Fig. 1: Study design, tDCS montages and E‑field distribution.
A, Study procedure. The experiment was conducted in a randomized, double-blind, sham-controlled parallel-group design. Patients were assigned to one of three tDCS groups: 2-mA conventional tDCS (n = 20), 4-mA optimized multichannel tDCS (n = 20), and sham tDCS (n = 20). All participants underwent 30 daily stimulation sessions. Depressive symptoms were evaluated before the intervention, at weeks 2 (sesson 10), 4 (session 20), 6 (session 30), and up to one and three months follow-up. Cognitive control performance and resting-EEG functional connectivity were assessed before and after the intervention. B, Distribution of the normal component of the E-field (En, in V/m) in a surface inside the grey matter tissue, between the grey matter-CSF and white matter-grey matter interfaces. The bottom panel shows the En-distribution for the conventional protocol using a bipolar electrode montage with rectangular 7×5 cm sponge electrodes (anode/cathode over F3/F4) at +2/-2-mA intensity (current density = 0.057 mA/cm²). The top panel illustrates the En-distribution for the optimized multichannel protocol using 7 small sponge electrodes (with a radius of 1.0 cm and a contact area of ≈3.14 cm2, anodes shown in red, cathodes in blue) targeting Brodmann area 46, intended to enhance and reduce excitability in the left and right dorsolateral prefrontal cortices, respectively. In the optimized multichannel stimulation condition, three anodes over the F3 (1.10 mA, current density ≈0.35 mA/cm²), AF3 (1.24 mA, current density ≈0.40 mA/cm²), and FC6 (1.65 mA, current density ≈0.53 mA/cm²) and four cathodes over the F4 (-1.28 mA, current density ≈0.41 mA/cm²), AF4 (-1.11 mA, current density ≈0.35 mA/cm²), F8 (-0.62 mA, current density ≈0.20 mA/cm²), and FC5 (-0.97 mA, current density ≈0.31 mA/cm²) were included for anodal and cathodal stimulation respectively. Positive En values indicate an inward-directed field, corresponding to increased cortical excitability (expected anodal effect), whereas negative En values indicate the opposite direction and reduced excitability (expected cathodal effect), according to the lambda-E model. Figure 1B shows that the target regions in the left and right prefrontal cortices are modulated more homogeneously in the optimized multichannel tDCS than in the conventional tDCS.
