Dear Lynne and Ged --
Yes I agree with Ged -- the ADNI dataset is a very nice one to try
out TBM on -- there will soon be some example TBM results available
in the ADNI database too.
We are finishing a TBM analysis of the ADNI scans - we'll upload the
results to the ADNI database at LONI (linked by Ged below). By
results, I mean the Jacobian maps that index the degree of atrophy in
each subject, relative to a common (average) template, or the same
subject's baseline scan. There'll also be the atlas template used for
normalization and some ROIs. So, although the data aren't there now,
in a couple of months there'll be a TBM-based map of relative atrophy
(per scan) uploaded back into the database. The main reason for doing
this is in case others developing nonlinear image registration
programs want to see if they get a similar answer - they could also
look at the maps to see if they can detect disease or cognitive
effects with greater power. People could also use the maps of atrophy
as part of a new analysis, e.g. for correlation with PET, serum
markers, or maybe developing a joint predictor of future conversion
to dementia, without having to compute the whole analysis again from
the raw data, if they don't want to (as it takes a while...). Other
ADNI analysis sites will be uploading their own results too, such as
processed images and statistical maps from VBM, cortical surface
mapping, hippocampal shape analysis, etc.
Paul
--
Paul Thompson
Professor of Neurology
Lab of Neuro Imaging
UCLA School of Medicine, Los Angeles, CA 90095, USA
http://www.loni.ucla.edu/~thompson/thompson.html
http://www.loni.ucla.edu/~thompson/projects.html
On Aug 12, 2007, at 4:52 AM, Ged Ridgway wrote:
> Hi Lynne,
>
>> I'm trying to learn longitudinal DBM and was wondering if anyone
>> has available a sample data set with results (even a single
>> subject would do) to verify that I am doing the method correctly.
>
> I don't know of any publicly available data-sets with DBM results
> available... Also, I think the term DBM seems to be used in
> slightly different ways by different researchers, and results might
> depend quite strongly on the software used, and the particular
> options etc. within it, so verifying your method will not be easy.
>
> The best, relatively easily available, longitudinal (structural)
> MRI data that I'm aware of is from the Alzheimer's Disease
> Neuroimaging Initiative:
> http://www.loni.ucla.edu/ADNI/
> Some groups have almost certainly run some kind of TBM/DBM study on
> some of these subjects, but I'm afraid I don't know the details.
> Maybe look for further work from some of the authors of this ADNI
> TBM preparatory study:
> http://dx.doi.org/10.1016/j.neuroimage.2005.12.013
>
> The OASIS project also promises longitudinal data "coming soon"
> http://www.oasis-brains.org/
>
> I hope that helps a bit, sorry if it's not the answer you hoped for!
>
> Ged.
|