Fig. 1.

Experimental design and results. Participants performed a detection task in which they were asked to indicate whether a given trial was a stimulation or sham trial. The experiment was split into two conditions and was always administered in the same order starting with unmasked trials. a, Unmasked trials. In the unmasked condition, participants received TUS in half of the trials, while in the other half no TUS was applied. Each TUS trial consisted of a 300 ms burst, which was made up of 0.5 ms pulses of 500 kHz ultrasound (250 cycles) alternating with 0.5 ms of no ultrasound (that is, a burst duty cycle (BDC) of 50% and PRF of 1 kHz). The maximum modelled peak positive pressure at the target cortical site (V1) was 0.6 MPa b, Masked trials. The masked condition comprised the same TUS stimulation protocol but with the addition of an audio mask delivered via earphones to the participants. The audio mask was applied during TUS and sham trials in the masked condition only. The audio mask consisted of a 1 kHz square wave starting approximately 110 ms before TUS onset and lasting 700 ms. c, Behavioural results. Boxplots depicting d’ values for unmasked and masked trials. Black lines indicate the median d’ value and filled circles correspond to individual data points. Detection rates were significantly lower in the masked condition compared to the unmasked condition. Three participants continued to detect stimulation in the presence of the masking signal. d, Stimulation sites. Stimulation sites shown in MNI space. Target brain sites were identified visually. Sites were chosen in the right visual cortex that could be reached easily with TUS.