NGC | Catalog
CatalogModelsMONAI Spleen CT Segmentation

MONAI Spleen CT Segmentation

For downloads and more information, please view on a desktop device.
Logo for MONAI Spleen CT Segmentation


A pre-trained model for volumetric (3D) segmentation of the spleen from CT image.



Latest Version



April 4, 2023


36.95 MB

Model Overview

A pre-trained model for volumetric (3D) segmentation of the spleen from CT images.

This model is trained using the runner-up [1] awarded pipeline of the "Medical Segmentation Decathlon Challenge 2018" using the UNet architecture [2] with 32 training images and 9 validation images.

model workflow


The training dataset is the Spleen Task from the Medical Segmentation Decathalon. Users can find more details on the datasets at

  • Target: Spleen
  • Modality: CT
  • Size: 61 3D volumes (41 Training + 20 Testing)
  • Source: Memorial Sloan Kettering Cancer Center
  • Challenge: Large-ranging foreground size

Training configuration

The segmentation of spleen region is formulated as the voxel-wise binary classification. Each voxel is predicted as either foreground (spleen) or background. And the model is optimized with gradient descent method minimizing Dice + cross entropy loss between the predicted mask and ground truth segmentation.

The training was performed with the following:

  • GPU: at least 12GB of GPU memory
  • Actual Model Input: 96 x 96 x 96
  • AMP: True
  • Optimizer: Adam
  • Learning Rate: 1e-4
  • Loss: DiceCELoss


One channel

  • CT image


Two channels

  • Label 1: spleen
  • Label 0: everything else


Dice score is used for evaluating the performance of the model. This model achieves a mean dice score of 0.96.

Training Loss

A graph showing the training loss over 1260 epochs (10080 iterations).

Validation Dice

A graph showing the validation mean Dice over 1260 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
Override the train config and evaluate config to execute multi-GPU evaluation:
torchrun --standalone --nnodes=1 --nproc_per_node=2 -m monai.bundle run evaluating --meta_file configs/metadata.json --config_file "['configs/train.json','configs/evaluate.json','configs/multi_gpu_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


[1] Xia, Yingda, et al. "3D Semi-Supervised Learning with Uncertainty-Aware Multi-View Co-Training." arXiv preprint arXiv:1811.12506 (2018).

[2] Kerfoot E., Clough J., Oksuz I., Lee J., King A.P., Schnabel J.A. (2019) Left-Ventricle Quantification Using Residual U-Net. In: Pop M. et al. (eds) Statistical Atlases and Computational Models of the Heart. Atrial Segmentation and LV Quantification Challenges. STACOM 2018. Lecture Notes in Computer Science, vol 11395. Springer, Cham.


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.