[Camino-users] Connectivity Mapping with Camino

[Ian Malone]

Hi,

conmap did work this way when it was first written, but that was with a 
much earlier SVN revision, I think there must have been a refactoring of 
the ROI code since then (quite a lot of functionality has been added in 
the meantime).

You said you were using a Linux system, so patching Camino should be as 
simple as:
1. Get the latest SVN release.
2. Save the patch in the SVN camino/ directory
3. cd into the camino SVN directory and run:
patch -p0 < connectivity-mapping.patch
4. do make/make install as normal

(Assuming you've already got a checked out/built camino you'd do step 3, 
'make clean' and then step 4)

Ian

Irvin Teh wrote:
> Dear Nagesh and Ian,
>
> Many thanks for the replies.
>
> Nagesh, I tried using two ROIs generated using Mricron's intensity filter (Range:1-178 and 78-255 or 178 bins each) to match my ROI regions reported by FreeFormROI. This wasn't ideal as the ROI using those settings didn't cover the whole brain, and in any case, resulted in the same error as before.
>
> Ian, thank you for sharing your patch. I've not installed patches before on Camino, and would appreciate it if you could point me in the right direction regarding installation before I get too clever and break something. 
>
> Is this a R1010/1011 issue - would it help to roll back to an earlier version?
>
> Thanks all,
>
> Irvin
>
>
> On Wed, Oct 12, 2011 at 5:13 AM, Ian Malone <i.malone at ucl.ac.uk> wrote:
>
>     Hi,
>
>     This is the issue I was talking about in my reply to Gavin Winston (Re:
>     Connectivity Mapping with Camino). You're right: the FreeFormROI class
>     assumes that ROI are sequential (they can actually start at a whole
>     number other than 1, so can range [m, m+N-1] for N ROI). The problem is
>     that if trying to use FreeSurfer as described for region labels you do
>     not have sequential ROI indices, for example 1xxx is used for one side
>     of the cortex and 2xxx for the other hemisphere. The patch in my other
>     email fixes (I think) the mapping that FreeFormROI uses between indices
>     and its internal ROI list (so far as conmap is concerned) to remove this
>     requirement. However I should warn I haven't tested extensively against
>     the other utilities that use ROIs (e.g. seeds and waypoints for
>     tractography).
>
>     Ian
>
>     On 11/10/11 20:32, Nagesh Adluru wrote:
>     > Hello Irvin,
>     >
>     > Your ROI /intensity /values should not be more than 178 if you have
>     > 178 ROIs. That said I think there are some things in conmap that I
>     > don't fully understand.
>     >
>     > On Tue, Oct 11, 2011 at 5:20 AM, Irvin Teh <medv820 at nus.edu.sg
>     > <mailto:medv820 at nus.edu.sg>> wrote:
>     >
>     >     Forgot to mention, we're on Camino R1010 running on 64-bit Linux
>     >     with Sun Java ver 1.6.0_27.
>     >
>     >     Thanks.
>     >
>     >     ________________________________________
>     >     From: Irvin Teh
>     >     Sent: Tuesday, October 11, 2011 5:28 PM
>     >     To: camino-users at www.nitrc.org <mailto:camino-users at www.nitrc.org>
>     >     Subject: Trouble getting conmap to work
>     >
>     >     Hello all,
>     >
>     >     I'm trying to setup a pipeline for connectivity mapping using
>     >     Camino & Freesurfer and followed the instructions here.
>     >     http://web4.cs.ucl.ac.uk/research/medic/camino/pmwiki/pmwiki.php?n=Tutorials.ConnectivityMapping
>     >
>     >     There seems to be no problem (eg. the wm_undersampled.nii.gz looks
>     >     correctly segmented and downsampled to the resolution of the
>     >     diffusion images) till I get to the final conmap step whereby I
>     >     get the error message below.
>     >     I've tried with to troubleshoot by using slightly different ROIs
>     >     of the whole brain (created using different intensity filters in
>     >     Mricron), and also tried using running conmap with a track file
>     >     created using PICo, but repeatedly faced the same error.
>     >
>     >
>     >     [irvin at server dti68]$ conmap -inputfile brain_track.Bdouble
>     >     -roifile wm_undersampled.nii.gz
>     >     Exception in thread "main" java.lang.IllegalArgumentException: ROI
>     >     index out of bounds. This ROI contains 178 regions
>     >            at tractography.FreeFormROI.getRegion(FreeFormROI.java:345)
>     >            at apps.ConnectivityMapping.main(ConnectivityMapping.java:238)
>     >     [irvin at server dti68]$
>     >
>     >
>     >     Could it be due to imperfect anatomical segmentation, or incorrect
>     >     ROI selection?
>     >     Would be very grateful for any suggestions and fixes, thanks!
>     >
>     >     Irvin
>     >     Singapore
>     >     _______________________________________________
>     >     Camino-users mailing list
>     >     Camino-users at www.nitrc.org <mailto:Camino-users at www.nitrc.org>
>     >     http://www.nitrc.org/mailman/listinfo/camino-users
>     >
>     >
>     >
>     >
>     > _______________________________________________
>     > Camino-users mailing list
>     > Camino-users at www.nitrc.org
>     > http://www.nitrc.org/mailman/listinfo/camino-users
>
>
>     _______________________________________________
>     Camino-users mailing list
>     Camino-users at www.nitrc.org
>     http://www.nitrc.org/mailman/listinfo/camino-users _______________________________________________
> Camino-users mailing list
> Camino-users at www.nitrc.org
> http://www.nitrc.org/mailman/listinfo/camino-users
>
>