Toy project to send time consuming BART command like ecalib
to gadgetron Server from MATLAB(Windows).
Experimental Docker container file has to be compiled.
cd docker
Docker build -t gadgetron_bart:devel .
docker save and load image
docker save mynewimage > gadgetron_bart.tar
scp gadgetron_bart user@192.168.2.1:/home/user
docker load < gadgetron_bart.tar
start the container: with ecalib.py on the gadgetron python path
docker run --rm -ti --net=host --volume /home/meduser/ecalib:/usr/local/share/gadgetron/python a40abc
at the moment we need to set TOOLBOX_PATH
and start gadgetron manually while running the container
export TOOLBOX_PATH=/opt/code/bart/
gadgetron -p 9020
Ideally this should be done in dockerfile.
now only ecalib is supported now.
See matalb\demo.m
to get started!.
The calibration data is Encapsulated into a ISMRMRD data file(h5) with BART command ( see matlab\calib2ismrmrd.m
) and sent to the gadgetron container we built before. You can also execute the system call on a command prompt to continue using MATLAB!
gadgetron_ismrmrd_client.exe -f Calibdata.h5 -C ecalib.xml -a 192.168.2.1 -p 9020 -o test1out.h5