Facilitates fast correlation computation for pairwise internodal functional connectivity estimation based on fMRI data. So far, two measures of association are supported: Pearson's r and the tetrachoric correlation coefficient. With Pearson's r, efficient implementation (based on CPU instruction set extensions) and parallelization yield speedups up to ~3x compared to Matlab's corrcoef (R2011b). With the tetrachoric correlation coefficient, data reduction in the temporal domain based on binarization of nodal time series combined with tetrachoric correlation estimation and efficient implementation (based on a 16-bit lookup table or CPU instruction set extensions) and parallelization yield speedups up to ~20x compared to Matlab's corrcoef (R2011b).
- 64-bit Linux
- 64-bit Matlab
- The GCC C compiler
- A working MEX setup that uses GCC to compile the MEX files
Obtain an archive containing the lastest version from http://www.kristianloewe.com or clone the repository using
$ git clone --recursive https://github.com/kloewe/corr-m.git
Change to the root directory of the extracted archive (or the cloned repository) and install corr-m and cpuinfo-m.
$ ./install-m.sh corr-m <target-dir-corr-m>
$ ./install-m.sh cpuinfo-m <target-dir-cpuinfo-m>
It is assumed here that Matlab is on your path and that MEX is set up.
Next, start Matlab and add the relevant directories to its path.
>> addpath <target-dir-corr-m>
>> addpath <target-dir-cpuinfo-m>
A description of each function can be displayed in Matlab using
>> help <function-name>
If you run into problems, please send an email to kl@kristianloewe.com.
If you use this program, please cite:
Loewe K, Grueschow M, Stoppel C, Kruse R, and Borgelt C (2014).
Fast construction of voxel-level functional connectivity graphs.
BMC Neuroscience 15:78.
doi