Ultrasound Modulation of Visual Circuits in Mice Independent of Auditory Confound

Jiaru He, Jiejun Zhu, Xinxin Wang, Zihao Chen, Jin Yang, Zhen Yuan, Hongzhi Xu, Lei Sun, Zhihai Qiu

Advanced Science 2026, 13 · 10.1002/advs.202515991

rodenthealthycellular imaging

Abstract

Despite its growing potential as a noninvasive neuromodulation technology, the fundamental biophysical mechanism of focused ultrasound neuromodulation (FUN) remains elusive. Progress has been hampered by a persistent auditory confound, making it difficult to isolate the direct mechanical effects on neural circuits. To overcome this barrier, we employed two-photon calcium imaging (2PCI) in the primary visual cortex (V1) of deafened mice. Here, this study identifies a sparse (<1.5%) and spatially distributed population of neurons that is robustly and directly activated by ultrasound in a pressure-dependent manner. Crucially, this sparse, direct activation is sufficient to modulate the broader visual circuit, altering the response dynamics of neighboring neurons to visual stimuli. By unequivocally decoupling direct neuromodulation from auditory artifacts, the findings identify a key cellular substrate for FUN and establish a robust framework for future studies. This work paves the way for targeted investigations into the precise biophysical mechanisms governing the interaction between ultrasound and neurons.

Abstract via europepmc.

Speciesmouse

Subjects39 animals

Sessions per subjectnot reported

Randomisednot reported

Blindingnot reported

Sham / controlinactive transducer

Auditory controldeafened subjects

Readout timingboth

Anaesthesianot reported

Readoutscellular imaging — Two-photon calcium imaging (GCaMP6s) of V1 neurons through a cranial glass window; auditory brainstem response (ABR) testing used to confirm surgical deafening

Direction of effectbidirectionalUltrasound directly and pressure-dependently activated a sparse population of ultrasound-sensitive neurons (UNs; up to 1.144% of neurons at 450 kPa) in V1 of deafened mice, independent of audition. When combined with visual stimulation, ultrasound bidirectionally modulated visual-sensitive neurons (VNs): one VN subgroup was inhibited by ultrasound (VNs2, p<0.0001) while another was excited (VNs3, p<0.0001), and a third was unaffected (VNs1).

Adverse eventsnot reportedCalculated mechanical index (MI = 0.45) at the maximum negative pressure of 450 kPa (1 MHz) was below the FDA limit of 1.9; ISPPA (6.75 W/cm2) and ISPTA (55.33 mW/cm2) at this pressure were well below FDA safety thresholds (190 W/cm2 and 720 mW/cm2 respectively); simulated focal temperature rise after 0.25 s of continuous ultrasound action at 450 kPa was approximately 0.055 degrees C.

Exposures

Exposure 1: Ultrasound stimulation of primary visual cortex (V1) through cranial glass window

Target: primary visual cortex — “primary visual cortex (V1)”
Device: custom-built · custom annular ultrasound transducer integrated with 2-photon objective lens

Pulse timing
Waveform	pulsed
Fundamental frequency (kHz)	1,000	✓✓✓
Pulse duration (ms)	50	✓✓✓
Pulse repetition frequency (Hz)	2.5, 10, 50, 250, 500, 1,000, 2,000swept	✓✓✓⚑
Duty cycle (%)	50	✓✓✓
Sonication duration (s)	0.5	✓✓✓

Pressure and intensity, by domain
Free-field pressure (kPa)	not reported
Free-field Isppa (W/cm²)	not reported
Free-field Ispta (W/cm²)	not reported
In-situ estimate	measurementsingle value
In-situ pressure (kPa)	69, 178, 315, 450swept	✓✓✓
In-situ Isppa (W/cm²)	6.75	✓✓✓⚑
In-situ Ispta (W/cm²)	0.0553	✓✓✓

Protocol, in the paper’s words

Smoothed ultrasonic pulses (sine-windowed to minimise auditory artifacts) were delivered through a cranial glass window to V1. At PRF 10 Hz: pulse duration (PD) 50 ms, pulse repetition interval (PRI) 100 ms, ramp duration (RD) 25 ms, stimulation duration (SD) 0.5 s (5 pulses), stimulation interval (SI) 30 s; each trial comprised 10 ultrasound stimulations (first discarded, 9 analysed). Pressure was varied (0, 69, 178, 315, 450 kPa) at fixed PRF 10 Hz; PRF was separately varied (2.5-2000 Hz) at a fixed pressure of 315 kPa to test auditory confound across cell types.

Flags from extraction

exposures[0].in_situ — Pressures were measured via hydrophone through a glass cranial window (using ex vivo mouse skull plus glass) rather than in free water alone; classified as in_situ (measurement) since this represents the field reaching the neurons, though the paper does not use the term 'in situ' itself.
exposures[0].in_situ.isppa_w_cm2 — ISPPA (6.75 W/cm2) is only explicitly reported for the maximum 450 kPa condition, not for the other tested pressures (69, 178, 315 kPa).
exposures[0].timing.pulse_repetition_frequency_hz — The main pressure-sweep experiments used a fixed PRF of 10 Hz; a separate experiment swept PRF (2.5-2000 Hz) at a fixed pressure of 315 kPa. Both parameter sweeps are combined into one exposure per the one-exposure-per-target-and-frequency rule, recorded as a list.
n_sessions_per_subject — The paper describes 10 ultrasound stimulation trials per recording session but does not state whether individual mice underwent one or multiple separate imaging sessions/days across the various sub-experiments (pressure sweep, PRF sweep, hearing vs deaf comparisons).
n_subjects — 39 mice total were used across viral injection/glass window/deafening procedures; individual sub-experiments report smaller overlapping cohort sizes in figure legends (e.g. n=7, n=8 mice per group) that could not be reliably summed without double-counting.
anaesthesia — Anaesthesia (ketamine/xylazine) was used for surgical procedures (viral injection, glass window, deafening), but the paper does not explicitly state whether mice were awake or anaesthetised during the 2PCI/ultrasound stimulation sessions themselves.