A pre-trained model for 3D segmentation of the spleen organ from CT images using DeepEdit.
DeepEdit is an algorithm that combines the power of two models in one single architecture. It allows the user to perform inference as a standard segmentation method (i.e., UNet) and interactively segment part of an image using clicks [2]. DeepEdit aims to facilitate the user experience and, at the same time, develop new active learning techniques.
The model was trained on 32 images and validated on 9 images.
The training dataset is the Spleen Task from the Medical Segmentation Decathalon. Users can find more details on the datasets at http://medicaldecathlon.com/.
The training as performed with the following:
Three channels
Two channels
Dice score is used for evaluating the performance of the model. This model achieves a dice score of greater than 0.90, depending on the number of simulated clicks.
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.
python -m monai.bundle run training --meta_file configs/metadata.json --config_file configs/train.json --logging_file configs/logging.conf
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.
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
python -m monai.bundle run evaluating --meta_file configs/metadata.json --config_file configs/inference.json --logging_file configs/logging.conf
[1] Diaz-Pinto, Andres, et al. DeepEdit: Deep Editable Learning for Interactive Segmentation of 3D Medical Images. MICCAI Workshop on Data Augmentation, Labelling, and Imperfections. MICCAI 2022.
[2] Diaz-Pinto, Andres, et al. "MONAI Label: A framework for AI-assisted Interactive Labeling of 3D Medical Images." arXiv preprint arXiv:2203.12362 (2022).
[3] Sakinis, Tomas, et al. "Interactive segmentation of medical images through fully convolutional neural networks." arXiv preprint arXiv:1903.08205 (2019).
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
http://www.apache.org/licenses/LICENSE-2.0
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.