Error when using recon-neonatal-cortex

[Thea-Eddie-Amy]

Hi, I was trying to use the MIRTK recon-neonatal-cortex, I run the command like this: mirtk recon-neonatal-cortex -v --threads=4 --config=C:\Users\lenovoM425\Desktop\dHCP_pipeline_test\dhcp-structural-pipeline-master\parameters\recon-neonatal-cortex.cfg --session=test-0 --prefix=surfaces\test-0\vtk\test-0 --temp=surfaces\test-0\vtk\temp-recon\test-0 --white --pial

however gets the error `Reconstructing cortical surfaces of test session 0 Dividing brain volume into disjoint regions

Traceback (most recent call last): File "C:\Users\lenovoM425\AppData\Local\Programs\Python\Python38\lib\configparser.py", line 789, in get value = d[option] File "C:\Users\lenovoM425\AppData\Local\Programs\Python\Python38\lib\collections_init_.py", line 898, in getitem return self.missing(key) # support subclasses that define missing File "C:\Users\lenovoM425\AppData\Local\Programs\Python\Python38\lib\collections_init_.py", line 890, in missing raise KeyError(key) KeyError: 'fill_wm_holes'

During handling of the above exception, another exception occurred:

Traceback (most recent call last): File "D:\CMake Program Files\MIRTK\Lib\Tools\recon-neonatal-cortex.cmd", line 923, in recon_neonatal_cortex(config=config, section=args.section, config_vars=config_vars, File "D:\CMake Program Files\MIRTK\Lib\Tools\recon-neonatal-cortex.cmd", line 480, in recon_neonatal_cortex require_brain_mask(config, section, config_vars, stack, verbose, File "D:\CMake Program Files\MIRTK\Lib\Tools\recon-neonatal-cortex.cmd", line 228, in require_brain_mask require_regions_mask(config, section, config_vars, stack, verbose) File "D:\CMake Program Files\MIRTK\Lib\Tools\recon-neonatal-cortex.cmd", line 207, in require_regions_mask fill_wm_holes=config.getboolean(section, 'fill_wm_holes'), File "C:\Users\lenovoM425\AppData\Local\Programs\Python\Python38\lib\configparser.py", line 828, in getboolean return self._get_conv(section, option, self._convert_to_boolean, File "C:\Users\lenovoM425\AppData\Local\Programs\Python\Python38\lib\configparser.py", line 808, in _get_conv return self._get(section, conv, option, raw=raw, vars=vars, File "C:\Users\lenovoM425\AppData\Local\Programs\Python\Python38\lib\configparser.py", line 803, in _get return conv(self.get(section, option, **kwargs)) File "C:\Users\lenovoM425\AppData\Local\Programs\Python\Python38\lib\configparser.py", line 792, in get raise NoOptionError(option, section) configparser.NoOptionError: No option 'fill_wm_holes' in section: 'recon-neonatal-cortex'`

Please help, thanks.

[schuhschuh]

Hi, you will see this error with an "older" version of MIRTK. I've added the fill_wm_holes option only in recent months. To use it, you have to use the latest HEAD revision of the master branch. Please note that due to a bug in VTK <9, you have to build MIRTK with VTK 9. This unfortunately affects the AppImage that is available on bintray. I should have submitted a fix for this by the end of this week.

Otherwise, if you want to keep using the MIRTK version you have, you will have to remove the new option from your input configuration file.

[Thea-Eddie-Amy]

Hi @schuhschuh, Thank you for your reply! Finally, I successfully processed my data with dhcp-structural-pipeline. But I am not sure whether the pipeline uses the latest version of recon-neonatal-cortex.

Here are a few questions I'm curious about:

1. Will refining the cGM label in tissue_labels.nii.gz manually be helpful to increase the accuracy of cortical surface reconstruction?
2. Is there any thing I can do to refine my cortical surface after running the whole pipeline?

Thanks again for your help! Best regards, Thea