Re: [caret-users] negative values displayed positive
Hi Dona, thank you very much for looking into this. The Volume I sent contained values that were thresholded for significance. This is why some values were zero. The blob probably results from the cluster-permutation approach that I applied in order to compute the significance. I should also say that the data comes from MEG source reconstruction and that I interpolated the data onto the SPM template brain (T1.nii). Based on your response I assume that the problem does not come from Caret but from the data that contains positive values. The problem must thus occur while I am exporting the data to a volume file. Thanks for the quick help! I know where to search now. Fred -- FR - Original Message - From: "Donna Dierker" To: "SureFit Caret, and SuMS software users" Sent: Monday, June 22, 2015 4:36:09 PM Subject: Re: [caret-users] negative values displayed positive Hi FR, This is rather an odd file. I did 3dhistog on it and got this output: http://brainmap.wustl.edu/pub/donna/SPAIN/BCBL/your_file.3dhistog.txt login pub password download Mostly zeroes, but 14184 voxels set to 32766 (and no voxels set to any values in between 0 and 32766). And when I map this to the PALS atlas, I get nothing on most vertices, but solid yellow over a perisylvian blob. And when I click in the yellow, sure enough the ID window confirms the vertex is set to 32766 intensity. Working as expected, but data a bit off the beaten path. Donna On Jun 21, 2015, at 4:55 PM, Frédéric Roux wrote: > Hi Dona, > > thanks for getting back on this. > I am using NIFTI, so nothing rare from that side. > > Just uploaded the file via the link you sent me. > > Thanks for looking into this. > > -- > FR > > - Original Message - > From: "Donna Dierker" > To: "SureFit Caret, and SuMS software users" > Sent: Sunday, June 21, 2015 6:37:50 PM > Subject: Re: [caret-users] negative values displayed positive > > I'm not sure what would do this. What format is your volume? NIFTI? Older > .hdr/.img pairs sometimes got x-flipped upon opening under some > circumstances, but nothing I know of (except explicit *(-1) using volume > math) does this. > > You can use something AFNI 3dinfo or 3dhistog to double-check your volume's > values, or you can upload your volume here and I can do so tomorrow: > > http://brainvis.wustl.edu/cgi-bin/upload.cgi > > You can also have a look at the palette format, and just make sure it's not > the palette that is making your values appear to be flipped: > > http://brainvis.wustl.edu/CaretHelpAccount/caret5_help/file_formats/file_formats.html#paletteFile > > Caret has some built-in palettes (see D/C: Metric Settings menu). > > > On Jun 21, 2015, at 3:55 AM, Frédéric Roux wrote: > >> Dear all, >> >> I'm trying to map a functional volume to the PLAS-B12 >> surface but the sign of the values gets flipped so that >> negative values appear as positive ones. >> >> Does anyone know what could be wrong? >> >> Many thanks. >> >> Fred >> >> -- >> FR >> >> - Original Message - >> From: "Matt Glasser" >> To: "SureFit Caret, and SuMS software users" >> Sent: Saturday, June 20, 2015 11:37:53 PM >> Subject: Re: [caret-users] Missing parts in MyelinMap+Different number of >> nodes! >> >> Yes please use the HCP pipelines for myelin mapping. >> >> Peace, >> >> Matt. >> >> On 6/19/15, 4:17 AM, "Donna Dierker" wrote: >> >>> Hi Xara, >>> >>> Because you are in the enviable position of having high res T1 AND T2, >>> you may be able to use the Human Connectome Project pipelines on your >>> data, rather than this (older generation) caret-based myelin mapping >>> script: >>> >>> http://www.humanconnectome.org/documentation/HCP-pipelines/ >>> >>> This also has the advantage of writing output files in workbench formats, >>> rather than older caret formats. I have a hunch that Matt Glasser will >>> point you in that direction (he is probably at OHBM conference or on his >>> way home). >>> >>> I confess I am not as familiar with this script as Matt is, but from what >>> I read at the link you sent, it appears the maps are on the native mesh, >>> rather than 164k. And I don't see myelin represented in the list of >>> freesurfer_to_fs_LR output files here: >>> >>> http://brainvis.wustl.edu/wiki/index.php/Caret:Operations/Freesurfer_to_fs >>> _LR/Output >>> >>> (These were independent effor
