Multiple scans per visit - where to start???

Questions about BrainVisa usage and installation

Moderators: denghien, riviere

Post Reply
ruslana
Posts: 14
Joined: Thu Mar 31, 2016 9:19 pm

Multiple scans per visit - where to start???

Post by ruslana »

Hi all,

I have a dataset with multiple scans per visit, and I need help to process this data with BrainVisa.
As far as I understand, the scans from the same visit should be averaged and co-registered with motion correction.
I was trying to import the data as is, but was not sure how to handle this - it seems like I can only have 1 scan per subject under each acquisition session. I could probably use various acquisition labels for multiple scans to store the data, but in my case the data is longitudinal, so I have to use this label to differentiate between different sessions. No idea how to differentiate between different scans in the same session then. :oops:
Does BrainVisa support such a data or should I pre-process it to create one average image with some external tool before importing it to the BrainVisa?
What would be the procedure in BrainVisa or what tools could I use to pre-process such a dataset?
Any advice would be much appreciated, since I have really no idea where to start!

Regards,
Ruslana.
User avatar
riviere
Site Admin
Posts: 1361
Joined: Tue Jan 06, 2004 12:21 pm
Location: CEA NeuroSpin, Saint Aubin, France
Contact:

Re: Multiple scans per visit - where to start???

Post by riviere »

Hi,

If I understand, you have longitudinal acquisitions, and several scans for each timepoint ?
I must admit that the current organization of Mosphologist does not support well this kind of schema. It is always possible to define a new files organization schema, but it is a fastidious work and you probably don't want to do that. Maybe the best option is to use the acquisition field to store both information, like acquisition=<timepoint>_<scan>.

Now, for longitudinal studies especially, we have added the possibility to store database attributes which are not part of the file names (such attributes are stored in json files). It allows to record more information without changing the files organization. There is another import process for T1 MRIs, "Import T1 MRI, longitudinal version" (it can be found both in morphologist/importation or in data management/importation/T1 MRI). This process allows to specify former attributes (center, subject, acquisition), but also time_point, time_duration, rescan and acquisition_date, which are additional attributes stored in a file. You have to use distinct names for acquisition in a single subject (you can use the rule above, say <time_point>_<scan>) and use time_point and rescan to easily distinguish scans and time points.
However I must also say that this additional attributes system has finally not been used for our longitudinal studies, so has remained experimental, and is not fully tested, so it may cause problems at some points (there are ambiguities actually on where to write these attributes files, especially when writing, importing or copying data between databases). You can try, see if it works for your needs.

Denis
ruslana
Posts: 14
Joined: Thu Mar 31, 2016 9:19 pm

Re: Multiple scans per visit - where to start???

Post by ruslana »

Denis,
Thanks a lot for your reply.
Yes, you understood correct - I do have longitudinal acquisitions, with several scans for each timepoint.
I think using the acquisition field to store both timepoint and scan, indeed may be a good solution to try.
My other concern is, however, if the BrainVisa supports processing these few scans (from the same timepoint) to create one image per visit - averaged and coregistered with motion correction? I could not find in documentation how to do this... Could you please advise?
Thanks a lot!
Ruslana.
User avatar
riviere
Site Admin
Posts: 1361
Joined: Tue Jan 06, 2004 12:21 pm
Location: CEA NeuroSpin, Saint Aubin, France
Contact:

Re: Multiple scans per visit - where to start???

Post by riviere »

Sorry I forgot to answer the second part of the question: no we have not developed motion correction and coregistration methods. Other groups and software have already spent lots of efforts on this topic, so we rather rely on external software for this.
Denis
Post Reply