Dear all,
I have been searching previous posts trying to piece together a pipeline for longitudinal imaging. What I have so far is below. I would appreciate any thoughts or comments that may help me progress. As a precursor, I am working with pre-clinical rodent mri so I have adjusted the steps accordingly (using the SPMmouse plug-in).
- Bulk manual register the images using SPMmouse - Both scan 1 (post-natal day 21, s1) and 2 (post-natal day 35, s2) to rat template (already in paxinos & Watson space) -> Both images s1 & s2 are now in template space & roughly aligned
(From what i've read it is also common to register scan1 to scan2, or vice versa - I am unsure as to which is the best method here)
- HDW to calculate jacobian determinants (j*). s1 = stationary image, s2= moved image
-> Creates jacobian (encodes relative volume change)
- Segment early scans (s1) using segmentation (pre-proc: segment - SPMmouse)
-> This will provide GM,WM & CSF maps in template space of the early scan time-point
- Using Utils:Image Calculator;
i1.*(i2-1)
Where i1 is the stationary image (s1) and i2 is the jacobian from the HDW (mapping s1 to s2).
Dartel: intial import - using the GM,WM,CSF from the initial segmentation of scan1 time-point above
Run Dartel: generetate flow fields - These flow fields encode the deformation of the segmented images back to the average template generated through the Dartel procedure.
-- At this point, previous posts have used the "normalise to MNI space' - in Dartel tools. However, I am not sure whether this module has MNI as its default and therefore I cannot use this to register the rodent brains to the rodent template I have already. Instead could I use;
Create Warped: Flow fields - from dartel
Images to be warped to average template are the output of the Image calculator (jacobian scaled grey matter).
Does this mean that I now have jacobian scaled grey matter images (of the second scan time-point) all registered to the average template generated in DARTEL (which was registered initially to the TPM I specified in the previous segmentation (prior to DARTEL)?
In my expectation I would need to then register these to the initial template so that they are in atlas space (or would they already be so because of the initial registration to the template for the first segmentation of which are the precursors files for the Dartel analysis)
I've ran exactly this and from the Image calculator I get a 'HDR' and 'Disc Image File' which I cannot open (it says the file is corrupted). I'm currently working on getting past this issue. I have tried various data type in the the image calculator - 1) INT26 & FLOAT64-double prec. float
I would be extremely grateful if you could shed any light on the matter. If anything needs clarifying please do let me know. Thanks in advance. Kind Regards, Jolyon.
|