File size: 2,326 Bytes
faf3b5a 3468859 cd73668 faf3b5a 19b6500 faf3b5a aeb290a faf3b5a bd05d51 ea90c07 bd05d51 3468859 faf3b5a bd05d51 3468859 bd05d51 3468859 bd05d51 3468859 faf3b5a 3468859 faf3b5a bd05d51 3468859 faf3b5a ad1ce7c bd05d51 19b6500 bd05d51 3468859 faf3b5a 3468859 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 |
# Automatic liver segmentation in CT using deep learning
[](https://github.com/DAVFoundation/captain-n3m0/blob/master/LICENSE)
[](https://github.com/andreped/livermask/actions)
#### Pretrained U-Net model is automatically downloaded when running the inference script and can be used as you wish, ENJOY! :)
<img src="figures/Segmentation_3DSlicer.PNG" width="70%" height="70%">
## Install
```
pip install git+https://github.com/andreped/livermask.git
```
## Usage:
```
livermask --input path-to-nifti.nii --output path-to-output-file.nii
```
In addition, there is the optional `--cpu` action to disable the GPU (force computations on CPU only) if necessary.
## DICOM/NIfTI format
Pipeline assumes input is in the NIfTI format, and output a binary volume in the same format (.nii).
DICOM can be converted to NIfTI using the CLI [dcm2niix](https://github.com/rordenlab/dcm2niix), as such:
```
dcm2niix -s y -m y -d 1 "path_to_CT_folder" "output_name"
```
Note that "-d 1" assumed that "path_to_CT_folder" is the folder just before the set of DICOM scans you want to import and convert. This can be removed if you want to convert multiple ones at the same time. It is possible to set "." for "output_name", which in theory should output a file with the same name as the DICOM folder, but that doesn't seem to happen...
## Troubleshooting
You might have issues downloading the model when using VPN. If any issues are observed, try to disable VPN and try again.
If the program struggles to install, attempt to install using:
```
pip install --force-reinstall --no-deps git+https://github.com/andreped/livermask.git
```
## Acknowledgements
The model was trained on the LITS dataset. The dataset is openly accessible and can be downloaded from [here](https://competitions.codalab.org). If this tool is used, please, consider citing the corresponding [LITS challenge dataset paper](https://arxiv.org/abs/1901.04056).
Disclaimer, I have no affiliation with the LITS challenge, the LITS dataset, or the challenge paper. I only wish to provide an open, free-to-use tool that people may find useful :)
------
Made with :heart: and python
|