Dear Matthew, thanks for your reply and tips! Naturally, I have a couple more questions.
1. What exactly the order of the pipelines would be? (I am still stuck at the functional preprocessing right now, so I'm not that far yet). Would it be both functional ones (volume and surface), then ICA+FIX and then MSMAll? 2. I'm sorry but I also didn't quite get the procedure for th e -cifti-average-roi-correlation. I only have one run per subject so I don't need to average the runs. So when I have the individual outputs after all the above mentioned pipelines (.dtseries.nii, right?) do I just use all of them (for one group) with a -cifti flag for each as input files in one command or do I run the command for every single participant of a group and afterwards merge the outputs? I want to do group-level comparison so at some step I need to create group maps. Or am I misunderstanding something? Sorry for such basic confused questions and thanks a lot! Lisa On 5 April 2017 at 21:05, Glasser, Matthew <[email protected]> wrote: > I recommend you use the MSMAll aligned, ICA+FIX denoised data and use > wb_command -cifti-average-roi-correlation. You may or may not chose to > do something like global signal regression to clean up residual global > artifact in the data (which ICA+FIX is not designed to remove) depending on > if you think leaving in global signals will create a bigger positive bias > than removing the mean of the RSNs will create a negative bias (or just > analyze things both ways). We are working on a better solution for this > issue that does not require removing the mean of the RSNs when cleaning up > global artifact (i.e. remove the positive bias in connectivity without > adding in a negative bias). > > You can make a -vol-roi of the hippocampus by extracting it from this file > ${StudyFolder}/${Subject}/MNINonLinear/Results/Atlas_ROIs.2.nii.gz. I > would use a -cifti flag for each run of a given subject, but run separate > commands per subject to generate one dense scalar correlation map per > subject. You can then do statistics on these maps (e.g. with the FSL PALM > software tool). You may wish to do the Fisher transform on the correlation > maps first with wb_command -cifti-math “atanh(x)” <output> -var x <input> > > Peace, > > Matt. > > From: <[email protected]> on behalf of Lisa > Kramarenko <[email protected]> > Date: Wednesday, April 5, 2017 at 4:40 AM > To: "[email protected]" <[email protected]> > Subject: [HCP-Users] Help with the group comparison of seed-based FC > > Hello dear experts, > > I am very new to HCP so I am struggling with a lot of confusion and hope > you can help. I would like to calculate seed-based FC of hippocampus of two > groups (patients/controls) and to perform a group comparison between them. > Now I am not sure about how to proceed. I see two possible ways: > > 1. I could merge .dtseries.nii files for all the subjects in a group with > cifti-merge to create a group-average dense connectome. However, how do I > extract the connectivity of the seed of interest and how do I I perform > statistical analysis on it? > 2. Or, if I understand correctly, I can use -cifti-average-roi-correlation. > However, I am not sure about the inputs. Should I first merge .dtseries.nii > files for all subjects in a group, take this as <cifti-in> and then extract > hippocampus from Atlas_ROIs.2.nii.gz with cifti-separate and use it > as <roi-vol>? Second question is when I managed to run it, what would the > output be and how do I perform statistical analysis on it? > > I would be super grateful if you could clarify what the right way is and > give me a short step-by-step of how to do a seed-based group-level analysis. > > Thanks a lot! > Lisa > > _______________________________________________ > HCP-Users mailing list > [email protected] > http://lists.humanconnectome.org/mailman/listinfo/hcp-users > > > ------------------------------ > > The materials in this message are private and may contain Protected > Healthcare Information or other information of a sensitive nature. If you > are not the intended recipient, be advised that any unauthorized use, > disclosure, copying or the taking of any action in reliance on the contents > of this information is strictly prohibited. If you have received this email > in error, please immediately notify the sender via telephone or return mail. > _______________________________________________ HCP-Users mailing list [email protected] http://lists.humanconnectome.org/mailman/listinfo/hcp-users
