VIP Split Brain ERROR

[schampus]

Hello!

We at the CRC (Liège) have the problem that for certain subjects we always encounter that problem during the Splin Brain Process. Until there everything seems to work fine... Could you please help us on that? What might be the problem here...



Process Vip Split Brain 13 started on 2005/09/16 21:53
in Vip Split Brain 13 RuntimeError: System command exited with non null value : 111

neuroProcesses.py (1476) in _processExecution:
result = process.execution( self )
VipSplitBrain.py (107) in execution:
apply( context.system, call_list+option_list )
neuroProcessesGUI.py (934) in system:
ret = apply( ExecutionContextGUI.system, (self,) + args, kwargs )
neuroProcesses.py (1556) in system:
raise RuntimeError( _t_( 'System command exited with non null value : %s' ) % str( ret ) )

[Jean-Francois Mangin]

Did you read other posts on Split Brain process?

You do not give enough information for us to understand what happen.

One advise: apply the bias correction of the pipeline 2005,
then lock the bias corrected image, and switch back to the usual
2004 pipeline. This new bias correction overcomes a lot
of problems.

[schampus]

Hi!

The 2005 pipeline bias unfortunately does not help. What information do you need in order to be able to help from here?

[schampus]

Perhaps this is an important point:

In the step before "split brain" I perform VIP get brain often getting the following error:

------------------------------
! VIP Warning:
Something was wrong with the iterative erosion!
------------------------------
unrecognized tag start: e

Therefore, I have to switch to 'Get Brain' mode: "standard" for many subjects... then continuing with VIP 'Split Brain' ...

[Jean-Francois Mangin]

You should open the log window providing the details of each processing

[Pierre]

Hello,

I have the same problem.

I first applied the T1 Bias Correction 2005 process, and then apply again the Ana Do A Lot of Things 2004.

Unfortunatly the error is still present.
There is a warning during VipHistoAnalysis, and the error occurs during VipSplitBrain.
Here is the log:

'VipHistoAnalysis' '-i' '/pierre/training/patient/mcd_049/mcd_049/anatomy/nobias_mcd_049' '-o' '/pierre/training/patient/mcd_049/mcd_049/anatomy/nobias_mcd_049' '-S' 'y' '-m' 'a'
Output
Reading volume...
read aimsdata: 181, vol: 181, border: 0
Computing histogram
Writing histogram
Computing scale space singularities up to order three...
(95 percent threshold: 1259)
Undersampling histogram... (factor 16 )
Detecting D1/D2 singularity matings and cascades...
Analysing histogram D1 and D2 singularities...
Main event scales: 1226 37 177
Main cascade volumes: 1172502 860556 360598

------------------------------
! VIP Warning:
Impossible to distinguish gray and white matter: insufficient grey white contrast, you have to increase field regularization in VipBiasCorrection)
------------------------------
Gray matter found at: [464-832-76] (sigma/mean/sigma)
White matter found at: [64-1040-64] (sigma/mean/sigma)
This T1-MR image seems to come from an Inversion/Recovery sequence (high gray/white contrast)
Result
Value returned = 0

Command line
'VipSplitBrain' '-input' '/pierre/training/patient/mcd_049/mcd_049/anatomy/nobias_mcd_049' '-brain' '/pierre/training/patient/mcd_049/mcd_049/segment/brain_mcd_049' '-analyse' 'r' '-hname' '/pierre/training/patient/mcd_049/mcd_049/anatomy/nobias_mcd_049' '-output' '/pierre/training/patient/mcd_049/mcd_049/segment/voronoi_mcd_049' '-erosion' '2.0' '-ccsize' '500' '-walgo' 'r' '-Points' '/pierre/training/patient/mcd_049/mcd_049/anatomy/mcd_049' '-template' '/pierre/Brainvisa/share/shfj/hemitemplate/closedvoronoi' '-TemplateUse' 'y'
Output
Reading /pierre/training/patient/mcd_049/mcd_049/anatomy/nobias_mcd_049...
read aimsdata: 181, vol: 183, border: 1
Reading /pierre/training/patient/mcd_049/mcd_049/segment/brain_mcd_049...
read aimsdata: 181, vol: 183, border: 1
Reading voronoi template /pierre/Brainvisa/share/shfj/hemitemplate/closedvoronoi...
read aimsdata: 157, vol: 157, border: 0
Reading AC, PC, IH coordinates: /pierre/training/patient/mcd_049/mcd_049/anatomy/mcd_049
Reading Coordinate file (AC, PC...)
Anterior commissure: 91 88 113
Posterior commissure: 91 115 109
Random interhemispheric point: 90 109 53
100%
Talairach bounding box (mm):
-92.443741 -96.429024 -116.403557
90.097755 143.031494 83.422363
Scaling: x: 0.860078, y:0.789274, z:0.680592
Reading histogram analysis: /pierre/training/patient/mcd_049/mcd_049/anatomy/nobias_mcd_049
Regularized classification
Keep only the largest 6 connected component...
Opening of size 1.6mm ...
1.0
dist: 0.0
Closing of size 1.6mm ...
dist: 0.0
dist: 0.0
AutoDeleting 2082 white matter points from interhemispheric plane and pons
Let us try to split this brain correctly...
erosion of size 2.000000 mm
dist: 0.0

