qicq1c/SuPreM

Instructions

1-Data preparation

This is how inputs_data organizes

    $inputs_data/
    ├── casename00001
    │   └── ct.nii.gz
    ├── casename00002
    │   └── ct.nii.gz
    ├── casename00003
    │   └── ct.nii.gz
    ...

2-Download

You can choose to use the docker image (recommand) or the singularity container.

Download the docker image.

docker pull qchen99/suprem:v1

or

Download the singularity container.

wget https://huggingface.co/qicq1c/SuPreM/resolve/main/suprem_final.sif

3-Inference

You can directly perform inference on your own data. Simply modify inputs_data into your data path and adjust outputs_data to specify the desired output location for the segmentation results.

Use Docker

sudo docker container run --gpus "device=0" -m 128G --rm -v $inputs_data:/workspace/inputs/ -v $outputs_data:/workspace/outputs/ qchen99/suprem:v1 /bin/bash -c "sh predict.sh"

or

Use Singularity

SINGULARITYENV_CUDA_VISIBLE_DEVICES=0 singularity run --nv -B $inputs_data:/workspace/inputs -B $outputs_data:/workspace/outputs suprem_final.sif

This is how outputs_data organizes

    $outputs_data/
    ├── casename00001
    ├── casename00002
    ├── casename00003
        │── combined_labels.nii.gz
        └── segmentations
            ├── aorta.nii.gz
            ├── gall_bladder.nii.gz
            ├── kidney_left.nii.gz
            ├── kidney_right.nii.gz
            ├── liver.nii.gz
            ├── pancreas.nii.gz
            ├── postcava.nii.gz
            ├── spleen.nii.gz
            ├── stomach.nii.gz
    │
    ...