Hi,
On 27 Aug 2009, at 21:05, Michael Scheel wrote:
> Dear fsl experts,
>
> I'm preparing a case report on a patient with a neurodegenerative
> disease affecting mainly the white matter.
> I want to comare his FA map with 20 controls (healthy and similar
> age). I was using the TBSS script to align the FA maps of Patient
> and controls.
> My question is: If I want to compare the FA map of the patient with
> the controls should I use the randomise tool with the 20 controls
> being one group and the patient being the 2nd group. Is that a valid
> analysis to say something about where the patient differs from
> normal controls?
Yes - it's ok but you only have 20 distinct permutations so your p-
values won't look good! You may be able to just interpret the raw
tstat image to get the information you want.
> I also have data for the patient after one year of treatment and a
> rescan on 2 of the controls.
> I want to ask: Did the treatment reversed some of the differences in
> FA map differences?
> My original idea was
> 1) Compare Patient_Before_Treatment with Controls and then Compare
> Patient_After_Treatment with Controls.
> 2) Compare the resulting tstat of Con_vs_BeforeTreatment with
> Con_vs_AfterTreatment
If by the latter you mean compare the longitudinal change in the
patient with corresponding longitudinal changes in controls then yes
that's fine. I don't think option 1 will tell you anything useful.
Cheers.
>
> I did the above analysis and have promising and hypothesis conform
> results but I have my doubts on the statistical validity.
> Can you give me some advice on this? Is randomise the right tool to
> compare single subjects agains a control group?
>
> Thanks, Michael
>
---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director, Oxford University FMRIB Centre
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------
|