master/PostFreeSurfer/scripts/FreeSurfer2CaretConvertAndRegisterNonlinear.sh *refsulc* get copied to subject dir from here: https://github.com/Washington-University/Pipelines/tree/7cc0bf1863cbc8a1a7ab9c6dd48a25de9be9bae7/global/templates/standard_mesh_atlases On Oct 12, 2016, at 10:53 AM, Jared P Zimmerman <jar...@mail.med.upenn.edu<mailto:jar...@mail.med.upenn.edu>> wrote: Thanks Michael, Donna, and Matt, RE: Michael, definitely aware this is an inferior approach to true subject-specific parcellation but given that we don’t have the requisite data to run the classifier this seems like an appropriate approach to use the parcellation. RE: Donna and Matt, thanks for the advice, I think using an MSM sulc approach is something we will definitely want to look into, but looking over the git repo quickly I don’t see an MSM sulc script. There’s the MSM all pipeline, but that seems like it can only use myelin maps, RSNs or topography and doesn’t have an option for sulcal depth. Where might I find an MSM algorithm that can use sulcal depth maps to register? Thanks, Jared On Oct 12, 2016, at 9:48 AM, Dierker, Donna <do...@wustl.edu<mailto:do...@wustl.edu>> wrote: If you can’t do MSMAll, then consider using just sulc, rather than sulc+curv. See figure 4, row B, in Emma Robinson’s paper: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4190319/figure/F4/ I have heard David describe curvature as capturing more local folding, sulc global folding. In my experience, curvature varies too much across subjects, while sulc seems a more reasonable driver for intersubject registration. For intrasubject registration across timepoints, however, curvature is your friend. On Oct 11, 2016, at 6:54 PM, Glasser, Matthew <glass...@wustl.edu<mailto:glass...@wustl.edu>> wrote: Using folding-based alignment won’t be as good as areal-feature-based alignment, but I wouldn’t worry about the alignment being any worse than any other comparison of two studies driven by folding-based alignment. Peace, Matt. From: <hcp-users-boun...@humanconnectome.org<mailto:hcp-users-boun...@humanconnectome.org>> on behalf of Jared P Zimmerman <jar...@mail.med.upenn.edu<mailto:jar...@mail.med.upenn.edu>> Date: Tuesday, October 11, 2016 at 3:12 PM To: Timothy Coalson <tsc...@mst.edu<mailto:tsc...@mst.edu>> Cc: "hcp-users@humanconnectome.org<mailto:hcp-users@humanconnectome.org>" <hcp-users@humanconnectome.org<mailto:hcp-users@humanconnectome.org>> Subject: Re: [HCP-Users] Glasser nature atlas I have a related question to this. I’d like to use the parcellation on some more traditional volume data (i.e. 3x3x3mm voxels, 3s TR). I have made fs_LR surfaces from free surfer using the wb_command shortcuts for it, as in here: https://wiki.humanconnectome.org/download/attachments/63078513/Resampling-FreeSurfer-HCP.pdf?version=1=1472225460934=v2 but I am missing the registration from subject surface to group average. Is this registration, or the validity of the parcellation, dependent on MSMAll or would an MSMSulc registration be effective (enough) here? The paper and parcellation are clearly based on MSMAll and this was essential in generating parcellation, but since I don’t have the requisite rest data to do an MSMAll register, would an MSMSulc registration be sufficient for ultimately translating back to the subject volume? I read the above discussion about a month ago and didn’t see this covered, but I don’t know if that was addressed in the discussion since. Tanks, Jared On Oct 4, 2016, at 5:45 PM, Timothy Coalson <tsc...@mst.edu<mailto:tsc...@mst.edu>> wrote: Making a "nifti version" isn't a simple conversion, as some lengthy conversations on this list have elaborated on. The good news: if you want a single-subject-specific version, that is possible, using -cifti-separate and -label-to-vo
Re: [HCP-Users] Glasser nature atlas
