This repository provides a trained HighRes3DNet1 for prostate segmentation. Training and inference is facilitated by NiftyNet2.
HighRes3DNet was trained using 79 T2-weighted images (T2WI) from the PICTURE study dataset3 and 47 T2WI from the PROMISE12 dataset4. On a small validation set (N=4), HighRes3DNet achieved a mean Dice score of 0.90.
If you use this repository, please cite the following publication:
Note: NiftyNet is not actively maintained anymore.
-
Clone/download this repository.
-
Create a Python virtual environment:
conda create -n prostate_segmentation_highres3dnet_venv python=3.6.3
-
Activate the virtual environment:
conda activate prostate_segmentation_highres3dnet_venv
-
Install CUDA Toolkit (required for running TensorFlow GPU):
conda install -c anaconda cudatoolkit=9.0
-
Locate cudnn64_7.dll and copy+paste into path e.g. in environment bin <FULL_PATH>\prostate_segmentation_highres3dnet_venv\Library\bin.
-
Install requirements, chiefly NiftyNet==0.3.0 and tensorflow-gpu==1.7:
pip install -r requirements.txt
-
Activate the virtual environment if not already activated:
conda activate prostate_segmentation_highres3dnet_venv
-
Change directory into repository.
-
Add T2WI for inference to: .\sample_data\sample_t2w
-
Run the following command:
net_segment inference -c .\configs\config_infer.ini -a niftynet.application.segmentation_application.SegmentationApplication
1 Li, W.; Wang, G.; Fidon, L.; Ourselin, S.; Cardoso, M.J.; Vercauteren, T. On the compactness, efficiency, and representation of 3D convolutional networks: Brain parcellation as a pretext task. In Proceedings of the Information Processing in Medical Imaging (IPMI 2017); 2017; Vol. 10265, pp. 348–360.
2 Gibson, E.; Li, W.; Sudre, C.; Fidon, L.; Shakir, D.I.; Wang, G.; Eaton-Rosen, Z.; Gray, R.; Doel, T.; Hu, Y.; et al. NiftyNet: a deep-learning platform for medical imaging. Comput. Methods Programs Biomed. 2018, 158, 113–122.
3 Simmons, L.A.M.; Kanthabalan, A.; Arya, M.; Briggs, T.; Barratt, D.; Charman, S.C.; Freeman, A.; Gelister, J.; Hawkes, D.; Hu, Y.; et al. The PICTURE study: Diagnostic accuracy of multiparametric MRI in men requiring a repeat prostate biopsy. Br. J. Cancer 2017, 116, 1159–1165.
4 Litjens, G.; Toth, R.; van de Ven, W.; Hoeks, C.; Kerkstra, S.; van Ginneken, B.; Vincent, G.; Guillard, G.; Birbeck, N.; Zhang, J.; et al. Evaluation of prostate segmentation algorithms for MRI: the PROMISE12 challenge. Med. Image Anal. 2014, 18, 359–373.