Cardiac T1 and T2 Mapping Segmentation

Source code for our paper "Automated T1 and T2 mapping segmentation on cardiovascular magnetic resonance imaging using deep learning".

Setup Docker Container

For accessing the container by directly connecting to it via ssh:

1. Create a keypair, copy the public key to the root of this repo and name it cm-docker.pub!
2. Run make ssh.
3. Connect on port 2233 ssh root@<hostname> -i <private_key_paht> -p 2233.

To run the container without starting an ssh server, run make run.

To customize docker build and run edit the Makefile.

⚠️ make ssh and make run starts the container with the --rm flag! Only contents of the /workspace persists if the container is stopped (via a simple volume mount)!

How to run

Supervised segmentation training

PYTHONPATH=. python supervised_segmentation/train.py -c supervised_segmentation/config.yaml

Evaluation

Edit the supervised_segmentation/inference.py by adding your checkpoint path and run the following command:

PYTHONPATH=. python supervised_segmentation/inference.py