Transcranial random noise stimulation over early visual cortex improves processing of noisy visual stimuli

Transcranial random noise stimulation (tRNS) is a form of non-invasive electrical brain stimulation, which has shown promise at improving perceptual processing and learning. Here, we sought to gain insights into the brain mechanism underlying these benefits by using an equivalent noise paradigm along with the Perceptual Template Model (PTM). Subjects performed a two-alternative, forced choice, orientation discrimination of a centrally-presented Gabor patch (1° radius, 1 cycle/°, ±12° tilt). Random white noise was added to each Gabor at one of 8, evenly log-spaced increments of luminance contrast, while contrast thresholds were measured using a custom parameterized adaptive staircase that measured the entire threshold-versus-noise curve. Nine subjects completed 3 days of psychophysical testing that included an initial test day to allow subjects practice on the task and two subsequent days where tRNS and sham stimulation were counterbalanced. On stimulation days, four sets of psychophysical measurements were taken: before stimulation, during stimulation, and both 20 and 60 min post-stimulation. Stimulation was high frequency RNS (20 minutes, 2mA, bilateral occipital stimulation over O1 and O2 EEG locations). Compared to same-day baselines and sham stimulation, subjects showed benefits during stimulation that was specific to higher noise levels—an improvement linked by the PTM to improved external noise filtering. This was further confirmed by a grouped hierarchical Bayesian model, in which group hyper-priors for the PTM were compared for stimulation and sham groups. Of the three hyper-priors, only values for the coefficient representing the ability to filter external noise were non-overlapping at 95% highest posterior density (Sham = 2.358, 95% HPD = 2.292-2.388, Stim = 2.682, 95% HPD = 2.587-4.525). These results help account for both positive and negative outcomes of online stimulation, suggesting that tRNS may help boost signals among higher noise stimuli while offering little behavioral benefit at low noise and low contrast. Meeting abstract presented at VSS 2018