LORETA is a term meaning “Low Resolution Electromagnetic Tomographic Analysis” created by Roberto Pascual-Marqui during work over many decades. He has worked in Cuba, Switzerland (Key Institute) and is now working in Japan. He created LORETA, and others also produced an algorithm VARETA (Variable resolution electromagnetic tomography). Other versions have appeared, including swLORETA and eLORETA.
sLORETA stands for “standardized low-resolution electromagnetic tomographic analysis” and is an enhanced type of LORETA
The following is a comparison of the LORETA and sLORETA algorithms.
Both are designed by Roberto Pascual-Marqui
Both solve the “inverse” problem of EEG electrophysiology
Both use a set of scalp EEG measurements to produce estimates of voxel activity
Both use the concept of current-source-density (CSD) as the quantity estimated.
Both use a linear matrix transform of the scalp “field” to produce voxel estimates
Both use a 3-dimensional “dipole” model of each voxel
Both produce a statistical estimate of the probability that each voxel is contributing to the surface data
Both have published and online resources for developers who want to develop their own software.
LORETA has 2,000+ voxels. Each voxel is 7mm on a side.
sLORETA uses 6,339 voxels. Each voxel is 5mm on a side.
sLORETA incorporates additional assumptions regarding smoothing and weighting of the values, hence is “standardized”
sLORETA has “guaranteed accuracy” in presence of a single dipole. LORETA does not.
BrainMaster’s implementation of sLORETA is an entirely BrainMaster developed package, using the scientific information available regarding sLORETA. We do not use any software taken from LORETA or sLORETA in our packages. There is also a significant difference in performance in different systems. The original LORETA software ran slower than real-time, so that it would take more than a minute to process a minute of EEG. In the BrainMaster implementation of sLORETA, it is over 100X faster, so that the system can easily keep up with real-time. This allows the BrainMaster BrainAvatar to actually be a live animation sLORETA “projector.” BrainAvatar is able to reconstruct every one of the 6,239 voxels in real time, in full color, as raw CSD or as z-scores (BrainDX or qEEGPro). It thus provides a true live animation image of brain activation. Other implementations, being slower, cannot construct even the 2000+ voxels of LORETA in real time. They therefore us a “representative voxel” approach, computing only a chosen small set of voxels. Therefore, only BrainAvatar can create a true full image of all the voxels activation, in high resolution and in real time, for live neurofeedback or for animations for review and assessment.
For information and videos on the BrainAvatar implementation of sLORETA, see: www.brainavatar.com
Relevant links:
http://www.uzh.ch/keyinst/loreta.htm
http://www.uzh.ch/keyinst/NewLORETA/sLORETA/sLORETA.htm