Thanks Michael, Donna, and Matt, RE: Michael, definitely aware this is an inferior approach to true subject-specific parcellation but given that we don’t have the requisite data to run the classifier this seems like an appropriate approach to use the parcellation. RE: Donna and Matt, thanks for the advice, I think using an MSM sulc approach is something we will definitely want to look into, but looking over the git repo quickly I don’t see an MSM sulc script. There’s the MSM all pipeline, but that seems like it can only use myelin maps, RSNs or topography and doesn’t have an option for sulcal depth. Where might I find an MSM algorithm that can use sulcal depth maps to register? Thanks, Jared > On Oct 12, 2016, at 9:48 AM, Dierker, Donna <do...@wustl.edu> wrote: > > If you can’t do MSMAll, then consider using just sulc, rather than sulc+curv. > > See figure 4, row B, in Emma Robinson’s paper: > > https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4190319/figure/F4/ > > I have heard David describe curvature as capturing more local folding, sulc > global folding. > In my experience, curvature varies too much across subjects, while sulc seems > a more reasonable driver for intersubject registration. > > For intrasubject registration across timepoints, however, curvature is your > friend. > > >> On Oct 11, 2016, at 6:54 PM, Glasser, Matthew <glass...@wustl.edu> wrote: >> >> Using folding-based alignment won’t be as good as areal-feature-based >> alignment, but I wouldn’t worry about the alignment being any worse than any >> other comparison of two studies driven by folding-based alignment. >> >> Peace, >> >> Matt. >> >> From: <hcp-users-boun...@humanconnectome.org> on behalf of Jared P Zimmerman >> <jar...@mail.med.upenn.edu> >> Date: Tuesday, October 11, 2016 at 3:12 PM >> To: Timothy Coalson <tsc...@mst.edu> >> Cc: "hcp-users@humanconnectome.org" <hcp-users@humanconnectome.org> >> Subject: Re: [HCP-Users] Glasser nature atlas >> >> I have a related question to this. I’d like to use the parcellation on some >> more traditional volume data (i.e. 3x3x3mm voxels, 3s TR). I have made >> fs_LR surfaces from free surfer using the wb_command shortcuts for it, as in >> here: >> https://wiki.humanconnectome.org/download/attachments/63078513/Resampling-FreeSurfer-HCP.pdf?version=1=1472225460934=v2 >> but I am missing the registration from subject surface to group average. >> >> Is this registration, or the validity of the parcellation, dependent on >> MSMAll or would an MSMSulc registration be effective (enough) here? The >> paper and parcellation are clearly based on MSMAll and this was essential in >> generating parcellation, but since I don’t have the requisite rest data to >> do an MSMAll register, would an MSMSulc registration be sufficient for >> ultimately translating back to the subject volume? >> >> I read the above discussion about a month ago and didn’t see this covered, >> but I don’t know if that was addressed in the discussion since. >> >> >> Tanks, >> Jared >> >> >> >>> On Oct 4, 2016, at 5:45 PM, Timothy Coalson <tsc...@mst.edu> wrote: >>> >>> Making a "nifti version" isn't a simple conversion, as some lengthy >>> conversations on this list have elaborated on. >>> >>> The good news: if you want a single-subject-specific version, that is >>> possible, using -cifti-separate and -label-to-volume-mapping, using the >>> subject's own surfaces. >>> >>> The bad news: if you want it to be a group atlas, the currently popular >>> volume registration methods aren't good enough at cortical area alignment >>> to make our parcellation meaningful in a group-average volume space, and >>> thus we do not recommend making a volumetric version in such a space. In >>> the meantime, surface registration avoids several difficulties inherent to >>> volume registration, and was instrumental in being able to create the >>> parcellation at all (via alignment of areal features). >>> >>> For more detailed discussion: >>> >>> https://www.mail-archive.com/hcp-users@humanconnectome.org/msg03078.html >>> >>> Tim >>> >>> >>> On Tue, Oct 4, 2016 at 9:47 AM, Elam, Jennifer <e...@wustl.edu> wrote: >>>> Hi Peter, >>>> The data for the new parcellation are available in the BALSA database, see >>>> this thread: >>>> http://www.mail-archive.com/hcp-user
Re: [HCP-Users] Glasser nature atlas
