This is an implementation of Continuous Connectivity from 1 and 2. It is meant to be a parcellation free representation of brain connectivity, i.e. a form of connecitivty outside of the discrete network frame. If none of that made any sense to you, we strongly recommend you, we strongly recommend you read the papers and email the first author (Daniel Moyer, moyerd [at] usc.edu) with any questions.

This project will be expanding as time goes on, but for now is clearly in pre-alpha! Sorry! DM will add tutorials and hopefully real documentation soon.

-A gcc compiler that supports c++11
-cmake
-OpenMP
-The Eigen library from http://eigen.tuxfamily.org/. This is only used by third party software.

Step 0: Install required software from Requirements. Don't forget to add Eigen's library to your CPLUS_INCLUDE_PATH, or equivalent. This is usually done on Unix systems by adding the following to your .bashrc:

export CPLUS_INCLUDE_PATH="/usr/include/eigen3/:${CPLUS_INCLUDE_PATH}"

For non-Ubuntu systems this path may vary.

Step 1: Create a build directory wherever you would like to build concon.

Step 2: Navigate to the build directory and run cmake /path/to/concon/repo/, with the appropriate path.

Step 3: Run make.

The binaries and static libraries for concon will be built in bin and lib.

We include the spherical harmonics package from here. Their original build instructions used Bazel; we have created a short cmake recipe. It is otherwise unchanged. Please see their License file for the corresponding copyright information.

Please see our License file (MIT).

[1] D Moyer et al. Continuous Representations of Brain Connectivity using Spatial Point Processes, Medical Image Analysis (MedIA) 2017.

[2] D Moyer et al. A Continuous Model of Cortical Connectivity, International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI) pp. 157-165. Springer International Publishing, 2016.