|
jsmith
BIAC Alum
USA
6 Posts |
 Posted - Dec 15 2003 : 10:27:58 AM
|
Evan and I have developed a MATLAB script for use in obtaining the correlation between any number of user-specified ROIs. Essentially, the script accepts ROIs, finds the average percent signal change value of said ROIs at the peak of the hemodynamic response to each trial of user-specified conditions/bins, and obtains a correlation coefficient between any/all ROIs.
The script is also capable of subtracting the activity of a whole-brain ROI from each trial's activation, correcting for any whole brain, across trial effects that might create an artifical correlation, or lack thereof.
The script to run is in:
\\Gall\Programs\User_scripts\Smith\event_correlation.m
It requires two other scripts listing the subjects to be analyzed and the bins/conditions to be examined--examples of both scripts are located in the same folder.
This is our first effort at creating a general-use matlab script, so it is possible (probable?) that we have failed to anticipate some of the ways in which people have set up their data. If you would like to use it and the help function instructions (help event_correlation) aren't enough of a guide, or if the script crashes, please send an email with the error details to the address below and we'll try to patch it up.
jonathan.smith@duke.edu |
Edited by - jsmith on Dec 15 2003 4:07:37 PM |
|
Correlation of ROI-activity
