Hi Anastasia,
There are several possible explanations. Could you please check if one
of the following has happened:
- you used a different bvalue in a subset of the subjects
- you specified a different bvalue for a subset of the subjects when
using dtifit
- you used a different echo time for different acquisitions in a
subset of the subjects
- you added the b=0 (instead of averaging them) for a subset of the
subjects
- you added the b!=0 (instead of averaging them) for a subset of the
subjects
(note: in the last two cases, if you have two repeats, it will create
a two-fold increase/decrease in MD)
Alternatively, there might be some signal drop in your data. But in
order for this to affect MD alone, it needs to be present in all
directions... It's worth looking at the data though.
Let us know if none if the above explain your finding..
Cheers,
Saad.
On 29 Jun 2009, at 19:20, Anastasia A Ford wrote:
> Hi,
>
> I am running TBSS-style analysis on MD maps for patients and
> controls. After
> all MD maps were registered to standard space, I run the movie loop
> to see
> how well the skeleton from the FA analysis fits each subject. I
> noticed that
> some of the subject have much higher (often twice the value of other
> subjects) MD values throughout the brain, compared to other
> subjects. These
> differences are seen in both controls and patient groups. The
> subjects that
> show high MD values to not seem to show any abnormalities in the FA
> values.
> What would be a possible explanation for these increased MD values?
>
> Sincerely,
>
> Anastasia
>
--
Saad Jbabdi
Oxford University FMRIB Centre
JR Hospital, Headington, OX3 9DU, UK
+44 (0) 1865 222545 (fax 717)
www.fmrib.ox.ac.uk/~saad
|