Which segmentation algorithm and which version of SPM? With many MR
sequences, most versions do not do an especially good job at
identifying CSF, although the "new segment" toolbox may be slightly
more accurate because it includes some information about non-brain
tissues that lie adjacent to CSF. Maybe also take a look at the
segmentation results to see how they appear visually.
Best regards,
-John
On 8 December 2010 17:21, R Duke <[log in to unmask]> wrote:
> Hello SPM Experts,
>
> I have recently collected some segmentation data (modulated normalised) on about 29 of my subjects (age = 60 to 69). After calculating ICV (GM+WM+CSF) I calculated the percentage of each tissue type (GM/ICV), and my results yielded some interesting values. It appears as if my CSF volume makes up about 18 to 41% (mean = 32%) of total ICV (see values below). My question is related to the high percentage of CSF volume. Are these values what other people see? Is this a common finding in "older" healthy adults using this methodology? Also, if anyone would be able to suggest some literature I would greatly appreciate it. I found one paper describing 'normative' values in healthy adults (Kruggel-2006-MRI-based volumetry of head compartments: Normative values of healthy adults), but the methodology that was used is different. Any suggestions would be great.
>
> Cheers!
>
> CSF/TBV
> 0.369677793
> 0.316596805
> 0.250886675
> 0.343763644
> 0.362200567
> 0.319824008
> 0.34150398
> 0.323900256
> 0.39456022
> 0.3267729
> 0.273886916
> 0.255621697
> 0.347943037
> 0.297922525
> 0.349123208
> 0.315401937
> 0.386075517
> 0.186514505
> 0.384746976
> 0.275407774
> 0.304729
> 0.377143486
> 0.393461728
> 0.242075638
> 0.294585911
> 0.269121855
> 0.312448977
> 0.412999751
> 0.311132561
>
|