Re: [caret-users] negative values displayed positive
Hi FR, This is rather an odd file. I did 3dhistog on it and got this output: http://brainmap.wustl.edu/pub/donna/SPAIN/BCBL/your_file.3dhistog.txt login pub password download Mostly zeroes, but 14184 voxels set to 32766 (and no voxels set to any values in between 0 and 32766). And when I map this to the PALS atlas, I get nothing on most vertices, but solid yellow over a perisylvian blob. And when I click in the yellow, sure enough the ID window confirms the vertex is set to 32766 intensity. Working as expected, but data a bit off the beaten path. Donna On Jun 21, 2015, at 4:55 PM, Frédéric Roux wrote: > Hi Dona, > > thanks for getting back on this. > I am using NIFTI, so nothing rare from that side. > > Just uploaded the file via the link you sent me. > > Thanks for looking into this. > > -- > FR > > - Original Message - > From: "Donna Dierker" > To: "SureFit Caret, and SuMS software users" > Sent: Sunday, June 21, 2015 6:37:50 PM > Subject: Re: [caret-users] negative values displayed positive > > I'm not sure what would do this. What format is your volume? NIFTI? Older > .hdr/.img pairs sometimes got x-flipped upon opening under some > circumstances, but nothing I know of (except explicit *(-1) using volume > math) does this. > > You can use something AFNI 3dinfo or 3dhistog to double-check your volume's > values, or you can upload your volume here and I can do so tomorrow: > > http://brainvis.wustl.edu/cgi-bin/upload.cgi > > You can also have a look at the palette format, and just make sure it's not > the palette that is making your values appear to be flipped: > > http://brainvis.wustl.edu/CaretHelpAccount/caret5_help/file_formats/file_formats.html#paletteFile > > Caret has some built-in palettes (see D/C: Metric Settings menu). > > > On Jun 21, 2015, at 3:55 AM, Frédéric Roux wrote: > >> Dear all, >> >> I'm trying to map a functional volume to the PLAS-B12 >> surface but the sign of the values gets flipped so that >> negative values appear as positive ones. >> >> Does anyone know what could be wrong? >> >> Many thanks. >> >> Fred >> >> -- >> FR >> >> - Original Message - >> From: "Matt Glasser" >> To: "SureFit Caret, and SuMS software users" >> Sent: Saturday, June 20, 2015 11:37:53 PM >> Subject: Re: [caret-users] Missing parts in MyelinMap+Different number of >> nodes! >> >> Yes please use the HCP pipelines for myelin mapping. >> >> Peace, >> >> Matt. >> >> On 6/19/15, 4:17 AM, "Donna Dierker" wrote: >> >>> Hi Xara, >>> >>> Because you are in the enviable position of having high res T1 AND T2, >>> you may be able to use the Human Connectome Project pipelines on your >>> data, rather than this (older generation) caret-based myelin mapping >>> script: >>> >>> http://www.humanconnectome.org/documentation/HCP-pipelines/ >>> >>> This also has the advantage of writing output files in workbench formats, >>> rather than older caret formats. I have a hunch that Matt Glasser will >>> point you in that direction (he is probably at OHBM conference or on his >>> way home). >>> >>> I confess I am not as familiar with this script as Matt is, but from what >>> I read at the link you sent, it appears the maps are on the native mesh, >>> rather than 164k. And I don't see myelin represented in the list of >>> freesurfer_to_fs_LR output files here: >>> >>> http://brainvis.wustl.edu/wiki/index.php/Caret:Operations/Freesurfer_to_fs >>> _LR/Output >>> >>> (These were independent efforts going on at roughly the same time.) >>> >>> You may already have considered the HCP pipeline route, but if not, it is >>> worth a look -- you've got the T2's at the right res, which is often a >>> hitch. I suspect myelin maps on 164k won't be your only benefit. >>> >>> Donna >>> >>> >>> On Jun 19, 2015, at 8:58 AM, "Shams, Z." wrote: >>> >>>> Dear users, >>>> >>>> >>>> >>>> I just started working with Caret for creating myelin maps. I used HCP >>>> imaging protocols for data acquisition, both T1 and T2 data have the >>>> voxel size of 0.7mm isotropic. >>>> >>>> I followed the instructions in >>>> http://brainvis.wustl.edu/wiki/index.php/Caret:Operations/My
