parenchyma of the kidney, kidney neoplasms, kidney arteries, kidney veins, kidney ureters

inference.json

{
    "imports": [
        "$import glob",
        "$import os"
    ],
    "bundle_root": ".",
    "ckpt_dir": "$@bundle_root + '/models'",
    "output_dir": "$@bundle_root + '/eval'",
    "dataset_dir": "$@bundle_root + '/data'",
    "images": "$[{'artery':a, 'vein':b, 'excret':c }for a,b,c in  zip(glob.glob(@dataset_dir + '/*/12.nii.gz'), glob.glob(@dataset_dir + '/*/22-.nii.gz'), glob.glob(@dataset_dir + '/*/32-.nii.gz'))]",
    "labels": "$[]",
    "device": "$torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')",
    "network_def": {
        "_target_": "SegResNet",
        "in_channels": 3,
        "out_channels": 6,
        "init_filters": 32,
        "upsample_mode": "deconv",
        "dropout_prob": 0.2,
        "norm_name": "group",
        "blocks_down": [
            1,
            2,
            2,
            4
        ],
        "blocks_up": [
            1,
            1,
            1
        ]
    },
    "network": "$@network_def.to(@device)",
    "preprocessing": {
        "_target_": "Compose",
        "transforms": [
            {
                "_target_": "LoadImaged",
                "keys": [
                    "artery",
                    "vein",
                    "excret"
                ],
                "image_only": false
            },
            {
                "_target_": "EnsureChannelFirstd",
                "keys": [
                    "artery",
                    "vein",
                    "excret"
                ]
            },
            {
                "_target_": "Orientationd",
                "keys": [
                    "artery",
                    "vein",
                    "excret"
                ],
                "axcodes": "LPS"
            },
            {
                "_target_": "Spacingd",
                "keys": [
                    "artery",
                    "vein",
                    "excret"
                ],
                "pixdim": [
                    0.8,
                    0.8,
                    0.8
                ],
                "mode": "bilinear"
            },
            {
                "_target_": "scripts.my_transforms.ConcatImages",
                "keys_merge": [
                    "artery",
                    "vein",
                    "excret"
                ],
                "keys_out": "image"
            },
            {
                "_target_": "ScaleIntensityRanged",
                "keys": "image",
                "a_min": -1000,
                "a_max": 1000,
                "b_min": 0.0,
                "b_max": 1.0,
                "clip": true
            },
            {
                "_target_": "EnsureTyped",
                "keys": "image"
            }
        ]
    },
    "dataset": {
        "_target_": "Dataset",
        "data": "@images",
        "transform": "@preprocessing"
    },
    "dataloader": {
        "_target_": "DataLoader",
        "dataset": "@dataset",
        "batch_size": 1,
        "shuffle": false,
        "num_workers": 4
    },
    "inferer": {
        "_target_": "SlidingWindowInferer",
        "roi_size": [
            96,
            96,
            96
        ],
        "sw_batch_size": 4,
        "overlap": 0.25
    },
    "postprocessing": {
        "_target_": "Compose",
        "transforms": [
            {
                "_target_": "Invertd",
                "transform": "$@preprocessing",
                "device": "@device",
                "keys": "pred",
                "orig_keys": "artery",
                "meta_keys": "pred_meta_dict",
                "nearest_interp": false,
                "to_tensor": true
            },
            {
                "_target_": "Activationsd",
                "keys": "pred",
                "softmax": false,
                "sigmoid": true
            },
            {
                "_target_": "AsDiscreted",
                "keys": "pred",
                "threshold": 0.5
            },
            {
                "_target_": "scripts.my_transforms.MergeClassesd",
                "keys": "pred"
            },
            {
                "_target_": "SaveImaged",
                "keys": "pred",
                "meta_keys": "pred_meta_dict",
                "data_root_dir": "@dataset_dir",
                "output_dir": "@output_dir"
            }
        ]
    },
    "handlers": [
        {
            "_target_": "CheckpointLoader",
            "load_path": "$@ckpt_dir + '/model.pt'",
            "load_dict": {
                "model": "@network"
            },
            "strict": "True"
        },
        {
            "_target_": "StatsHandler",
            "iteration_log": false
        }
    ],
    "evaluator": {
        "_target_": "SupervisedEvaluator",
        "device": "@device",
        "val_data_loader": "@dataloader",
        "network": "@network",
        "inferer": "@inferer",
        "postprocessing": "@postprocessing",
        "val_handlers": "@handlers",
        "amp": false
    },
    "inference": [
        "$setattr(torch.backends.cudnn, 'benchmark', True)",
        "$@evaluator.run()"
    ]
}

kidneys_monai_bundle.sh

#--------------------------------
# commands on this part need to be executed one by one

curl -fsSL https://pyenv.run | bash

# execute the following line by line
echo 'export PYENV_ROOT="$HOME/.pyenv"' >> ~/.bashrc
echo '[[ -d $PYENV_ROOT/bin ]] && export PATH="$PYENV_ROOT/bin:$PATH"' >> ~/.bashrc
echo 'eval "$(pyenv init - bash)"' >> ~/.bashrc

touch ~/.bash_profile
# execute the following line by line
echo 'export PYENV_ROOT="$HOME/.pyenv"' >> ~/.bash_profile
echo '[[ -d $PYENV_ROOT/bin ]] && export PATH="$PYENV_ROOT/bin:$PATH"' >> ~/.bash_profile
echo 'eval "$(pyenv init - bash)"' >> ~/.bash_profile

exec "$SHELL" # you need to restart the shell

#--------------------------------


MYMONAIDIR=/home/$USER/my_monai
MYPYENVNAME=kidney_bundle2
MYPYENVVERSION=3.10.12
pyenv install $MYPYENVVERSION
pyenv virtualenv $MYPYENVVERSION $MYPYENVNAME
mkdir -p $MYMONAIDIR
cd $MYMONAIDIR
pyenv activate $MYPYENVNAME
# pyenv local $MYPYENVNAME # this line does not work

pip install "monai[all]"

python -m monai.bundle download "renalStructures_CECT_segmentation" --bundle_dir "bundles/"

cd bundles/renalStructures_CECT_segmentation
python -m monai.bundle run download_data --meta_file configs/metadata.json --config_file "['configs/train.json', 'configs/evaluate.json']"
rm -f ./data/*/merged.nii.gz

# please overwrite the configs/inference.json with the other file in this gist

python -m monai.bundle run inference --meta_file configs/metadata.json --config_file configs/inference.json

# check segmentation results on eval folder

Ok. Now it should work for batch inference.

Please have in mind that you need to achieve a folder structure like this for inference:

    "images": [
        {"artery": "./data/patient1/12.nii.gz", "vein": "./data/patient1/22-.nii.gz", "excret": "./data/patient1/32-.nii.gz"},
        {"artery": "./data/patient2/12.nii.gz", "vein": "./data/patient2/22-.nii.gz", "excret": "./data/patient2/32-.nii.gz"},
        {"artery": "./data/patient3/12.nii.gz", "vein": "./data/patient3/22-.nii.gz", "excret": "./data/patient3/32-.nii.gz"}
    ],