UGadget: A python library, providing some mri reconstruction cases, based on Gadgetron framework
UGadget is an application-level library, containing some cases, based on Gadgetron framework, to do reconstruction on MRI images.
- clone gadgetron to your workspace
git clone https://github.com/gadgetron/gadgetron.git
cd gadgetron- create a conda environment using environment.yml
conda env create -f environment.yml If this step cost forever, you may try mamba, it may save you:
conda activate
conda install mamba -c conda-forge
mamba env create -f environment.yml- start gadgetron environment, then build gadgetron
conda activate gadgetron
mkdir -p build
cd build
cmake -GNinja -DCMAKE_BUILD_TYPE=Release -DUSE_MKL=ON -DCMAKE_INSTALL_PREFIX=${CONDA_PREFIX} ../
ninja
ninja install- test gadgetron
# check gadgetron info
$ gadgetron --info
02-17 10:50:55.456 DEBUG [gadgetron_paths.cpp:117] Executable path: "/home/uih/miniconda3/envs/gadgetron/bin/gadgetron"
02-17 10:50:55.456 DEBUG [gadgetron_paths.cpp:123] Default Gadgetron home: "/home/uih/miniconda3/envs/gadgetron"
02-17 10:50:55.456 WARNING [initialization.cpp:38] Environment variable 'OMP_WAIT_POLICY' not set to 'PASSIVE'.
02-17 10:50:55.456 WARNING [initialization.cpp:39] Gadgetron may experience serious performance issues under heavy load (multiple simultaneous reconstructions, etc.)
02-17 10:50:56.418 INFO [system_info.cpp:104] CUDA DEVICE COUNT 0 and error number 35
02-17 10:50:56.418 INFO [main.cpp:86] Gadgetron Version Info
-- Version : 4.4.5
-- Git SHA1 : 41dd90ee85ffe5fd3a0cc6a43422854541931be6
-- System Memory size : 11964 MB
-- Python Support : YES
-- Julia Support : NO
-- Matlab Support : NO
-- CUDA Support : YES
-- NVCC Flags : -gencode arch=compute_60,code=sm_60;-gencode arch=compute_61,code=sm_61;-gencode arch=compute_70,code=sm_70;-gencode arch=compute_75,code=sm_75;-gencode arch=compute_80,code=sm_80;-gencode arch=compute_86,code=sm_86 --std=c++17
* Number of CUDA capable devices: 0
# start a gadgetron server
$ gadgetron
02-17 10:50:31.535 DEBUG [gadgetron_paths.cpp:117] Executable path: "/home/uih/miniconda3/envs/gadgetron/bin/gadgetron"
02-17 10:50:31.535 DEBUG [gadgetron_paths.cpp:123] Default Gadgetron home: "/home/uih/miniconda3/envs/gadgetron"
02-17 10:50:31.535 WARNING [initialization.cpp:38] Environment variable 'OMP_WAIT_POLICY' not set to 'PASSIVE'.
02-17 10:50:31.535 WARNING [initialization.cpp:39] Gadgetron may experience serious performance issues under heavy load (multiple simultaneous reconstructions, etc.)
02-17 10:50:31.535 INFO [main.cpp:90] Gadgetron 4.4.5 [41dd90ee85ffe5fd3a0cc6a43422854541931be6]
02-17 10:50:31.536 DEBUG [storage.cpp:68] Found storage server: /home/uih/miniconda3/envs/gadgetron/bin/mrd-storage-server
02-17 10:50:31.536 INFO [storage.cpp:69] Starting storage server on port 9112
02-17 10:50:31.537 INFO [storage.cpp:28] Verifying connectivity to storage server...
{"level":"info","time":"2023-02-17T10:50:31.582+08:00","message":"Listening on port 9112"}
{"level":"info","requestId":"9c1d8c90-ab1f-4fd7-a963-89103fc90ef6","status":200,"method":"GET","path":"/healthcheck","query":"","latencyMs":1.56,"time":"2023-02-17T10:50:31.743+08:00","message":"request completed"}
02-17 10:50:31.743 INFO [storage.cpp:33] Received successful response from storage server.
02-17 10:50:31.743 INFO [main.cpp:99] Running on port 9002
02-17 10:50:31.743 INFO [Server.cpp:25] Gadgetron home directory: "/home/uih/miniconda3/envs/gadgetron"
02-17 10:50:31.743 INFO [Server.cpp:26] Gadgetron working directory: "/tmp/gadgetron/"Generate an ismrmrd raw data and do reconstruction using gadgetron
# generate a dataset in the current working directory (./testdata.h5) with 8 coils and 10 repetitions.
$ ismrmrd_generate_cartesian_shepp_logan -r 10
Generating Cartesian Shepp Logan Phantom!!!
Acceleration: 1
$ gadgetron_ismrmrd_client -f testdata.h5
Gadgetron ISMRMRD client
-- host : localhost
-- port : 9002
-- hdf5 file in : testdata.h5
-- hdf5 group in : /dataset
-- conf : default.xml
-- loop : 1
-- hdf5 file out : out.h5
-- hdf5 group out : 2023-02-17 11:31:57Now you could see a out.h5 file in the current directory.
For who wants to install this repo directly in your conda environment, without modifying the code:
pip install git+https://github.com/medlab/ugad4fun.gitFor who wants to contribute to this repo, you may need follow steps below to install ugad4fun:
- clone this repository to your workspace
git clone https://github.com/medlab/ugad4fun.git
cd ugad4fun- use pip to install this package in editable mode:
pip install -e .- link the config directory to gadgetron-required location
sh scripts/link_config.shClone the repository to your workspace, then create a Docker image using scripts provided.
git clone https://github.com/medlab/ugad4fun.git
cd ugad4fun/scripts
bash create_gad_image.sh
bash start_gad_container.sh