Re: [caret-users] negative values displayed positive
Hi Dona, thanks for getting back on this. I am using NIFTI, so nothing rare from that side. Just uploaded the file via the link you sent me. Thanks for looking into this. -- FR - Original Message - From: "Donna Dierker" To: "SureFit Caret, and SuMS software users" Sent: Sunday, June 21, 2015 6:37:50 PM Subject: Re: [caret-users] negative values displayed positive I'm not sure what would do this. What format is your volume? NIFTI? Older .hdr/.img pairs sometimes got x-flipped upon opening under some circumstances, but nothing I know of (except explicit *(-1) using volume math) does this. You can use something AFNI 3dinfo or 3dhistog to double-check your volume's values, or you can upload your volume here and I can do so tomorrow: http://brainvis.wustl.edu/cgi-bin/upload.cgi You can also have a look at the palette format, and just make sure it's not the palette that is making your values appear to be flipped: http://brainvis.wustl.edu/CaretHelpAccount/caret5_help/file_formats/file_formats.html#paletteFile Caret has some built-in palettes (see D/C: Metric Settings menu). On Jun 21, 2015, at 3:55 AM, Frédéric Roux wrote: > Dear all, > > I'm trying to map a functional volume to the PLAS-B12 > surface but the sign of the values gets flipped so that > negative values appear as positive ones. > > Does anyone know what could be wrong? > > Many thanks. > > Fred > > -- > FR > > - Original Message - > From: "Matt Glasser" > To: "SureFit Caret, and SuMS software users" > Sent: Saturday, June 20, 2015 11:37:53 PM > Subject: Re: [caret-users] Missing parts in MyelinMap+Different number of > nodes! > > Yes please use the HCP pipelines for myelin mapping. > > Peace, > > Matt. > > On 6/19/15, 4:17 AM, "Donna Dierker" wrote: > >> Hi Xara, >> >> Because you are in the enviable position of having high res T1 AND T2, >> you may be able to use the Human Connectome Project pipelines on your >> data, rather than this (older generation) caret-based myelin mapping >> script: >> >> http://www.humanconnectome.org/documentation/HCP-pipelines/ >> >> This also has the advantage of writing output files in workbench formats, >> rather than older caret formats. I have a hunch that Matt Glasser will >> point you in that direction (he is probably at OHBM conference or on his >> way home). >> >> I confess I am not as familiar with this script as Matt is, but from what >> I read at the link you sent, it appears the maps are on the native mesh, >> rather than 164k. And I don't see myelin represented in the list of >> freesurfer_to_fs_LR output files here: >> >> http://brainvis.wustl.edu/wiki/index.php/Caret:Operations/Freesurfer_to_fs >> _LR/Output >> >> (These were independent efforts going on at roughly the same time.) >> >> You may already have considered the HCP pipeline route, but if not, it is >> worth a look -- you've got the T2's at the right res, which is often a >> hitch. I suspect myelin maps on 164k won't be your only benefit. >> >> Donna >> >> >> On Jun 19, 2015, at 8:58 AM, "Shams, Z." wrote: >> >>> Dear users, >>> >>> >>> >>> I just started working with Caret for creating myelin maps. I used HCP >>> imaging protocols for data acquisition, both T1 and T2 data have the >>> voxel size of 0.7mm isotropic. >>> >>> I followed the instructions in >>> http://brainvis.wustl.edu/wiki/index.php/Caret:Operations/MyelinMapping. >>> >>> Actually I managed to display the four outputs, but some parts of the >>> maps are missing when overlaying on a surface(they¹re not complete maps: >>> either raw or corrected myelin maps). >>> >>> Another problem is that I used the freesurfer to fs LR script to >>> convert all my freesurfer files to the 164k_fs_LR mesh, >>> >>> but when I try to show the myelin maps on e.g. very-inflated 164k fs >>> LR, It fails with the error of containing a different number of nodes >>> thanŠ . >>> >>> >>> >>> I¹d appreciate any help and instructions. >>> >>> >>> >>> Thanks, >>> >>> >>> Xara >>> >>> ___ >>> caret-users mailing list >>> caret-users@brainvis.wustl.edu >>> http://brainvis.wustl.edu/mailman/listinfo/caret-users >> >> >> _
Re: [caret-users] negative values displayed positive