If you can’t do MSMAll, then consider using just sulc, rather than sulc+curv. See figure 4, row B, in Emma Robinson’s paper: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4190319/figure/F4/ I have heard David describe curvature as capturing more local folding, sulc global folding. In my experience, curvature varies too much across subjects, while sulc seems a more reasonable driver for intersubject registration. For intrasubject registration across timepoints, however, curvature is your friend. > On Oct 11, 2016, at 6:54 PM, Glasser, Matthew <glass...@wustl.edu> wrote: > > Using folding-based alignment won’t be as good as areal-feature-based > alignment, but I wouldn’t worry about the alignment being any worse than any > other comparison of two studies driven by folding-based alignment. > > Peace, > > Matt. > > From: <hcp-users-boun...@humanconnectome.org> on behalf of Jared P Zimmerman > <jar...@mail.med.upenn.edu> > Date: Tuesday, October 11, 2016 at 3:12 PM > To: Timothy Coalson <tsc...@mst.edu> > Cc: "hcp-users@humanconnectome.org" <hcp-users@humanconnectome.org> > Subject: Re: [HCP-Users] Glasser nature atlas > > I have a related question to this. I’d like to use the parcellation on some > more traditional volume data (i.e. 3x3x3mm voxels, 3s TR). I have made fs_LR > surfaces from free surfer using the wb_command shortcuts for it, as in here: > https://wiki.humanconnectome.org/download/attachments/63078513/Resampling-FreeSurfer-HCP.pdf?version=1=1472225460934=v2 > but I am missing the registration from subject surface to group average. > > Is this registration, or the validity of the parcellation, dependent on > MSMAll or would an MSMSulc registration be effective (enough) here? The > paper and parcellation are clearly based on MSMAll and this was essential in > generating parcellation, but since I don’t have the requisite rest data to do > an MSMAll register, would an MSMSulc registration be sufficient for > ultimately translating back to the subject volume? > > I read the above discussion about a month ago and didn’t see this covered, > but I don’t know if that was addressed in the discussion since. > > > Tanks, > Jared > > > >> On Oct 4, 2016, at 5:45 PM, Timothy Coalson <tsc...@mst.edu> wrote: >> >> Making a "nifti version" isn't a simple conversion, as some lengthy >> conversations on this list have elaborated on. >> >> The good news: if you want a single-subject-specific version, that is >> possible, using -cifti-separate and -label-to-volume-mapping, using the >> subject's own surfaces. >> >> The bad news: if you want it to be a group atlas, the currently popular >> volume registration methods aren't good enough at cortical area alignment to >> make our parcellation meaningful in a group-average volume space, and thus >> we do not recommend making a volumetric version in such a space. In the >> meantime, surface registration avoids several difficulties inherent to >> volume registration, and was instrumental in being able to create the >> parcellation at all (via alignment of areal features). >> >> For more detailed discussion: >> >> https://www.mail-archive.com/hcp-users@humanconnectome.org/msg03078.html >> >> Tim >> >> >> On Tue, Oct 4, 2016 at 9:47 AM, Elam, Jennifer <e...@wustl.edu> wrote: >>> Hi Peter, >>> The data for the new parcellation are available in the BALSA database, see >>> this thread: >>> http://www.mail-archive.com/hcp-users@humanconnectome.org/msg02876.html >>> >>> We recommend using the included Connectome Workbench scene file to >>> conveniently view the data in the wb_view part of the software platform. >>> Get Connectome Workbench here: >>> https://www.humanconnectome.org/software/get-connectome-workbench.html >>> >>> Many of the files are in the CIFTI data format, which has a lot of >>> advantages, but if you need NIFTI versions, you can use wb_command >>> (distributed as part of Connectome Workbench) to do the conversion. >>> >>> There's lots of help for using wb_command in the HCP-Users mail archive, >>> and/or you could post more questions to the list. >>> >>> Best, >>> Jenn >>> >>> >>> Jennifer Elam, Ph.D. >>> Scientific Outreach, Human Connectome Project >>> Washington University School of Medicine >>> Department of Neuroscience, Box 8108 >>> 660 South Euclid Avenue >>> St. Louis, MO 63110 >>> 314-362-9387 >>> e...@wustl.edu >>> www.hu
Re: [HCP-Users] Glasser nature atlas
