Experiments on ADNI Database

Click here for more information about the database.

Preparation

Get the raw data

Get raw data from here.

Serialize the raw data

Use the following command to run the serialization:

python -m datasets.brain.serialize -m sacred

Unconditioned brain GAN

The following experiment shows how to generate random brains with the GanEstimator module. For implementation details see Components of Reproducible-ML or API Documentation

To run the code with the default parameters defined in the config function just type from the root directory:

python -m exps.brain.train

Or use the command line arguments to set different parameters:

python -m exps.brain.train with brain_feeder.batch_size=16

Example output

With the default parameters we will obtain the following result for training steps 100, 1000 and 2000: