Releases: MaastrichtU-CDS/hn_cnn
Release v1.5.0
Due to vulnerabilities, we updated python and the versions for the libraries used.
Previous docker images (pmateus/hn-cnn:1.4.0) and versions (branch 'reproduce') of the code still allow to recreate the environment used to train the model.
New docker image: pmateus/hn-cnn:1.5.0
Full Changelog: v1.4.0...v1.5.0
Release 1.4.0
- Improved the scripts to run the model.
- Added scripts to predict/classify using one of the models available.
- Included additional information on dependencies and docker image for the Jupyter Notebook and VS code environment used.
- Create docker image for reproducibility: pmateus/hn-cnn:1.4.0-rep2
Full Changelog: v1.2.0...v1.4.0
Docker image: pmateus/hn-cnn:1.4.0
Release 1.2.0
- Fixed a bug when saving the model.
- Improved documentation regarding the pre-processing.
New docker image: pmateus/hn-cnn:1.2.0
Release 1.1.0
Now including the imaging pre-processing scripts in the package: allows to automatically pre-process (choice of the CT slice, cropping and windowing the slice) before starting the training. With this change, it's possible to use an input of any size.
Reselase 1.0.0
First release with the scripts to train, validate, and test the CNN for head and neck cancer prognosis.
Includes the necessary scripts to pre-process the data and the seeds to reproduce the original work.
Official docker image: pmateus/hn-cnn:1.0.0