From eskild at gmail.com  Mon Feb  2 05:17:25 2015
From: eskild at gmail.com (Simon Eskildsen)
Date: Mon, 2 Feb 2015 11:17:25 +0100
Subject: [MINC-users] Isolated Spots in mincbeast masks?
In-Reply-To: <CAFQOoOo-1Mnzjn0-+_KjWogA869ZixeGY6YfgoiC+zTY1LN2Lg@mail.gmail.com>
References: <CAFQOoOo-1Mnzjn0-+_KjWogA869ZixeGY6YfgoiC+zTY1LN2Lg@mail.gmail.com>
Message-ID: <CAD3v434t=3QRCtgOawhBb7KocDEdt-=NkqJ2q3mE7Sx3LTwidg@mail.gmail.com>

Hi Gabriel,

The -fill option simply performs a post-processing step where possible
holes in the mask are filled. Holes are in this context defined as
background objects completely surrounded by foreground object (the mask).

mincbeast always removes isolated mask components as a post-processing step
(irrespective of the -fill option). If the isolated mask components, which
can be seen in the sagittal view of your figure, are truly isolated in 3D,
then mincbeast has a bug. However, I suspect that these voxels are
connected to the main component in 3D.

Let me know if this is not the case and I can take a closer look at your
example.

Simon

On Wed, Jan 28, 2015 at 11:33 PM, Gabriel A. Devenyi <gdevenyi at gmail.com>
wrote:

> Hi all,
>
> I?m having some trouble with mincbeast producing masks with isolated
> spots, despite using the -fill option when invoking. See attached
> screenshot of a brain-mask overlay.
>
> I would?ve expected that the mask would be expanded to encompass the
> isolated mask component, is that not the case?
>
> I invoked mincbeast with the following
>
> mincbeast -clobber -verbose -fill -median -same_res -flip -conf /opt/quarantine/resources/BEaST_libraries/combined/default.1mm.conf /opt/quarantine/resources/BEaST_libraries/combined preprocessed/OAS1_0019_MR1_mpr_n4_anon_sbj_111.n4correct.affine_register_masked.cutneckapply.beastnormalize.mnc preprocessed/OAS1_0019_MR1_mpr_n4_anon_sbj_111.n4correct.affine_register_masked.cutneckapply.beastnormalize.beastmask.mnc
>
> ?
> ?Am I missing something or is this a bug??
>
>
> --
> Gabriel A. Devenyi B.Eng. Ph.D.
> Research Computing Associate
> Computational Brain Anatomy Laboratory
> Cerebral Imaging Center
> Douglas Mental Health University Institute
> McGill University
> t: 514.761.6131x4781
> e: gdevenyi at gmail.com
>

From gdevenyi at gmail.com  Mon Feb  2 11:05:06 2015
From: gdevenyi at gmail.com (Gabriel A. Devenyi)
Date: Mon, 2 Feb 2015 11:05:06 -0500
Subject: [MINC-users] Isolated Spots in mincbeast masks?
In-Reply-To: <CAD3v434t=3QRCtgOawhBb7KocDEdt-=NkqJ2q3mE7Sx3LTwidg@mail.gmail.com>
References: <CAFQOoOo-1Mnzjn0-+_KjWogA869ZixeGY6YfgoiC+zTY1LN2Lg@mail.gmail.com>
	<CAD3v434t=3QRCtgOawhBb7KocDEdt-=NkqJ2q3mE7Sx3LTwidg@mail.gmail.com>
Message-ID: <CAFQOoOpCqJw8T8HvQsVt7DhkqOerMeL6wN9+KFuqyN5AwQuKQA@mail.gmail.com>

Simon,

You are correct, the region was connected elsewhere, it was the slice I was
looking at that showed no connection. Thanks for your help.

Thanks for your feedback!

-- 
Gabriel A. Devenyi B.Eng. Ph.D.
Research Computing Associate
Computational Brain Anatomy Laboratory
Cerebral Imaging Center
Douglas Mental Health University Institute
McGill University
t: 514.761.6131x4781
e: gdevenyi at gmail.com

On Mon, Feb 2, 2015 at 5:17 AM, Simon Eskildsen <eskild at gmail.com> wrote:

> Hi Gabriel,
>
> The -fill option simply performs a post-processing step where possible
> holes in the mask are filled. Holes are in this context defined as
> background objects completely surrounded by foreground object (the mask).
>
> mincbeast always removes isolated mask components as a post-processing
> step (irrespective of the -fill option). If the isolated mask components,
> which can be seen in the sagittal view of your figure, are truly isolated
> in 3D, then mincbeast has a bug. However, I suspect that these voxels are
> connected to the main component in 3D.
>
> Let me know if this is not the case and I can take a closer look at your
> example.
>
> Simon
>
> On Wed, Jan 28, 2015 at 11:33 PM, Gabriel A. Devenyi <gdevenyi at gmail.com>
> wrote:
>
>> Hi all,
>>
>> I?m having some trouble with mincbeast producing masks with isolated
>> spots, despite using the -fill option when invoking. See attached
>> screenshot of a brain-mask overlay.
>>
>> I would?ve expected that the mask would be expanded to encompass the
>> isolated mask component, is that not the case?
>>
>> I invoked mincbeast with the following
>>
>> mincbeast -clobber -verbose -fill -median -same_res -flip -conf /opt/quarantine/resources/BEaST_libraries/combined/default.1mm.conf /opt/quarantine/resources/BEaST_libraries/combined preprocessed/OAS1_0019_MR1_mpr_n4_anon_sbj_111.n4correct.affine_register_masked.cutneckapply.beastnormalize.mnc preprocessed/OAS1_0019_MR1_mpr_n4_anon_sbj_111.n4correct.affine_register_masked.cutneckapply.beastnormalize.beastmask.mnc
>>
>> ?
>> ?Am I missing something or is this a bug??
>>
>>
>> --
>> Gabriel A. Devenyi B.Eng. Ph.D.
>> Research Computing Associate
>> Computational Brain Anatomy Laboratory
>> Cerebral Imaging Center
>> Douglas Mental Health University Institute
>> McGill University
>> t: 514.761.6131x4781
>> e: gdevenyi at gmail.com
>>
>
>