------------------------------
! VIP Warning:
Empty volume in VipGet3DConnex
------------------------------
Not enough cc
-------------
erosion of size 2.500000 mm
dist: 0.0

------------------------------
! VIP Warning:
Empty volume in VipGet3DConnex
------------------------------
Not enough cc
-------------
erosion of size 3.000000 mm
dist: 0.0

------------------------------
! VIP Warning:
Empty volume in VipGet3DConnex
------------------------------
Not enough cc
-------------
erosion of size 3.500000 mm
dist: 0.0

------------------------------
! VIP Warning:
Empty volume in VipGet3DConnex
------------------------------
Not enough cc
-------------
erosion of size 4.000000 mm
dist: 0.0

------------------------------
! VIP Warning:
Empty volume in VipGet3DConnex
------------------------------
Not enough cc
-------------
erosion of size 4.500000 mm
dist: 0.0

------------------------------
! VIP Warning:
Empty volume in VipGet3DConnex
------------------------------
Not enough cc
-------------
erosion of size 5.000000 mm
dist: 0.0

------------------------------
! VIP Warning:
Empty volume in VipGet3DConnex
------------------------------
Not enough cc
-------------
erosion of size 5.500000 mm
dist: 0.0

------------------------------
! VIP Warning:
Empty volume in VipGet3DConnex
------------------------------
Not enough cc
-------------
erosion of size 6.000000 mm
dist: 0.0

------------------------------
! VIP Warning:
Empty volume in VipGet3DConnex
------------------------------
Not enough cc
-------------
erosion of size 6.500000 mm
dist: 0.0

------------------------------
! VIP Warning:
Empty volume in VipGet3DConnex
------------------------------
Not enough cc
-------------
erosion of size 7.000000 mm
dist: 0.0

------------------------------
! VIP Warning:
Empty volume in VipGet3DConnex
------------------------------
Not enough cc
-------------
erosion of size 7.500000 mm
dist: 0.0

------------------------------
! VIP Warning:
Empty volume in VipGet3DConnex
------------------------------
Not enough cc
-------------
erosion of size 8.000000 mm
dist: 0.0

------------------------------
! VIP Warning:
Empty volume in VipGet3DConnex
------------------------------
Not enough cc
-------------
erosion of size 8.500000 mm
dist: 0.0

------------------------------
! VIP Warning:
Empty volume in VipGet3DConnex
------------------------------
Not enough cc
-------------
erosion of size 9.000000 mm
dist: 0.0

------------------------------
! VIP Warning:
Empty volume in VipGet3DConnex
------------------------------
Not enough cc
-------------
erosion of size 9.500000 mm
dist: 0.0

------------------------------
! VIP Warning:
Empty volume in VipGet3DConnex
------------------------------
Not enough cc
-------------
Result
Value returned = 111

[Jean-Francois Mangin]

Did you lock the bias corrected image before triggering the 2004 pipeline?

If yes, can you post your data somewhere?

[Pierre]

Hello,

Thank you for your response.
I'm using the last version of BV (3.0.1) and there is no lock anymore. So, I supposed the nobias image used for "Ana Do A Lot..." was the latest generated (i.e. with "T1 Bias Correction").

If it is not the case, how to lock an image ?

[Pierre]

I'm sorry for my precedent post, i just saw that nobias image was an output of "Ana Do A Lot...", not an input.

Anaway, i'm still looking for the way to use the nobias image generated by T1 Bias Correction.

Thanks a lot,

Pierre

[Manik]

To use the 2005 bias correction, you should run the bias correction process of 2005, then lock it to prevent the 2004 pipeline from destroying it.
To lock it, just use the process in Anatomy/T1 Pipeline 2004/validation/Validation_1 Bias Correction from T1 MRI with the validation parameter set to "Lock".
You can then run the Ana Do All process of 2004 without destroying it.

[Pierre]

Hi,

Thank you for your help ! I tried with 2 unsolved cases, one is working (great), but the other one is not. There is a RuntimeError during the Ana Split Brain process.

