weakly

Weakly-supervised registration tutorial

Yipeng Hu

Tutorial

The complete tutorial can be run on Azure.

Local version

If you would like to run locally, there are two parts, the Training on Notebook and the Inference on Notebook.

For example, you can use a tested enviroment based on Anaconda:

Instructions for Anaconda enviroment settings:

1 - Install Anaconda

Download

2 - Install tensorflow

In Anaconda Prompt, type:

conda create -n tensorflow_env tensorflow
conda activate tensorflow_env

or

conda create -n tensorflow_gpuenv tensorflow-gpu
conda activate tensorflow_gpuenv

3 - Install nibabel

conda install -c conda-forge nibabel

4 - Install notebook in tensorflow_env

Use the Anaconda Navigator:
Change the enviroment to tensorflow_env (or tensorflow_gpuenv);
Click to install Jupyter Notebook.

5 - Open the notebook

In Anaconda Navigator, click to launch the notebook. Or,
In Anaconda Prompt, type:

conda activate tensorflow_env
jupyter notebook

Original version

This tutorial was adapted from the demo code on GitHub.