Moved chainer imports inside functions [no ci]

Attempt to copy model inside /code container

Enabled to disable multiprocessing for inference + disabled mp for app

Disable GPU for app + moved CPU disable logic to inside run_analysis method

Download models in Dockerfile + revert request changes

Test requests Session trust_env=False

Added support for accessing livermask as python package [no ci]

fixed argparse to support shortcuts

added fix for handling one or none unique liver segments

Added log

Change extension condition

Updated README

Added .nii.gz support

improved download logic to catch incomplete downloads

added github release download for parenchyma model

refactored to PEP8

tried to remove md5 for parenchyma model

filtered predicted vessels outside the predicted liver parenchyma region

changed tqdm for vessel seg for consistency

mp fix

potential multiprocessing fix for macos

fix for chainer inference when cupy is not installed and cpu is not specific

working solution

typo

potential fix for path

fix for deadlock issue on macOSX

removed silly argument thingy. Refactored

more refactoring

refactoring, adding informative prints for liver vessel segmentation in log, which supports verbose mode

finally working. Now the CLI properly installs using pip and it can be used without any dependency issues across all relevant operating systems

removed wrong import

tried importing Config directly, instead of the script

test

typo

worked, now have to add similar fix for all scripts

added '.' to paths when importing

tried removed redundant py_module

added init file to module

moved file to utils

tried fixing dependencies when installing

Added option to segment the portal and hepatic veins (one class). Liver segmentation is now run in a separate Process, as TF is unable to clear the GPU manually - only when a session is killed, which is only when a Process is finished/killed.

typo fix '.ini' instead of '.nii'. Also fixed casting to reduce memory usage

added check to skip non-nifti files in for loop

added batch mode. Can now provide directory of CTs/.nii files to be predicted on. Also refactored build workflow

added checks

CLI works well on Windows

potential bug fix

debugging

added GPU support and bumped tensorflow to v2.4

improved argparse stuff