The Source Information Flow Toolbox (SIFT) is an GUI-enabled EEGLAB plugin for modeling and visualizing dynamical interactions between electrophysiological signals (EEG, ECoG, MEG, etc), preferably after transforming signals into the source domain. The toolbox consists of four modules: (1) Data Preprocessing, (2) Model Fitting and Connectivity Estimation, (3) Statistical Analysis, (4) Visualization, with a fifth Group Analysis module in development. Module 2 currently includes several adaptive multivariate autoregressive modeling (AMVAR) algorithms, including segmentation AMVAR and Kalman filtering. This subsequently allows the user to validate the model and estimate (in the time-frequency domain) a wide range of multivariate Granger-causal and coherence measures published to date. Module 3 includes routines for parametric and non-parametric significance testing. Module 4 contains routines for interactive visualization of dynamical interactions across time, frequency and anatomical source location.
The GNU General Public License; see: http://www.gnu.org/copyleft/gpl.html
The GPL requires that source code to a tool or resource be made available, and that modified versions of the tool or resource (including new tools which include part of the original tool's source) make the modified source available to users.
A developer of a tool or resource distributed under the GPL ensures that users of the tool and of modified versions of the tool may always inspect the source code and modify the source to suit their needs, and that users may modify the tool as they see fit without being obliged to distribute their changes. While this may seem to encourage multiple versions of the same tool, the result is often that changes are passed back to the tool originator and incorporated into the original tool.