Summed contrasts can conceal reversed regional preferences: directional and dose criteria for screening transcranial direct current stimulation montage pairs

Abstract

Background: The regional-matching design evaluated here requires two transcranial direct current stimulation (tDCS) montages, each stronger in its own target region, with adequate exposure. Summed contrasts can be positive when only one preference holds. Objective: To test whether a summed double-dissociation index (DDI), overlap and cohort-median dose rules identify such pairs, for a focused-attention versus open-monitoring experiment. Methods: Twenty-four montage pairs were simulated in 20 anatomical head models (SimNIBS 4.6, 2 mA). A four-rule screen (DDI, overlap, cohort-median dose) was recorded before cohort results; a post hoc conjunction required both regional contrasts positive in >=80% of heads. Two leading pairs were compared on per-head dose, a second solver and 60 conductivity stress-test draws. Results: Eighteen pairs passed the screen, but 13 failed the conjunction; one with positive median DDI (0.782) reversed the salience-region preference in all 20 heads. The nominated pair (F3-centered 4x1 ring; F8-Fp2 sponges; median DDI 1.561; both preferences in 20/20 heads) brought both anatomical targets to the 0.20 V/m operational floor in 11/20 heads; a conventional sponge pair (DDI 0.627) did so in 20/20, though both passed the cohort-median dose rule. The nominated pair's separation held across a second solver and all 60 stress-test draws, but both network targets reached the floor in 19/60. Conclusions: Summed indices and cohort-median dose rules can pass pairs that fail reciprocal-exposure criteria. Such screens should pre-specify the directional conjunction and report per-head joint dose coverage. These are modeled exposures, not physiological or behavioral effects.