Hi
> Den 15. jun. 2018 kl. 11.05 skrev SUBSCRIBE SPM Wendel Liou <[log in to unmask]>:
>
> Dear all,
>
> I collect some DWI images, and I get some questions when using SPM.
>
> MR imaging protocol
> 1) Information of MR imaging sequence:
> T2WI(b-value=0 s/(mm^2)) and DWI(b-value=200 s/(mm^2)) were acquired alternatively.
> The direction of the diffusion gradient is fixed at X axes.
> So, the acquisition of the EPI series is T2WI-DXIx-T2WI-DWIx.... for the 150 excitations. (i.e, after a whole acquisition will get 75 T2WI and 75 DWI )
>
> 2) Then, I use this sequence to perform two experiments
> a.
> Each excitation obtained 21 axial slices covering from the cerebellum down to the angle of the jaw.
> Slice thickness: 5mm
> Slice gap: 0mm
> TR: 2s
> TE: 53.3 ms
>
> b.
> Each excitation obtained 11 axial slices covering from the cerebellum down to the angle of the jaw.
> Slice thickness: 5mm
> Slice gap: 5mm
> TR: 1.053s
> TE: 53.3 ms
>
> Question
> 1) Slice timing
>
> My data consists of T2WI images and DWI images because the sequence switches the diffusion gradient alternatively.
> Should I separate my data into two group: T2WI and DWI ?
That would probably make sense as you don’t want to mix T2WI and DWI when doing the interpolation needed for slice timing
> Then, how about the value of TR and TA field?
You can use the same TR and TA as you would have done in a setting with pure T2WI. The information is only needed to find out how the slices were distributed inside the TR used to acquire a given volume.
> Is TR field same as my parameter (i.e TR field=2 for exp.a and TR field=1.053 for exp.b)?
> Can I use TR-(TR/slice) for TA field?
yep if you don’t have a large break in the end of each TR, alternatively consider using the alternative option of specifying acquisition times (if you have siemens dicoms they can be found like this):
hdr=spm_dicom_headers(‘YourDicomFile.dcm’);
timings=hdr{1}.Private_0019_1029
> And, whether slice gap influence the quality or not?
>
no impact
> 2)Realign
>
> Because the coverage of my data is nearly neck images, not brain images. Can I use Realign in SPM for motion correction?
give it a try I think it will work, spatial normalisation might be tricky though.
Best
Torben
|