Gadgetron Cases of Medical Images Reconstruction

UGadget: A python library, providing some mri reconstruction cases, based on Gadgetron framework

Introduction

UGadget is an application-level library, containing some cases, based on Gadgetron framework, to do reconstruction on MRI images.

Prerequisite

1. clone gadgetron to your workspace

git clone https://github.com/gadgetron/gadgetron.git
cd gadgetron

2. 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

3. 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

4. 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 11:31:57

Now you could see a out.h5 file in the current directory.

Usage

Researchers

For who wants to install this repo directly in your conda environment, without modifying the code:

pip install git+https://github.com/medlab/ugad4fun.git

Developers

For who wants to contribute to this repo, you may need follow steps below to install ugad4fun:

1. clone this repository to your workspace

git clone https://github.com/medlab/ugad4fun.git
cd ugad4fun

2. use pip to install this package in editable mode:

pip install -e .

3. link the config directory to gadgetron-required location

sh scripts/link_config.sh

Deployment

Clone 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

Reference:

1. https://gadgetron2020.sciencesconf.org/
2. https://gadgetron.readthedocs.io/
3. https://github.com/gadgetron/GadgetronOnlineClass
4. https://github.com/chidiugonna/learn-gadgetron
5. https://github.com/ismrmrd/ismrmrdviewer