Distributions of current produced by transcranial direct current stimulation (tDCS) in humans were predicted by a finite-element model representing several individual and collective refinements over prior efforts. A model of the entire human head and brain was made using a finely meshed (1.1x1.1x1.4mm(3) voxel) tissue dataset derived from the MRI data set of a normal human brain. The conductivities of ten tissues were simulated (bone, scalp, blood, CSF, muscle, white matter, gray matter, sclera, fat, and cartilage). We then modeled the effect of placing a "stimulating" electrode with a saline-like conductivity over F3, and a similar "reference" electrode over a right supraorbital (RS) location, as well as the complements of these locations, to compare expectations derived from the simulation with experimental data also using these locations in terms of the presence or absence of subjective and objective effects. The sensitivity of the results to changes in conductivity values were examined by varying white matter conductivity over a factor of ten. Our simulations established that high current densities were found directly under the stimulating and reference electrodes, but values of the same order of magnitude occurred in other structures, and many areas of the brain that might be behaviorally active were also subjected to what may be substantial amounts of current. The modeling also suggests that more targeted stimulations might be achieved by different electrode topologies.
Copyright 2010 Elsevier Inc. All rights reserved.