NGC | Catalog
CatalogModelsMONAI Swin UNETR BTCV Segmentation

MONAI Swin UNETR BTCV Segmentation

For downloads and more information, please view on a desktop device.
Logo for MONAI Swin UNETR BTCV Segmentation


A pre-trained model for volumetric (3D) multi-organ segmentation from CT image.



Latest Version



April 4, 2023


244.51 MB

Model Overview

A pre-trained Swin UNETR [1,2] for volumetric (3D) multi-organ segmentation using CT images from Beyond the Cranial Vault (BTCV) Segmentation Challenge dataset [3].

model workflow


The training data is from the BTCV dataset (Register through Synapse and download the Abdomen/

  • Target: Multi-organs
  • Task: Segmentation
  • Modality: CT
  • Size: 30 3D volumes (24 Training + 6 Testing)


The dataset format needs to be redefined using the following commands:

mv RawData/Training/img/ RawData/imagesTr
mv RawData/Training/label/ RawData/labelsTr
mv RawData/Testing/img/ RawData/imagesTs

Training configuration

The training as performed with the following:

  • GPU: At least 32GB of GPU memory
  • Actual Model Input: 96 x 96 x 96
  • AMP: True
  • Optimizer: Adam
  • Learning Rate: 2e-4


1 channel

  • CT image


14 channels:

  • 0: Background
  • 1: Spleen
  • 2: Right Kidney
  • 3: Left Kideny
  • 4: Gallbladder
  • 5: Esophagus
  • 6: Liver
  • 7: Stomach
  • 8: Aorta
  • 9: IVC
  • 10: Portal and Splenic Veins
  • 11: Pancreas
  • 12: Right adrenal gland
  • 13: Left adrenal gland


Dice score was used for evaluating the performance of the model. This model achieves a mean dice score of 0.8269

Training Loss

The figure shows the training loss curve for 10K iterations.

Validation Dice

A graph showing the validation mean Dice for 5000 epochs.

MONAI Bundle Commands

In addition to the Pythonic APIs, a few command line interfaces (CLI) are provided to interact with the bundle. The CLI supports flexible use cases, such as overriding configs at runtime and predefining arguments in a file.

For more details usage instructions, visit the MONAI Bundle Configuration Page.

Execute training:
python -m monai.bundle run training --meta_file configs/metadata.json --config_file configs/train.json --logging_file configs/logging.conf
Override the train config to execute multi-GPU training:
torchrun --standalone --nnodes=1 --nproc_per_node=2 -m monai.bundle run training --meta_file configs/metadata.json --config_file "['configs/train.json','configs/multi_gpu_train.json']" --logging_file configs/logging.conf

Please note that the distributed training-related options depend on the actual running environment; thus, users may need to remove --standalone, modify --nnodes, or do some other necessary changes according to the machine used. For more details, please refer to pytorch's official tutorial.

Override the train config to execute evaluation with the trained model:
python -m monai.bundle run evaluating --meta_file configs/metadata.json --config_file "['configs/train.json','configs/evaluate.json']" --logging_file configs/logging.conf
Execute inference:
python -m monai.bundle run evaluating --meta_file configs/metadata.json --config_file configs/inference.json --logging_file configs/logging.conf
Export checkpoint to TorchScript file:

TorchScript conversion is currently not supported.


[1] Hatamizadeh, Ali, et al. "Swin UNETR: Swin Transformers for Semantic Segmentation of Brain Tumors in MRI Images." arXiv preprint arXiv:2201.01266 (2022).

[2] Tang, Yucheng, et al. "Self-supervised pre-training of swin transformers for 3d medical image analysis." arXiv preprint arXiv:2111.14791 (2021).

[3] Landman B, et al. "MICCAI multi-atlas labeling beyond the cranial vault–workshop and challenge." In Proc. of the MICCAI Multi-Atlas Labeling Beyond Cranial Vault—Workshop Challenge 2015 Oct (Vol. 5, p. 12).


This training and inference pipeline was developed by NVIDIA. It is based on a model developed by NVIDIA researchers. This software has not been cleared or approved by FDA or any regulatory agency. MONAI pre-trained models are for developmental purposes only and cannot be used directly for clinical procedures.


Copyright (c) MONAI Consortium

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.