John,
If you have an M0 image that is the same as the calibration image (and is commonly a proton density weighted image). If there are multiple volumes you can feed them all into oxford_asl (within one image) and it will take the mean for you. The key thing is that you need to know some information about your M0 image, what the parameters were for its acquisition and whether any preprocessing has been done to it. This might be quite important in your case as CBF values in the range 150-250 are too high for absolute perfusion - we might expect mean grey matter to be in the range 30-50 from a typical ASL image, with values in some places potentially approaching ~100. I am afraid you cannot really tell anything from the numerical values in your perfusion.nii.gz image, these depend entirely on your scanner - hence the need for the calibration image. It is not immediately obvious to me, though, why you get different values in your perfusion.nii.gz image when you add the calibration image - as the two shouldn’t interact, unless this is triggering another correction to be performed.
Michael
> On 30 Sep 2017, at 00:52, John anderson <[log in to unmask]> wrote:
>
> Dear Dr Michael your detailed responses are highly appreciated!! Thank you so much! Kindly, I have one additional question and I appreciate the response as well!
>
> Regarding calibration using the flag -c in the command oxford_asl: I don't have proton density image and I am not sure what is the meaning of mean control image? Can I feed M0 (two columes merged together) to this flag?
> When I feed M0 to the flag -c in the command oxford_asl. I get perfusion_calib.nii.gz images and the range of CBF in these images is ~150-250 while in the perfusion.nii.gz images, the CBF range is ~15-30. Are these ranges correct? Notably, when I don't include flag -c in oxford_als the range of CBF in the perfusion.nii.gz images is ~40-90 which is close to real CBF?
>
>
>
>
>
>
>
> I have three questions regarding oxford_asl which i am using to process pcasl data
> 1. Can I use eddy current correct on pcasl data to correct for motion
> This is not for ASL, but specifically for diffusion data. You would be better doing standard motion correction and distorition correction as appropriate.
>
> 2. Are the final CBF maps (perfusion_calib.nii.gz) represent an absolute cbf within each voxel. If so, can I normalize these images to whole brain mean using fslmaths and the flag -inm 1 to get the % of cbf within each voxel freom whole brain cbf
> perfusion_calib should be in ml/100g/min and represent absolute CBF. YOu could then normalise these (and in fact oxford_asl can do that for you), rather than whole brain I would probably recommend using a grey matter ROI (the approach that oxford_asl does when doing normalisation).
>
> 3. Are the images (perfusion_calib.nii.gz) smoothed during prepossessing using oxford_asl. If not are there any recommendations for smoothing threshold. e.g. FWHM=6?
>
> If you use the --spatial option then there is automated (data-driven) spatial regularisation applied. However, if you are then proceeding to group analysis you might want to do some spatial smoothing as is conventional in BOLD analysis (and for the same reasons, with the same tradeoff in spatial localisation vs. SNR).
>
> Thanks
> John
>
> ---
> Michael Chappell MEng DPhil
> W: http://www.ibme.ox.ac.uk/qubic
> T: +44 1865 617657
> Associate Professor, Institute of Biomedical Engineering, University of Oxford.
> http://www.ibme.ox.ac.uk
> Director of Training, EPSRC-MRC CDT in Biomedical Imaging
> http://www.onbicdt.ox.ac.uk
> Governing Body Fellow, Wolfson College, Oxford.
> http://www.wolfson.ox.ac.uk
>
> PhysiologyforEngineers.org
> NeuroimagingPrimers.org
>
>
>
>
>
>
>
|
