Hi Timothy, sorry for the late response. Thanks, this helped. I ran the command you wrote but used the value of 1.3(=p-value of 0.05). After this, when using cifti-stats with the created ROI the output values were different from when I used the whole output of PALM as ROI.
Thanks again for your help! Lisa On 21 October 2017 at 06:34, Timothy Coalson <tsc...@mst.edu> wrote: > I think I misunderstood the problem, a single ROI should work fine. > > Are you talking about thresholding in the display settings (the wrench > icon on the layer in wb_view)? That doesn't actually threshold the file > into an ROI, that just changes the display of it to not color the > outside-threshold values. To actually threshold it to make an ROI file, > you need to do something like this: > > wb_command -cifti-math 'x > 5.3' ROI.dscalar.nii -var x > inputstatistic.dscalar.nii > > Tim > > > On Thu, Oct 19, 2017 at 2:02 AM, Lisa Kramarenko < > lisa.kramare...@gmail.com> wrote: > >> Hey Timothy, >> >> thanks for your response. Before continuing with the questions about the >> exact implementation I am not very sure as to why you refer to my ROIs in >> plural. As ROI, I used the dscalar file which I got as an output of PALM, >> i.e. it is a file showing where there was a significant difference between >> two groups. I threshholded it in the Workbench and saved the changes, so >> that the file now only shows the really significantly different areas. In >> my understanding this is only one ROI/one map. Do I understand it wrongly? >> I can attach the file if you want. >> >> I just want to see what the mean myelination in these exact areas that >> turned out to be significant is in the groups I compared. Hope it makes >> sense. >> >> Thanks again! >> >> Lisa >> >> On 18 October 2017 at 01:12, Timothy Coalson <tsc...@mst.edu> wrote: >> >>> The -cifti-parcellate command is currently the easier way to do this, if >>> your ROIs don't overlap. The -roi option to -cifti-stats currently uses >>> only the first map, and only tests for whether the value is greater than 0, >>> so probably the first map in your roi input is actually all positive. Also >>> note that MEAN does not take a dash in front of it, which would cause an >>> error. >>> >>> Tim >>> >>> >>> On Tue, Oct 17, 2017 at 5:57 PM, Glasser, Matthew <glass...@wustl.edu> >>> wrote: >>> >>>> It might only be using the first ROI. Tim will know better what is the >>>> issue. You might try converting the ROIs to a parcellation and then >>>> parcellating the myelin map and converting the parcellated file to text or >>>> whatever you need. >>>> >>>> Peace, >>>> >>>> Matt. >>>> >>>> From: <hcp-users-boun...@humanconnectome.org> on behalf of Lisa >>>> Kramarenko <lisa.kramare...@gmail.com> >>>> Date: Tuesday, October 17, 2017 at 7:11 AM >>>> To: "hcp-users@humanconnectome.org" <hcp-users@humanconnectome.org> >>>> Subject: [HCP-Users] Extraction of mean myelination values from a >>>> custom roi >>>> >>>> Dear experts, >>>> >>>> I have a question about extracting average myelination value from a >>>> custom roi. What I did was the following: >>>> >>>> wb_command -cifti-stats merged_myelin_maps_group1.dscalar.nii -reduce >>>> -MEAN -roi file_with_significant_differences_between_the_groups_thresho >>>> lded_at_1.3.dscalar.nii >>>> >>>> I get a series of numbers (for every of the merged maps), however they >>>> don't change if I omit the roi part which leads me to believe that the >>>> average is calculated across the whole cortex and not specifically in areas >>>> I want to specify. >>>> >>>> What would be the best way to find out the mean myelination values in a >>>> specific roi (which is defined by the file with significantly different >>>> areas)? >>>> >>>> I hope you understand what I mean. >>>> >>>> Thanks a lot for your help! >>>> >>>> Best, >>>> Lisa >>>> >>>> _______________________________________________ >>>> 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
