Re: Inquiry about ExploreASL

[Mutsaerts, H.J.M.M. (Henk Jan)]

Dear Anna,

Currently, the NIfTI images are stored by default without PVC, which can easily be appreciated visually by the effective spatial resolution that the final maps have.

After PVC, the separate 100%GM and 100%WM CBF images would be much smoother by design, unless they were multiplied by the PV maps, which could make them artificially appear much sharper. We have the option to export with PVC in both native and standard space but it needs to be applied “under-the-hood”, Jan might be able to help you with this.

The ROI values that are stored in the Stats folder do include both corrected and uncorrected values, though, as explained in the tutorial on the ExploreASL website.

May I ask you what kind of clinical analyses you are doing? Perhaps it would also be interesting to have a look at the spatial CoV/vascular component.

Also, did you know that you have a colleague working with ASL at the TUM (we - mainly Jan - also collaborate with him)?

BW, Henk

Henk(-Jan) MM Mutsaerts, MD PhD

Amsterdam UMC/UZ Ghent/UCL
Phone: +31 6 4390 8284

On 13 Nov 2020, at 14:55, Rubinski, Anna <Anna.R...@med.uni-muenchen.de> wrote:

Dear Dr. Mutsaerts, 

First of all, thank you for creating such a wonderful tool for preprocessing ASL data!

I was running the automatic pipeline on pCASL data that I have, and I was wondering whether the CBF final image both in the subject as well as the population folder are partial-volume corrected?

Is there a possibility to get also the uncorrected images? as you are probably well aware, many reviewers are asking for repetition of the analyses both using the PV corrected and un-corrected data. Therefore, having both versions will benefit the analysis. I am trying to look at which of the steps/scripts to modify, but I would really appreciate if you will point me in the right direction. 

​

Thank you for your help,

Anna Rubinski, M.Sc

Institute for Stroke and Dementia Research (ISD)
Klinikum der Universität München
Feodor-Lynen-Straße 17, D-81377 Munich, Germany

Phone: +49 (0)89 4400 - 46161

E-Mail: anna.r...@med.uni-muenchen.de
Internet: www.isd-muc.de

______________________________________________________
VUmc disclaimer : www.vumc.nl/disclaimer
AMC disclaimer : www.amc.nl/disclaimer

[Mutsaerts, H.J.M.M. (Henk Jan)]

From: Jan Petr <j.p...@hzdr.de>
Date: Tuesday, 17 November 2020 at 14:58
To: Rubinski, Anna <Anna.R...@med.uni-muenchen.de>, Mutsaerts, H.J.M.M. (Henk Jan) <h.j.mu...@amsterdamumc.nl>
Subject: Re: Inquiry about ExploreASL

Hi Anna and Henk,

With the bNativeSpaceAnalysis, you should be able to get the *_PVC0.tsv and *_PVC2.tsv files in the stats folder with the native space results with and without PVC. However, these are all computed for each region separately and within the whole region.

If you are interested in having the CBF maps with PV corrected GM and WM CBF, we currently do not export these.
However, you can find the files CBF.nii, PVgm.nii, and PVwm.nii in each subject's folder. These are all co-registered and sampled in the same resolution. So they are ready for partial volume correction.
You can either use the xASL_im_PVCkernel function yourself to calculate the maps, or wait until I add this option to ExploreASL.

It's on our list for a long time, I can try to make it mid next week.

best regards,
Jan

PS: Henk, note that Anna seems to work at LMU, which is a different university than TUM.

Dear Henk and Jan,

 

Thank you for the detailed reply, I understand now that the images are not PVC corrected but the values in the stats are provided in both versions. However, I only get stats file in standard space and not in native space, should I define something specifically in the parameters file for native space? Or the output should be automatic both in standard and native space?

 

Jan – I would highly appreciate if you could let me know whether it is possible to export the images with PVC in native space and how.

 

I am mainly working with participants on the Alzheimer’s disease spectrum, therefore it will be interesting at some point to look at the vascular components as well.

Regarding the colleague from TUM, no I did not know 😊