ConnectomeViewer and Surface Labels - Issue

[Tim Schultz]

Greetings everyone,

Quick question about using the ConnectomeViewer after successfully completing all connectomemapper tasks.

I'm interested in isolating a single ROI in CSurface/surface_rh.white.gii (say, the Corpus Callosum) for visual purposes.  It would be great if I could keep all other regions, but reduce their alpha (i.e. alpha = 0.25).  I found the Label Table values within the XML in CSurface/surface_label_rh.aparc.annot.gii.  Example:

      <Label Key="3294840"  Red="1" Green="0" Blue="0" Alpha="1"><![CDATA[corpuscallosum]]></Label>

      <Label Key="14433500" Red="0" Green="1" Blue="0" Alpha="0.25"><![CDATA[inferiorparietal]]></Label>

      <Label Key="7874740"  Red="0" Green="1" Blue="0" Alpha="0.25"><![CDATA[inferiortemporal]]></Label>

However, editing these values don't seem to have an effect when I apply the label surface to the RH white surface.  If I load the RH White surface, right click it in the MayaVi Visualization Tree, and select 'Add Modules' -> 'Labels', the application crashes with the following message:

ERROR: In /home/tim/projects/VTK5.10.1/Common/vtkDataArrayTemplate.txx, line 144

vtkFloatArray (0x9db9940): Unable to allocate 9223372036854775805 elements of size 4 bytes. 

Aborted (core dumped)

My Linux box has 32GB or RAM, but 9223372036854775805 is a lot of elements.  Where exactly is this number coming from?  Note: I also get this same error with the example CFF file.

If I manually dig into the guts of the data object contained within CSurface/surface_label_rh.aparc.annot.gii, here's what I see:

In [78]: x

Out[78]: <nibabel.gifti.gifti.GiftiImage at 0x43c9b90>

In [79]: x.print_summary()

----start----

Source filename:  /home/tim/data/dti/myproject/DTI2/ADNI/098_S_2079/2010-09-22_14_37_47.0/CMP/cff/CSurface/surface_label_rh.aparc.annot.gii

Number of data arrays:  1

Version:  1.0

----

Metadata:

{u'UserName': u'tim', u'Date': u'Thu Mar 28 07:14:42 2013', u'gifticlib-version': u'gifti library version 1.09, 28 June, 2010'}

None

----

Labeltable:

{9180300: u'isthmuscingulate', 9182740: u'lateraloccipital', 13145750: u'transversetemporal', 1639705: u'unknown', 2647065: u'bankssts', 1316060: u'postcentral', 14433500: u'inferiorparietal', 3302560: u'middletemporal', 3296035: u'lateralorbitofrontal', 1326300: u'parstriangularis', 7874740: u'inferiortemporal', 9231540: u'fusiform', 3988540: u'paracentral', 3988500: u'parahippocampal', 11832480: u'precuneus', 11146310: u'temporalpole', 4924360: u'medialorbitofrontal', 14474380: u'superiortemporal', 6558940: u'cuneus', 8204875: u'rostralmiddlefrontal', 9180240: u'rostralanteriorcingulate', 3294840: u'corpuscallosum', 6553700: u'frontalpole', 660700: u'entorhinal', 9211105: u'lingual', 6500: u'caudalmiddlefrontal', 1351760: u'supramarginal', 14423100: u'precentral', 10542100: u'superiorfrontal', 14464220: u'posteriorcingulate', 9221340: u'parsopercularis', 3957880: u'pericalcarine', 3302420: u'parsorbitalis', 10511485: u'caudalanteriorcingulate', 9221140: u'superiorparietal', 2146559: u'insula'}

None

----

DataArray 0:

Intent:  NIFTI_INTENT_LABEL

DataType:  NIFTI_TYPE_INT32

ArrayIndexingOrder:  RowMajorOrder

Dimensionality:  1

Dimensions:  [143611]

Encoding:  GZipBase64Binary

Endian:  GIFTI_ENDIAN_LITTLE

ExternalFileName:  

ExternalFileOffset:  

----

Coordinate System:

Dataspace:  NIFTI_XFORM_UNKNOWN

XFormSpace:  NIFTI_XFORM_UNKNOWN

Affine Transformation Matrix: 

[[ 1.  0.  0.  0.]

 [ 0.  1.  0.  0.]

 [ 0.  0.  1.  0.]

 [ 0.  0.  0.  1.]]

None

None

----end----

The data appears to be a 1 x 143611 array.

Is there any quick and dirty way of fixing this issue or working around it?  Manually editing the CSurface/surface_label_rh.aparc.annot.gii Label Table with desired color and alpha values doesn't appear to have any effect.  I would just like to isolate a few regions, while making all other regions (mostly) transparent.

Any suggestions would be greatly appreciated.

Thanks!

Regards,

Tim