I'm not sure what would do this. What format is your volume? NIFTI? Older .hdr/.img pairs sometimes got x-flipped upon opening under some circumstances, but nothing I know of (except explicit *(-1) using volume math) does this. You can use something AFNI 3dinfo or 3dhistog to double-check your volume's values, or you can upload your volume here and I can do so tomorrow: http://brainvis.wustl.edu/cgi-bin/upload.cgi You can also have a look at the palette format, and just make sure it's not the palette that is making your values appear to be flipped: http://brainvis.wustl.edu/CaretHelpAccount/caret5_help/file_formats/file_formats.html#paletteFile Caret has some built-in palettes (see D/C: Metric Settings menu). On Jun 21, 2015, at 3:55 AM, Frédéric Roux wrote: > Dear all, > > I'm trying to map a functional volume to the PLAS-B12 > surface but the sign of the values gets flipped so that > negative values appear as positive ones. > > Does anyone know what could be wrong? > > Many thanks. > > Fred > > -- > FR > > - Original Message - > From: "Matt Glasser" > To: "SureFit Caret, and SuMS software users" > Sent: Saturday, June 20, 2015 11:37:53 PM > Subject: Re: [caret-users] Missing parts in MyelinMap+Different number of > nodes! > > Yes please use the HCP pipelines for myelin mapping. > > Peace, > > Matt. > > On 6/19/15, 4:17 AM, "Donna Dierker" wrote: > >> Hi Xara, >> >> Because you are in the enviable position of having high res T1 AND T2, >> you may be able to use the Human Connectome Project pipelines on your >> data, rather than this (older generation) caret-based myelin mapping >> script: >> >> http://www.humanconnectome.org/documentation/HCP-pipelines/ >> >> This also has the advantage of writing output files in workbench formats, >> rather than older caret formats. I have a hunch that Matt Glasser will >> point you in that direction (he is probably at OHBM conference or on his >> way home). >> >> I confess I am not as familiar with this script as Matt is, but from what >> I read at the link you sent, it appears the maps are on the native mesh, >> rather than 164k. And I don't see myelin represented in the list of >> freesurfer_to_fs_LR output files here: >> >> http://brainvis.wustl.edu/wiki/index.php/Caret:Operations/Freesurfer_to_fs >> _LR/Output >> >> (These were independent efforts going on at roughly the same time.) >> >> You may already have considered the HCP pipeline route, but if not, it is >> worth a look -- you've got the T2's at the right res, which is often a >> hitch. I suspect myelin maps on 164k won't be your only benefit. >> >> Donna >> >> >> On Jun 19, 2015, at 8:58 AM, "Shams, Z." wrote: >> >>> Dear users, >>> >>> >>> >>> I just started working with Caret for creating myelin maps. I used HCP >>> imaging protocols for data acquisition, both T1 and T2 data have the >>> voxel size of 0.7mm isotropic. >>> >>> I followed the instructions in >>> http://brainvis.wustl.edu/wiki/index.php/Caret:Operations/MyelinMapping. >>> >>> Actually I managed to display the four outputs, but some parts of the >>> maps are missing when overlaying on a surface(they¹re not complete maps: >>> either raw or corrected myelin maps). >>> >>> Another problem is that I used the freesurfer to fs LR script to >>> convert all my freesurfer files to the 164k_fs_LR mesh, >>> >>> but when I try to show the myelin maps on e.g. very-inflated 164k fs >>> LR, It fails with the error of containing a different number of nodes >>> thanŠ . >>> >>> >>> >>> I¹d appreciate any help and instructions. >>> >>> >>> >>> Thanks, >>> >>> >>> Xara >>> >>> ___ >>> caret-users mailing list >>> caret-users@brainvis.wustl.edu >>> http://brainvis.wustl.edu/mailman/listinfo/caret-users >> >> >> ___ >> caret-users mailing list >> caret-users@brainvis.wustl.edu >> http://brainvis.wustl.edu/mailman/listinfo/caret-users > > > > ___ > caret-users mailing list > caret-users@brainvis.wustl.edu > http://brainvis.wustl.edu/mailman/listinfo/caret-users > > ___ > caret-users mailing list > caret-users@brainvis.wustl.edu > http://brainvis.wustl.edu/mailman/listinfo/caret-users ___ caret-users mailing list caret-users@brainvis.wustl.edu http://brainvis.wustl.edu/mailman/listinfo/caret-users
