Repository files navigation

This repository contains the source-code for the paper "Spinal cord gray matter segmentation using deep dilated convolutions", available aspre-print on ArXiv.

Note: this repository is made for researchers in deep learning. If you just would like to use the method on your data, this method has been implemented in theSpinal Cord Toolbox (SCT), where youcan find pre-trained models on much larger datasets and a user-friendly command-line tool called sct_deepseg_gm.

You can see the MRIex-vivosegmentation video. Anothermanuscript is under review for the MRIex-vivo data.

To use this repository, you'll need to install the following requirements:

• Clone the repository
• Install Python requirements withpip install -r pip-requirements.txt
• Open the Jupyter Notebook located atnotebooks folder

This repository contains two notebooks:

• TensorFlow model
• Pytorch model(experimental)

Some remarks regarding the model:

• This model was trained on a common space with a voxel size of 0.25mm x 0.25mm, soyou'll have to resample your data to this space if you want good results;
• This repository contains the model trained on the GM Challenge Dataset (both train and validation),the model is located on the directory calledmodels together with a json file containingthe mean/std that was used to standardize the training data;
• For the training procedure, please see the original paper for more information;

If you use this work in your research, please cite:

@article{arxiv1710.01269,  author = {Christian S. Perone, Evan Calabrese, Julien Cohen-Adad},  title = {Spinal cord gray matter segmentation using deep dilated convolutions},  journal = {arXiv preprint arXiv:1710.01269},  year = {2017}}

MIT LicenseCopyright (c) 2017 NeuroPolyPermission is hereby granted, free of charge, to any person obtaining a copyof this software and associated documentation files (the "Software"), to dealin the Software without restriction, including without limitation the rightsto use, copy, modify, merge, publish, distribute, sublicense, and/or sellcopies of the Software, and to permit persons to whom the Software isfurnished to do so, subject to the following conditions:The above copyright notice and this permission notice shall be included in allcopies or substantial portions of the Software.THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS ORIMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THEAUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHERLIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THESOFTWARE.