Code from the ML4MI Boot Camp.
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.


Code from the UW-Madison Machine Learning for Medical Imaging (ML4MI) Boot Camp. For more information about ML4MI go to:

All the excercises are written in Keras which is corrently integrated into tensorflow. We use the Keras functional model which is a lot more flexible than the commonly used sequential model in examples.

Keras Documentations:

Bootcamp organizers:

System Requirments

Code has been tested on a machine with a NVIDIA K80 (11gb of GPU ram). To run this you need: python 3 ( ) tensorflow ( , install tensorflow-gpu if you have one)

We installed these with the following commands.

pip install tensorflow-gpu
pip install keras
pip install matplotlib
pip install numpy
pip install livelossplot
pip install conda
pip install jupyterlab
conda install scikit-image
conda install scipy
conda install -c conda-forge --no-deps pydicom

Colab from Google Research

Some of these will run on the Google research supported Colab. This is a free cloud based enviroment supported by Google. You can click on the link in the source code or go to

Working Examples:

  • FunctionFitting - Some very basic networks used for learning functions
  • ImageReconstruction - Training of an neural network to reconstruct MRI images using 1D operations
  • MaleFemaleRadiograph - Classify chest xrays as male or female
  • ImageSegmentation - Lung segmentation from CT data (need to download data yourself)

Examples missing data (work in progress):

  • AgeRegression - Regression for Age
  • ImageSynthesis - Image synthesis of brats data

Note on commits:

If you aim to push changes to this repository, please edit the jupyter notebooks and then run clear_and_convert_bash (in linux or WSL). This will convert the notebooks to python and clear the ouput.