Build FAST on Ubuntu Linux

Install requirements

1. First, make sure you have all the tools necessary to download and compile the code:

   sudo snap install cmake --classic # Use snap to get more recent version of cmake on Ubuntu 18.04
   sudo apt install g++ git patchelf
   sudo apt install '^libxcb.*-dev' libx11-xcb-dev libglu1-mesa-dev libxrender-dev libxi-dev libxkbcommon-dev libxkbcommon-x11-dev
   sudo apt install pkgconf libusb-1.0-0-dev # Needed for realsense

2. Install OpenCL and OpenGL.
   • OpenGL: Usually installed along with your graphics driver.
   • OpenCL: To install OpenCL on Linux, download an implementation depending on the CPU/GPU you have:
     • NVIDIA - Install CUDA
     • Intel - Install the OpenCL NEO driver
     • AMD - Install the ROCm stack
     • If none of the above fits, you might want to check out the open source OpenCL implementation Portable Computing Lanauge (PCOL), although reduced performance is likely.
3. FAST will download all other dependencies (Qt5, eigen, zlib, DCMTK, OpenVINO, tensorflow, +++) automatically. Note that some optional requirements are needed for video streaming and GPU neural network inference, this can be installed later.

Clone and Configure

Clone the code using Git and configure the project using CMake:

git clone https://github.com/smistad/FAST.git
cd FAST
mkdir build
cd build
# Default cmake configuration. Options may be added like so: 
# cmake .. -DFAST_BUILD_TESTS=OFF -DFAST_BUILD_EXAMPLES=ON -DFAST_MODULE_TensorFlow=ON
cmake ..

FAST has several optional modules and build options. These are enabled using cmake options named FAST_MODULE_<Name> and FAST_BUILD_<Name>. Here is a list of some options which might be useful:

• FAST_BUILD_TOOLS
• FAST_BUILD_EXAMPLES
• FAST_BUILD_TESTS
• FAST_BUILD_DOCS
• FAST_MODULE_TensorFlow
• FAST_MODULE_TensorRT
• FAST_MODULE_OpenVINO
• FAST_MODULE_Dicom
• FAST_MODULE_WholeSlideImaging
• FAST_MODULE_OpenIGTLink
• FAST_MODULE_Clarius
• FAST_MODULE_Python
• FAST_MODULE_HDF5
• FAST_MODULE_Plotting
• FAST_MODULE_RealSense

Compile

This will build all configured targets:

make -j8 # Use 8 parallel jobs for compiling

Test

To test that your build was successful, run the systemCheck application:

./bin/systemCheck

This should display the FAST logo and some FAST+OpenCL information.

Troubleshoot

If you get the error saying something like "qfontengine_ft_p.h:56:22: fatal error: ft2build.h: No such file or directory" you have to install the fontconfig library "libfontconfig1-dev".

If you get an error like "ERROR: Feature 'xcb' was enabled, but the pre-condition 'features.thread && features.xkbcommon && libs.xcb' failed." This means you forgot to install the xcb an xkb libraries needed: sudo apt install '^libxcb.*-dev' libx11-xcb-dev libglu1-mesa-dev libxrender-dev libxi-dev libxkbcommon-dev

Still stuck? Get help at GitHub discussions.

Install

Set CMAKE_INSTALL_PREFIX to where you want to install FAST on your drive. Then run:

make -j8 install

Build Debian package

To build a Debian package run:

make -j8 package

Build the Python bindings (pyFAST)

Install the python development libraries and some dependencies:

sudo apt install python3 libpython3-dev python3-pip python3-setuptools
sudo pip3 install --upgrade pip
pip3 install pylddwrap==1.2.0 twine wheel==0.37.1

Then configure cmake with Python enabled: cmake .. -DFAST_MODULE_Python=ON

Finally, build the python-wheel target: make -j8 python-wheel

The wheel will appear in the python/dist folder.

Build the documentation

Install dependencies:

sudo apt install bison flex # Needed by Doxygen
pip3 install jinja2         # Needed by M.css

Configure cmake for documentation:

mkdir build-docs
cd build-docs
cmake ../doc/ # This will only build the documentation of FAST
make -j4 documentation

Then build the documentation target: make -j8 documentation

The documentation will be in the folder build-docs/documentation/html/