Neural Noise- Non-invasive focused ultrasound can modulate the human nucleus accumbens.1
- This FUS modulation produces measurable behavioral (reward sensitivity) readouts in humans.1
- The result is relevant for neuromodulation device development and therapeutic applications; Nature, human, tier-1.1 1
Weekly enrichment (2026-07-20)
- Published in Nature Communications (2025, Yaakub, Eraifej, Fouragnan et al.), the study tested whether non-invasive transcranial ultrasound stimulation (TUS) of the human nucleus accumbens (NAcc) causally alters reward-guided behaviour.23
- Design: within-subject, 26 healthy adults, four visits (screening/MRI plus three counterbalanced sessions ≥1 week apart) comparing TUS-NAcc, TUS-dACC (dorsal anterior cingulate control site) and Sham; fMRI recordings began ~10 min post-TUS during a probabilistic reversal-learning task (320 trials, reversal every 25 trials).2
- Device: bespoke CTX-500 NeuroFUS TPO system (Brainbox Ltd., Cardiff, UK) with a four-element annular transducer (64 mm diameter, 500 kHz central frequency, steering range 27.3-82.6 mm) specified to reach the deep NAcc.2
- Protocol: 80 s of 5 Hz repetitive TUS at 10% duty cycle (20 ms pulse duration, 200 ms pulse repetition interval, 400 total pulses); free-field spatial-peak pulse-average intensity (ISPPA) held constant at 35 W/cm².2
- Safety: exposures kept within ITRUSST guidelines; simulated maximum skull temperature rise generally <2 °C and CEM43 always <0.1 (well below the 0.25 threshold); Sham used bone-conduction sound and participants could not distinguish conditions.2
- TUS-NAcc increased reward-expectation BOLD in the NAcc versus Sham and TUS-dACC (ANOVA F₂,₇₅ = 7.15, p = 0.001; NAcc vs Sham Cohen’s d = 0.47), with whole-brain spread to adjacent striatum, thalamus, amygdala, precuneus and PCC.2
- Behaviourally, TUS-NAcc raised reward-linked learning rate (ANOVA F₂,₇₅ = 5.68, p = 0.005; vs Sham t₂₅ = -3, p = 0.006, d = 0.59) and increased win-stay strategy use and selection of the high-probability option—effects specific to reward rather than loss learning.2
- Target engagement was confirmed with invasive DBS-NAcc in 3 implanted patients: DBS ON reduced selection of the high-probability option and reward sensitivity (mixed-effects t₄₅₂ = -5.65, p = 2.7e-08), mirroring the TUS-driven features and supporting TUS as a non-invasive alternative to deep brain stimulation.2