An electronically steerable epidural ultrasound interface for deep brain neuromodulation in freely moving rats

L Ratz, H Rivandi, GK Wardhana, E Sarkar, M Aqamolaei, Y Tong, S Desmarais, L Sohail, PD Donaldson, V Aloia, M Arlotti, VA Coenen, DG Muratore, GD Spyropoulos, TLM da Costa, MD Döbrössy

2026 · 10.64898/2026.09.10.750612

rodenthealthyhistology molecularother

Abstract

Low-intensity focused ultrasound (LI-FUS) clinical trials exploit either the neural activity modulating, or the blood-brain barrier opening, capacity of this stimulation modality. However, LI-FUS currently is applied only transcranially which means that it is conducive only for episodic and intermittent stimulation, although, clinical data shows that in numerous neurological and psychiatric applications, chronic and continuous stimulation is required for long-term, stable therapeutic effect. The paper presents experimental data on a novel and innovative approach describing an implantable, epidural focus ultrasound (eFUS) device, designed for continuous, chronic and multi-site steerable neuromodulation. The miniaturized eFUS device consists of a two-dimensional piezoelectric transducer array directly integrated onto a custom ASIC, specifically engineered for proof-of-principle neuromodulation studies in the rat brain. The system generates electronically steerable focused ultrasound with software-defined focal coordinates, sufficient to stimulate neuronal activity in deep brain structures. In vitro acoustic characterization confirmed accurate beam steering and focusing, while in vivo validation demonstrated reliable stimulation of a deep subcortical target with measurable physiological effects. eFUS-mediated targeting of the ventral tegmental area in awake and freely moving rats produced increase in dopamine release in the nucleus accumbens as confirmed using fiber photometry recordings. Post-mortem histological analysis of the target regions showed the absence of inflammatory markers, although the epidural placement of the eFUS device was associated with mild tissue damage. Overall, the study provides in vitro data demonstrating the energy efficiency, and steerability of the technology, and in vivo physiological evidence of neuromodulatory ability of a deep, subcortical brain structure.

Abstract via europepmc.

Speciesrat (Sprague Dawley)

Subjects4 animals

Sessions per subjectnot reported

Randomisednot reported

Blindingnot reported

Sham / controlinactive transducer

Auditory controlnone

Readout timingboth

Anaesthesiaawake

Readoutshistology molecular, other — fiber photometry of GRABDA2m dopamine sensor in nucleus accumbens; ED1 and GFAP immunohistochemistry; Evans Blue BBB opening for targeting

Direction of effectexcitatoryStimulation of the VTA at 10% duty cycle and 2.5 s sonication produced a significant post-stimulation increase in nucleus accumbens dopamine versus sham; 2.5% and 5% duty cycles produced no significant increase.

Adverse eventsobservedSuperficial cortical damage from the skull screws fixing the chip was seen; no inflammation or tissue damage at the VTA target; temperature rise of the device below 1 C at 10% duty cycle (about 6 C at 50%) in bench testing.

Exposures

Exposure 1: epidural eFUS to VTA (parameter sets p1-p5 analysed in detail)

Target: ventral tegmental area — “the left VTA”
Device: custom-built · eFUS device (CMOS ASIC with integrated 2D PZT array) ✓

Pulse timing
Waveform	pulsed
Fundamental frequency (kHz)	4,000	✓✓✓⚑
Pulse duration (ms)	0.025, 0.05, 0.1, 0.2, 0.5swept	✓✓✓⚑
Pulse repetition frequency (Hz)	1,000	✓✓✓⚑
Duty cycle (%)	2.5, 5, 10, 20, 50swept	✓✓✓⚑
Sonication duration (s)	10, 5, 2.5, 1, 0.5swept	✓✓✓

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	deratingmean or range across subjects
In-situ pressure (kPa)	260, 574swept	✓✓✓⚑
In-situ Isppa (W/cm²)	not reported
In-situ Ispta (W/cm²)	not reported

Protocol, in the paper’s words

Twelve parameter sets were tested; five (p1 to p5) analysed in detail all at PRF 1 kHz with DC 2.5, 5, 10, 20, 50% and SD 10, 5, 2.5, 1, 0.5 s. Each trial had a 30 s baseline, a sonication, and an inter-stimulus interval of at least 45 s, repeated 3 times per site.

Flags from extraction

n_subjects — Four rats stated; one was excluded for misalignment and two further animals were used separately for BBB-opening targeting checks
exposures[0].in_situ.pressure_kpa — 0.26 to 0.574 MPa is stated for all in vivo sets; intro says peak focal pressures were derived assuming 0.8 dB/cm/MHz brain attenuation from water measurements, so the domain is inferred as derated
exposures[0].free_field — Water-tank pressures are shown only as derated estimates or in figures; no free-field value for the in vivo exposures is given in the main text
exposures[0].timing.pulse_duration_ms — Pulse width '25 µs to 80.000 µs' (European decimal separator, taken as 80 ms) is the range across all twelve sets, not only p1-p5
exposures[0].timing.pulse_repetition_frequency_hz — PRF 0.0025 to 1.5 kHz across twelve sets; p1-p5 all at 1 kHz
exposures[0].timing.duty_cycle_pct — Range 2.5 to 50% across all sets
exposures[0].fundamental_frequency_khz — Designed 4 MHz; individual devices peaked between 3.7 and 4.0 MHz
auditory_control — Paper states there was no correction for a possible auditory confounder