You can get the help info of commands by running them with no arguments, or from here: http://www.humanconnectome.org/software/workbench-command.php?function=-label-to-volume-mapping Note that the ribbon mapping method (which is generally recommended when you have pial and white surfaces) of surface to volume mapping is "generous" in that it assigns values to voxels that are only partially inside the cortical ribbon - this can be dealt with if it is a problem for your use. Tim On Tue, Oct 11, 2016 at 2:11 AM, Peter McColgan <pmccolga...@qub.ac.uk> wrote: > > Thanks Matt, I am keen to generate a single-subject-specific version using > -cifti-separate and -label-to-volume-mapping, using the subject's own > surfaces. > > > Could you advise how I go about doing this? > > > Thanks again for your help. > > > bw > > > Peter > -- > *From:* Glasser, Matthew <glass...@wustl.edu> > *Sent:* 11 October 2016 00:09:24 > *To:* Peter McColgan; NEUROSCIENCE tim; Elam, Jennifer > *Cc:* hcp-users@humanconnectome.org > > *Subject:* Re: [HCP-Users] Glasser nature atlas > > A number of people have been using this file: HCP_PhaseTwo/Q1-Q6_ > RelatedValidation210/MNINonLinear/fsaverage_LR32k/ > Q1-Q6_RelatedValidation210.CorticalAreas_dil_Final_Final_ > Areas_Group_Colors.32k_fs_LR.dlabel.nii > > Peace, > > Matt. > > From: <hcp-users-boun...@humanconnectome.org> on behalf of Peter McColgan > <pmccolga...@qub.ac.uk> > Date: Monday, October 10, 2016 at 3:22 PM > To: Timothy Coalson <tsc...@mst.edu>, "Elam, Jennifer" <e...@wustl.edu> > Cc: "hcp-users@humanconnectome.org" <hcp-users@humanconnectome.org> > Subject: Re: [HCP-Users] Glasser nature atlas > > Thanks Tim, I have found a number of files at https://balsa.wustl.edu/ > sceneFile/show/L731 can I ask which is the best one to use for the nature > paper parcellation? > > Thanks > > Peter > > ---------- > *From:* Timothy Coalson <tsc...@mst.edu> > *Sent:* 04 October 2016 22:45:46 > *To:* Elam, Jennifer > *Cc:* Peter McColgan; hcp-users@humanconnectome.org > *Subject:* Re: [HCP-Users] Glasser nature atlas > > Making a "nifti version" isn't a simple conversion, as some lengthy > conversations on this list have elaborated on. > > The good news: if you want a single-subject-specific version, that is > possible, using -cifti-separate and -label-to-volume-mapping, using the > subject's own surfaces. > > The bad news: if you want it to be a group atlas, the currently popular > volume registration methods aren't good enough at cortical area alignment > to make our parcellation meaningful in a group-average volume space, and > thus we do not recommend making a volumetric version in such a space. In > the meantime, surface registration avoids several difficulties inherent to > volume registration, and was instrumental in being able to create the > parcellation at all (via alignment of areal features). > > For more detailed discussion: > > https://www.mail-archive.com/hcp-users@humanconnectome.org/msg03078.html > > Tim > > > On Tue, Oct 4, 2016 at 9:47 AM, Elam, Jennifer <e...@wustl.edu> wrote: > >> Hi Peter, >> >> The data for the new parcellation are available in the BALSA database, >> see this thread: http://www.mail-archive.com/hc >> p-us...@humanconnectome.org/msg02876.html >> >> >> We recommend using the included Connectome Workbench scene file to >> conveniently view the data in the wb_view part of the software platform. >> Get Connectome Workbench here: https://www.humanconnectome.or >> g/software/get-connectome-workbench.html >> >> >> Many of the files are in the CIFTI data format >> <http://www.humanconnectome.org/software/workbench-command.php?function=-cifti-help>, >> which has a lot of advantages, but if you need NIFTI versions, you can use >> wb_command >> <http://www.humanconnectome.org/software/workbench-command.php> >> (distributed as part of Connectome Workbench) to do the conversion. >> >> >> There's lots of help for using wb_command in the HCP-Users mail archive, >> and/or you could post more questions to the list. >> >> >> Best, >> >> Jenn >> >> >> Jennifer Elam, Ph.D. >> Scientific Outreach, Human Connectome Project >> Washington University School of Medicine >> Department of Neuroscience, Box 8108 >> 660 South Euclid Avenue >> St. Louis, MO 63110 >> 314-362-9387 >> e...@wustl.edu >> www.humanconnectome.org >> >> --
Re: [HCP-Users] Glasser nature atlas
Making a "nifti version" isn't a simple conversion, as some lengthy conversations on this list have elaborated on. The good news: if you want a single-subject-specific version, that is possible, using -cifti-separate and -label-to-volume-mapping, using the subject's own surfaces. The bad news: if you want it to be a group atlas, the currently popular volume registration methods aren't good enough at cortical area alignment to make our parcellation meaningful in a group-average volume space, and thus we do not recommend making a volumetric version in such a space. In the meantime, surface registration avoids several difficulties inherent to volume registration, and was instrumental in being able to create the parcellation at all (via alignment of areal features). For more detailed discussion: https://www.mail-archive.com/hcp-users@humanconnectome.org/msg03078.html Tim On Tue, Oct 4, 2016 at 9:47 AM, Elam, Jennifer <e...