If white matter is lost, it is difficult to predict exactly what would
take its place within the intracranial volume.  Perhaps a bit more CSF
would enter into the ventricles, or maybe there would be a bit more
CSF outside the brain.

I'll let the proper neuroscientists/medics provide more authoritative
answers to your questions about whether changes are permanent or
transient, but I would imagine that inflammation may cause apparent
tissue expansion and dehydration could cause apparent shrinkage.  Some
of the papers by Bydder and Hajnal could be useful in terms of global
brain shrinkage.

You may also be interested in this one, which seems to show a rather
surprising cause of volume loss...
Effects of Aβ immunization (AN1792) on MRI measures of cerebral volume
in Alzheimer disease
NC Fox, RS Black, S Gilman, MN Rossor… - Neurology, 2005

Signal intensity changes in the MRI could also result in differences
in the way the data is segmented, which could lead to apparent
growth/atrophy.  For example, a small stroke could change the signal
intensity of WM so that it is segmented as GM.  One of my least
favourite confounding effects is the one that comes from partial
volume - particularly around the ventricles.  Voxels that contain a
mix of CSF and WM often have a signal intensity similar to GM, so
expansion of the ventricles can appear as if there is more GM - simply
because the surface area is larger.

Best regards,
-john


On 20 September 2011 16:18, Becky Haynes <[log in to unmask]> wrote:
> I've done a VBM anaysis using DARTEL, and looked at grey and white matter volume.
>
> My questions is how to interpret white matter volume loss. I can conceptualise that grey matter loss, would be replaced by the expansion of CSF. Would this also be the case in white matter?
>
> Also are these changes generally considered to be permanent or transient? and are they likely to be affected by inflammatory processes? if anyone can point me in the direction of any references that would be great.
>
> Thanks
>