Support for GE P-files

Support for GE P-files in Suspect is provided by the GE Orchestra library. This has the advantage of maximum compatibility with the many P-file revisions out there, but the disadvantage that Orchestra is not a freely distributable library: GE’s current license prohibits the sharing of even the compiled library. Therefore, in order to enable P-file support, you must download the Orchestra library yourself and add it to the OpenMRSLab Docker container. This is a relatively simple process, and the rest of this page gives a detailed guide to the necessary steps.

Step 1. Register for a GE MR Collaboration account

Go to the GE MR Collaboration Community website and register for an account. This account is free, but only available for users of GE scanners, so there is a manual confirmation step which may take a day or two before your account is active. As well as letting you download Orchestra, this account will also give you access to GE’s MR forums.

Step 2. Download the Orchestra Python SDK

On the GE MR Collaboration website, go to https://collaborate.mr.gehealthcare.com/docs/DOC-1727 to download the Orchestra SDK for Python (current version tested with Suspect is v1.10 on 01/10/20). This is a rather large download at around 416MB, this is mostly due to the large amount of sample data included with the code, the actual code library is only 10% of that, unfortunately there is no way to get only the SDK itself.

Step 3. Copy the code into OpenMRSLab

After extracting the Orchestra files from the downloaded archive, you will find that as well as the example processing scripts and datasets, there are several versions of the code library, covering Windows, MacOS and Linux. The file that we need is the Linux version called GERecon.so.python35. This is because the OpenMRSLab Docker container is based on Linux, no matter what the host operating system for your computer is.

Inside the Docker container, OpenMRSLab is already set up to look for the Orchestra .so file in a directory at the path: /home/jovyan/work/orchestra. If you have followed the recommended approach of mapping a host directory to the /home/jovyan/work directory then the simplest solution is simply to create a folder called orchestra in that mapped directory and copy the GERecon.so.python35 file into it. Note that it is important to rename the file GERecon.so (dropping the .python35 part) to allow Python to recognise and load it.

If you do not have a mapped directory, or if you are accessing your OpenMRSLab installation remotely and don’t have easy access to the host computer file system then it is also possible to upload the file through the Jupyter server itself. The home directory for the Jupyter server is /home/jovyan/work so using the Jupyter file browser interface you can first select New->Folder (from the top right hand corner of the interface) and name the resulting folder orchestra, then click Upload (next to the New button) and select the GERecon.so.python35 file from your local system. Again, note that it is important to rename the file to GERecon.so.

Step 4. Test the installation

To test your installation, simply open a Jupyter notebook and run import GERecon. If you have correctly installed the SDK then this will complete without an error, and you will then be able to use Suspect to work with GE P-files.