@wustl.edu> wrote: > Hi Peter, > > The data for the new parcellation are available in the BALSA database, see > this thread: http://www.mail-archive.com/hcp-users@humanconnectome.org/ > msg02876.html > > > We recommend using the included Connectome Workbench scene file to > conveniently view the data in the wb_view part of the software platform. > Get Connectome Workbench here: https://www.humanconnectome. > org/software/get-connectome-workbench.html > > > Many of the files are in the CIFTI data format > <http://www.humanconnectome.org/software/workbench-command.php?function=-cifti-help>, > which has a lot of advantages, but if you need NIFTI versions, you can use > wb_command <http://www.humanconnectome.org/software/workbench-command.php> > (distributed as part of Connectome Workbench) to do the conversion. > > > There's lots of help for using wb_command in the HCP-Users mail archive, > and/or you could post more questions to the list. > > > Best, > > Jenn > > > Jennifer Elam, Ph.D. > Scientific Outreach, Human Connectome Project > Washington University School of Medicine > Department of Neuroscience, Box 8108 > 660 South Euclid Avenue > St. Louis, MO 63110 > 314-362-9387 > e...@wustl.edu > www.humanconnectome.org > > -- > *From:* hcp-users-boun...@humanconnectome.org <hcp-users-bounces@ > humanconnectome.org> on behalf of Peter McColgan <pmccolga...@qub.ac.uk> > *Sent:* Tuesday, October 4, 2016 9:29:30 AM > *To:* hcp-users@humanconnectome.org > *Subject:* [HCP-Users] Glasser nature atlas > > > > Dear HCP team, > > > Is it possible to download a nifti version of the Glasser atlas featured > recently in nature? > > > Thanks > > > bw > > > Peter > > ___ > HCP-Users mailing list > HCP-Users@humanconnectome.org > http://lists.humanconnectome.org/mailman/listinfo/hcp-users > > ___ > HCP-Users mailing list > HCP-Users@humanconnectome.org > http://lists.humanconnectome.org/mailman/listinfo/hcp-users > ___ HCP-Users mailing list HCP-Users@humanconnectome.org http://lists.humanconnectome.org/mailman/listinfo/hcp-users
Re: [HCP-Users] Glasser nature atlas
Hi Peter, The data for the new parcellation are available in the BALSA database, see this thread: http://www.mail-archive.com/hcp-users@humanconnectome.org/msg02876.html We recommend using the included Connectome Workbench scene file to conveniently view the data in the wb_view part of the software platform. Get Connectome Workbench here: https://www.humanconnectome.org/software/get-connectome-workbench.html Many of the files are in the CIFTI data format<http://www.humanconnectome.org/software/workbench-command.php?function=-cifti-help>, which has a lot of advantages, but if you need NIFTI versions, you can use wb_command<http://www.humanconnectome.org/software/workbench-command.php> (distributed as part of Connectome Workbench) to do the conversion. There's lots of help for using wb_command in the HCP-Users mail archive, and/or you could post more questions to the list. Best, Jenn Jennifer Elam, Ph.D. Scientific Outreach, Human Connectome Project Washington University School of Medicine Department of Neuroscience, Box 8108 660 South Euclid Avenue St. Louis, MO 63110 314-362-9387 e...@wustl.edu<mailto:e...@wustl.edu> www.humanconnectome.org<http://www.humanconnectome.org/> From: hcp-users-boun...@humanconnectome.org <hcp-users-boun...@humanconnectome.org> on behalf of Peter McColgan <pmccolga...@qub.ac.uk> Sent: Tuesday, October 4, 2016 9:29:30 AM To: hcp-users@humanconnectome.org Subject: [HCP-Users] Glasser nature atlas Dear HCP team, Is it possible to download a nifti version of the Glasser atlas featured recently in nature? Thanks bw Peter ___ HCP-Users mailing list HCP-Users@humanconnectome.org http://lists.humanconnectome.org/mailman/listinfo/hcp-users ___ HCP-Users mailing list HCP-Users@humanconnectome.org http://lists.humanconnectome.org/mailman/listinfo/hcp-users
[HCP-Users] Glasser nature atlas
Dear HCP team, Is it possible to download a nifti version of the Glasser atlas featured recently in nature? Thanks bw Peter ___ HCP-Users mailing list HCP-Users@humanconnectome.org http://lists.humanconnectome.org/mailman/listinfo/hcp-users
