LIBRA mis-match CXXABI on Ubuntu 19.04 and above

[sarthakpati]

Describe the bug When running LIBRA Single Image mode, users see a CXX_ABI related error when trying to write DICOM masks.

To Reproduce Steps to reproduce the behavior:

./captk BreastTexturePipeline -i ~/Desktop/3251790_01.dcm -o ~/Desktop/
Running LIBRA Single Image with command '/tmp/.mount_captkYrOtwe/usr/bin/libra /home/ABC/Desktop/3251790_01.dcm /home/ABC/Desktop/temp/3251790_01'
Running LIBRA-1.0.4 (Revision: 603) on MATLAB R2014b
  >> Analzying single file...
  >> Processing File 1/1: /home/ABC/Desktop/3251790_01.dcm
   > Checking header
     > WARNING: Age information is missing. Use Default.
     > WARNING: PixelSpacing information is missing. Defaulting to detector resolution
   > Checking passed.
...identifying appropriate training model
...using model:GE-Processed...
Masking out Breast
Running 15-class FCM
...Getting dense-cluster metrics (i.e. hierarchical agglomeration)
Trying to write the result
MATLAB:dispatcher:loadLibrary Can't load '/tmp/.mount_captkYrOtwe/usr/libexec/MCR/v84/bin/glnxa64/libmwosgserver.so': /tmp/.mount_captkYrOtwe/usr/bin/../libexec/MCR/v84/sys/os/glnxa64/libstdc++.so.6: version `CXXABI_1.3.8' not found (required by /lib/x86_64-linux-gnu/libGLU.so.1).

# information about the stack comes here, which is not relevant for this repoprt

This is reproducible on GUI as well.

Expected behavior The library should work correctly.

CaPTk Version 1.8.0.Beta

Desktop (please complete the following information):

• OS: Linux
• Version: Ubuntu 20.04

Additional context

• Reported by @omaghsoudi
• Possible solution: https://askubuntu.com/a/1176351

[sarthakpati]

Possible inter-compatibility with #1226

[ashishsingh18]

Do the masks get written eventually or not?

[sarthakpati]

Do the masks get written eventually or not?

Nope, they do not.

[MarkBergman-cbica]

What about creating a new LIBRA exe from a more modern Matlab? Check w. Walt & Aimilia about which Matlab version broke (changed the results) of LIBRA due to updated functions, though Michael's probably the only know who knows the history.

[MarkBergman-cbica]

Possible inter-compatibility with #1226

I do not believe that 1226 and 1229 are related.