Here is the log:

Ana Do A Lot of Things from T1 MRI
Process identifier
AnaDoAll
Parameters
Processing_type = Do All
Side = Both
T1mri = /pierre/training/patient/mcd_050/mcd_050/anatomy/mcd_050.ima
Contrast = High grey/white contrast
Bias_type = Standard bias field
Compress_results = False
lesion_mask = None
mri_corrected = /pierre/training/patient/mcd_050/mcd_050/anatomy/nobias_mcd_050.ima
histo_analysis = /pierre/training/patient/mcd_050/mcd_050/anatomy/nobias_mcd_050.han
brain_mask = /pierre/training/patient/mcd_050/mcd_050/segment/brain_mcd_050.ima
Use_template = True
voronoi_template = /pierre/Brainvisa/share/shfj/hemitemplate/closedvoronoi.ima
brain_voronoi = /pierre/training/patient/mcd_050/mcd_050/segment/voronoi_mcd_050.ima
Commissure_coordinates = /pierre/training/patient/mcd_050/mcd_050/anatomy/mcd_050.APC
left_grey_white = /pierre/training/patient/mcd_050/mcd_050/segment/Lgrey_white_mcd_050.ima
right_grey_white = /pierre/training/patient/mcd_050/mcd_050/segment/Rgrey_white_mcd_050.ima
left_hemi_cortex = /pierre/training/patient/mcd_050/mcd_050/segment/Lcortex_mcd_050.ima
right_hemi_cortex = /pierre/training/patient/mcd_050/mcd_050/segment/Rcortex_mcd_050.ima
left_hemi_mesh = /pierre/training/patient/mcd_050/mcd_050/tri/mcd_050_Lhemi.mesh
right_hemi_mesh = /pierre/training/patient/mcd_050/mcd_050/tri/mcd_050_Rhemi.mesh
left_white_mesh = /pierre/training/patient/mcd_050/mcd_050/tri/mcd_050_Lwhite.mesh
right_white_mesh = /pierre/training/patient/mcd_050/mcd_050/tri/mcd_050_Rwhite.mesh
head_mesh = /pierre/training/patient/mcd_050/mcd_050/tri/mcd_050_head.mesh
iterations = 10
rate = 0.2
Lskeleton = /pierre/training/patient/mcd_050/mcd_050/segment/Lskeleton_mcd_050.ima
Rskeleton = /pierre/training/patient/mcd_050/mcd_050/segment/Rskeleton_mcd_050.ima
Lroots = /pierre/training/patient/mcd_050/mcd_050/segment/Lroots_mcd_050.ima
Rroots = /pierre/training/patient/mcd_050/mcd_050/segment/Rroots_mcd_050.ima
Lgraph = /pierre/training/patient/mcd_050/mcd_050/graphe/Lmcd_050.arg
Rgraph = /pierre/training/patient/mcd_050/mcd_050/graphe/Rmcd_050.arg
compute_fold_meshes = Yes
Output
Process Ana Do A Lot of Things from T1 MRI 1 started on 2006/01/16 12:46


in Ana Do A Lot of Things from T1 MRI 1 RuntimeError: System command exited with non null value : 111

neuroProcesses.py (1612) in _processExecution:
result = process.execution( self )
AnaDoAll.py (148) in execution:
?
neuroProcesses.py (1505) in runProcess:
result = self._processExecution( _process )
neuroProcesses.py (1612) in _processExecution:
result = process.execution( self )
AnaSplitBrainFromBrainMask.py (69) in execution:
?
neuroProcesses.py (1505) in runProcess:
result = self._processExecution( _process )
neuroProcesses.py (1612) in _processExecution:
result = process.execution( self )
VipSplitBrain.py (107) in execution:
?
neuroProcessesGUI.py (932) in system:
ret = apply( ExecutionContextGUI.system, (self,) + args, kwargs )
neuroProcesses.py (1692) in system:
raise RuntimeError( _t_( 'System command exited with non null value : %s' ) % str( ret ) )

Process Ana Do A Lot of Things from T1 MRI 1 finished on 2006/01/16 12:47 (1 minutes 1 seconds)

I think it is not a problem with the lock, because in the process window i have a confirmation:

/pierre/training/patient/mcd_050/mcd_050/anatomy/nobias_mcd_050 has been locked
Remove /pierre/training/patient/mcd_050/mcd_050/anatomy/nobias_mcd_050 .loc if you want to trigger a new correction

[Jean-Francois Mangin]

Go to the pipeline05 components directory. There open the Split Brain Process and
try variants of the process. You may be more lucky.
You can also go the the correction directory of 2004 pipeline, where you will get similar alternatives with some more help in the decumentation.

Anyway, there is some trouble with your data, and it may be interesting for us
to see one. Can you post it somewhere?

[Pierre]

Hi,

Thank you for your answer. I would like to help you, but what do you mean exactly by data ? What do you want me to post ?

A+

[Jean-Francois Mangin]

I was just assking for you to put on some ftp site or send by Email (after compression) one of the MR scan leading to trouble. If we find a fix,
we can improve the next release of brainVISA...

[Jean-Francois Mangin]

Thanks for your dataset Pierre, I tried to discover why you had some problem.
After a while I observed that:

No comment