Demo minimum requirement

v0.3 - moved from #9

• Each demo will have an independent folder directly under the 'Demos';
• Name the folder as [loader-type]_[image-modality]_[organ-disease]_[optional:brief-remark], e.g. unpaired_ultrasound_prostate or grouped_mr_brain_logitudinal;
• For simplicity and ease to convert to notebooks later, avoid sub-folders and separate files for additional functions/classes;
• Experiment using cross-validation or random splitting is NOT encouraged, unless the purpose of the demo is to demonstrate how to design experiments.

Open accessible data

• Each demo should have a 'demo_data.py' script to automatically download demo data;
• Data should be hosted in a reliable and efficient (not stored in this repo please) online storage, Kaggle, GitHub and zenodo are all options for non-login access (avoid google drive for known accessibility issues);
• Relevant dataset folder structure to utilise the supported loaders can be either pre-arranged in data source or scripted after downloading; Avoid slow and large data set download.

Training

• Each demo should have a 'demo_train.py' script;
• This is accompanied by a config file in the same folder.

Predicting

• Each demo should have a 'demo_predict.py' script;
• Ideally, a pre-trained model will be available for downloading, e.g. the same as your data (not stored in this repo please);
• Results: Provide at least one piece of numerical metric (Dice, distance error, etc) and one piece of visualisation to show the efficacy of the registration (optimum performance is not required here).

A 'readme.md' file

• Author name and email;
• Briefly describe the clinical application and the need for registration;
• Acknowledge data source.

Code style

• Please restrict using external libraries or anything unsupported by Colab or Azure;
• See general Contribution Guide.