Hi -
  We've been using the IBSR-18 dataset to evaluate our segmentation results, and were curious if the ages of the 18 participants are available anywhere. Any help would be greatly appreciated.
All the best,
Paul Fillmore

Andy.

Thanks Andy, much appreciated. The readme that is distributed with the "IBSR_V2.0 skull-stripped NIfTI" download doesn't list the ages, but I see now that the other readme file does - sorry for missing that. So in this case I'm guessing that "JUV" stands for "juvenille"? Are these ages just unknown, or are they a catch all for some range under 7 years? We're doing work related to creation of age-appropriate MRI templates and evaluating the errors caused by use of segmentation priors from very different ages, so you can see how detailed age info would be crucial. Also, is there an easy way to translate the IBSR20 images into NIFTI format? As those images seem to be from a more homogenous age range than the IBSR18, those might also be quite useful. It's great that you guys have put together this resource for the neuroimaging community, it's very helpful!
All the best,
Paul

You're right, Paul.  "JUV" means an unknown aged juvenile.  You might also be interested in the CANDI scans here on NITRC, and while it is not free, the labeled scans from Neuromorphometrics (full disclosure: that was a shameless plug for my company!)

The IBSR20 scans are available in ANALYZE format which is the precursor to NIFTI so they should be easily converted (even with the sample code from http://nifti.nimh.nih.gov/nifti-1/).  Setting the proper orientations will take some additional work though.  Has anybody done this already?  If so, I'd be happy to add that here.

Andy.

Thanks Andy - I'll definitely check out CANDI, and will also suggest the possibility of using Neuromorphometrics to our group here. Do you know anything (or now of a good resource) about the history of the two IBSR datasets? Were they all part of one large study, done on the same scanner, or are they from multiple places, or just a collection of clinical scans? I've just learned of the dataset recently, and it's a bit tough to piece it together from the papers I've been able to find. I'll keep you posted if I get the NIFTI conversion done successfully. Thanks!

Hi Paul,

The somewhat random collection of labeled brain scans was put together by David Kennedy starting in 1996 at the Center for Morphometric Analysis (Massachusetts General Hospital).  I did the web site (which now redirects to NITRC).  It was not easy to convince the PIs of the various studies to give away their expensive results for nothing.  It took weeks of work to label each scan.  One of the stipulations was that not much information was to be given about the scans so all there is to know can be found in the associated README files.  The scans were from various studies, but all, as far as I know, came from the same couple of scanners.  

Andy.

Thanks Andy - that definitely helps give it some context. I appreciate your help.