Hi,
In general the DTI model is valid for protocols that contain low-b and high-b value acqusitions. The b value that then appears in the DTI equation is the difference between the low and high b factors (when two of them are used), so that should be reflected in the bvals specification. In your case you should replace the b=0 entries with b=100 in the bvals files (alternatively, it would be OK to have b=0 entries, but the high-b ones should be 900).
An effect of using a low-b reference signal rather than a b=0, will be that the S0 estimate will represent the signal intensity at b=100, rather than at b=0, i.e. the S0 is underestimated. Furthermore, the signal attenuation (due to diffusion) modelled will be smaller than when a b=0 is used as a reference, so that should result in slightly overestimated diffusivity values.
Cheers,
Stam
On 14 Nov 2010, at 16:47, Tim Coolen wrote:
> Hi all!
>
> I have a question regarding the interpretability of FA and MD maps constructed from the following DTI acquisition sequence.
> The first so-called B0 volumes are in fact diffusion-weighted with b=100, the actual diffusion sequence having b=1000.
> Unaware of that at first, I used DTFIT with 0, 0, 0 as bvalues for the first volumes and got FA values that were compatible with the tissue types. Moreover, during my TBSS analysis, a correct FA skeleton was successfully obtained with threshold = 0,2.
> How did that affect my results? Will FSL manage "B0" volumes with non zero bvalues?
> Is the present analysis valid, even though the bvalues aren't correct?
>
>
> Thank you for you help!
>
|