Avinash

@johnnychhsu were you able to use placeholders with niftynet pipeline? I'm unable to restore the model when using a placeholder, please let me know you fixed this issue. Thank you

Hi @gmarceca, @Talkwarrior, did you find any fix for the problem?

@ndvbd how did you fix this issue?

Currently incorporated HD-BET tool for skull stripping, will be updating our custom model very soon...

check our BioExp framework for more detailed interpretability analysis including uncertainty and causality based analysis. Framework: https://github.com/koriavinash1/BioExp Interpretability Experiments: https://github.com/koriavinash1/BioExp_Experiments

initial implementation: https://github.com/koriavinash1/BioExp/blob/master/BioExp/helpers/visualizations.py

Hey @zhongzisha, Can you share a script for an end to end segmentation generation given a raw CT and PET registered volumes? Thank you

Hi @sunshiding, This repo was used for all our initial experiments if you wish to finetune the model or train from scratch, please check https://github.com/koriavinash1/DeepBrainSeg/blob/master/examples/finetune.py the example provides information about...

removed that, it was just to limit the iterations for 3D network

yes, they are the extracted patches for training