Build context for a Flywheel Gear to execute
Chris Rorden's dcm2niix is a popular tool for converting images from the complicated formats used by scanner manufacturers (DICOM, PAR/REC) to the simple NIfTI format used by many scientific tools. dcm2niix works for all modalities (CT, MRI, PET, SPECT) and sequence types.
Build the Image
To build the image, either download the files from this repo or clone the repo:
docker build --no-cache -t scitran/dcm2niix
The input to this gear can be either a zip, a tgz, or a directory containing DICOMs.
Default options are set in
manifest.json and copied to the container on build. Defaults are then loaded by the
run script when the algorithm is executed.
To run dcm2niix locally from this container you can do the following:
Ensure your input file is in it's own directory.
docker run --rm -ti \ -v </path/to/input/data_dir>:/flywheel/v0/input/dcm2niix_input \ -v </path/to/output>:/flywheel/v0/output \ scitran/dcm2niix