Hi Matt Thanks for the reply! I have taken a look at the FAQ already, but I'm still not sure how to use the file.
I've tried loading a *_Atlas.dtseries.nii in matlab using https://github.com/Washington-University/cifti-matlab and in python using the https://github.com/nipy/nibabel/pull/249 addition to nibabel. In matlab, there's a 96854 x 284 dtseries matrix which I assume is the data - in python theres a 1 x 1 x 1 x 1 x 284 x 91282 data matrix. 284 corresponds to the number of frames, but what is the other dimensions exactly? And what is the reasons for the discrepancy between when loading to python and matlab? (I realise you have not implemented either of the libraries, but it might be something independent of the implementation that you know of) I apologize if my questions are somewhat daft - this is a new area for me. Thank you again for the assistance. Best regards Philip On Fri, May 29, 2015 at 5:33 PM Glasser, Matthew <[email protected]> wrote: > Here is a page that should be helpful to getting started: > > https://wiki.humanconnectome.org/display/PublicData/HCP+Users+FAQ > > Peace, > > Matt. > > From: Philip Henningsen <[email protected]> > Date: Friday, May 29, 2015 at 7:29 AM > To: "[email protected] Users" <[email protected]> > Subject: [HCP-Users] Help understanding *_Atlas.dtseries.nii > > HI HCP experts > > I'm trying to understand how the *_Atlas.dtseries.nii files are > structured exactly. As I understand it, they are the fMRI images mapped to > coordinates in the brain, but I'm having trouble understanding exactly how > to use them. I haven't been successful in finding more detailed information > about it on the HCP website, but I might simply not know the correct place > to look. If anyone can point me in the correct direction, it would be much > appreciated. > > Some context: > I'm attempting to classify the actions done in the MOTOR task using > various forms of neural networks. So far, I've tried using the sequences of > fMRI data (the 284x91x109x91 .nii files), but have not achieved better than > random classification accuracy. A colleague told me that the > *_Atlas.dtseries.nii files contained the fMRI data mapped to a 2D atlas, > which would enable me to do 2D convolution without losing information (as I > would by e.g. collapsing the 3D images to 2D by taking the mean of a > direction). > > Thank you for the attention. > > Best regards > > Philip > > _______________________________________________ > 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
