Code for running a segmentation model in a Docker container

Code for running preprocessing and segmentation of MR images in a Docker container, for example for integration with the SAINT platform.

The code runs preprocessing of MR DICOM files, then runs a trained segmentation model on the preprocessed data, and finally converts the segmentations from binary volumes to a DICOM RTSTRUCT file to be imported into radiotherapy software.

Docker needs to be installed. MR images need to be in DICOM format.

Anyone working with segmentation in radiotherapy

See GitHub repository