MRI Reconstruction C++ Demo

This demo demonstrates MRI reconstruction model described in and implemented in The model is used to restore undersampled MRI scans which is useful for data compression.

Supported Models

  • hybrid-cs-model-mri


Refer to the tables Intel’s Pre-Trained Models Device Support and Public Pre-Trained Models Device Support for the details on models inference support at different devices.


  1. Running the application with the -h option yields the following usage message:

    $ ./mri_reconstruction_demo -h
    mri_reconstruction_demo [OPTION]
        -h                                Print a usage message.
        -i "<path>"                       Required. Path to input .npy file with MRI scan data.
        -p "<path>"                       Required. Path to sampling mask in .npy format.
        -m "<path>"                       Required. Path to an .xml file with a trained model.
        -d "<device>"                     Optional. Specify the target device to infer on; CPU or GPU is acceptable (CPU by default).
        --no_show                         Optional. Disable results visualization.
  2. To run the demo, you need to have