From gdevenyi at gmail.com  Thu Feb  5 08:57:04 2015
From: gdevenyi at gmail.com (Gabriel A. Devenyi)
Date: Thu, 5 Feb 2015 08:57:04 -0500
Subject: [MINC-users] CIVET-1.1.12 and -model icbm152nl_09s
Message-ID: <CAFQOoOqGA94Bf_Yvd1gHP_QcLSa1OgJ9OqDzcpmKn2Y0_fBv3w@mail.gmail.com>

Hi all,

I?m trying to run CIVET-1.1.12 with the more up-to-date 09a symmetric model
included with it, however the verify_image stage fails with:

Overlay /opt/quarantine/CIVET/1.1.12/build/share/mni-models//mni_icbm152_t1_tal_nlin_sym_09a_mask.obj
is not a file

I see that the mni_ADNI and symmetric_VI models have obj files but the
other model files are missing. Are these files available? Can anyone share
them?

Thanks!
?
-- 
Gabriel A. Devenyi B.Eng. Ph.D.
Research Computing Associate
Computational Brain Anatomy Laboratory
Cerebral Imaging Center
Douglas Mental Health University Institute
McGill University
t: 514.761.6131x4781
e: gdevenyi at gmail.com

From claude at bic.mni.mcgill.ca  Thu Feb  5 09:12:46 2015
From: claude at bic.mni.mcgill.ca (Claude LEPAGE)
Date: Thu, 5 Feb 2015 09:12:46 -0500
Subject: [MINC-users] CIVET-1.1.12 and -model icbm152nl_09s
In-Reply-To: <CAFQOoOqGA94Bf_Yvd1gHP_QcLSa1OgJ9OqDzcpmKn2Y0_fBv3w@mail.gmail.com>
Message-ID: <201502051412.t15ECkFe003991@agrippa.bic.mni.mcgill.ca>

Hi Gabriel,


> I'm trying to run CIVET-1.1.12 with the more up-to-date 09a symmetric model
> included with it, however the verify_image stage fails with:
> 
> Overlay /opt/quarantine/CIVET/1.1.12/build/share/mni-models//mni_icbm152_t1_tal_nlin_sym_09a_mask.obj
> is not a file
> 
> I see that the mni_ADNI and symmetric_VI models have obj files but the
> other model files are missing. Are these files available? Can anyone share
> them?

You should post such issues at civet-users at bic.mni.mcgill.ca.

Where are you running this quarantine from? It is not a quarantine that
I installed. It seems to have been copied from somewhere else or installed
by someone else. Up-to-date and complete versions are available in CBRAIN.

The file that you are missing is used to create the quality control image
only. It is not used for calculations, so missing it will not alter the 
results. This .obj is used to trace the outline of the model's mask in 
the top 2 rows of the verify.png image (red lines). You can replace the
missing sym_09a_mask.obj by the symmetric_VI one. Both are in MNI ICBM152
space. 

Claude

From gdevenyi at gmail.com  Fri Feb 13 11:31:53 2015
From: gdevenyi at gmail.com (Gabriel A. Devenyi)
Date: Fri, 13 Feb 2015 11:31:53 -0500
Subject: [MINC-users] mincresample while avoiding re-slicing
Message-ID: <CAFQOoOonCKE1WaMKucLX5EuxyGgKnAU5gR9YoLpFOSg2AjqyOQ@mail.gmail.com>

Hi all,

A question about mincresample, sometimes I'd like to use it to apply a
transform (say to move a brain into MNI space) according to some model minc
file (using -like) but I don't want to re-slice (change the step size) the
volume according to that -like input. Is there a way to prevent re-slicing
while still changing the dimensions and transforming the data?

Thanks!

-- 
Gabriel A. Devenyi B.Eng. Ph.D.
Research Computing Associate
Computational Brain Anatomy Laboratory
Cerebral Imaging Center
Douglas Mental Health University Institute
McGill University
t: 514.761.6131x4781
e: gdevenyi at gmail.com

From vladimir.fonov at gmail.com  Fri Feb 13 11:45:20 2015
From: vladimir.fonov at gmail.com (Vladimir S. FONOV)
Date: Fri, 13 Feb 2015 11:45:20 -0500
Subject: [MINC-users] mincresample while avoiding re-slicing
In-Reply-To: <CAFQOoOonCKE1WaMKucLX5EuxyGgKnAU5gR9YoLpFOSg2AjqyOQ@mail.gmail.com>
References: <CAFQOoOonCKE1WaMKucLX5EuxyGgKnAU5gR9YoLpFOSg2AjqyOQ@mail.gmail.com>
Message-ID: <54DE2A20.3090403@gmail.com>

Hello,


you can change step size, direction cosines and origin by using 
-tfm_input_sampling , without -like.

You can also change the dimensions using mincreshape.

On 15-02-13 11:31 AM, Gabriel A. Devenyi wrote:
> Hi all,
>
> A question about mincresample, sometimes I'd like to use it to apply a
> transform (say to move a brain into MNI space) according to some model minc
> file (using -like) but I don't want to re-slice (change the step size) the
> volume according to that -like input. Is there a way to prevent re-slicing
> while still changing the dimensions and transforming the data?


-- 
Best regards,

  Vladimir S. FONOV ~ vladimir.fonov <at> gmail.com