← Explore

Low intensity focused ultrasound stimulation in stroke: A phase I safety & feasibility trial

Ziping Huang, Charalambos C. Charalambous, Mengyue Chen, Taewon Kim, Estate Sokhadze, Allen Song, Sin-Ho Jung, Shashank Shekhar, Jody A. Feld, Xiaoning Jiang, Wuwei Feng

Brain Stimulation 2025, 18, 179-187 · 10.1016/j.brs.2025.01.015

human patientstrokeemg mepbehaviourother mri

Abstract

Objective We aimed to determine the maximum safe spatial-peak pulse-average intensity (I SPPA ) of low-intensity focused ultrasound stimulation (LIFUS) in stroke patients and explore its effect on motor learning and corticospinal excitability. Methods We adopted the classic 3 + 3 design to escalate I SPPA (estimated in-vivo transcranial value) from 0, 1, 2, 4, 6, to 8 W/cm 2 . Stopping rules were pre-defined: 2 nd -degree scalp burn, clinical seizure, new lesion on diffusion-weighted imaging or major reduction in apparent diffusion coefficient, and participant discontinuation due to any reason. We applied 12-min LIFUS over the ipsilesional motor cortex while participants were concurrently practicing 3 blocks of a motor sequence learning (MSL) task using the affected hand. We measured MSL (response time) and corticospinal excitability (motor evoked potential) pre- and post-stimulation and compared MSL and corticospinal excitability between the LOW (0, 1, and 2 W/cm 2 ) and HIGH (4, 6, and 8 W/cm 2 ) groups. Results I SPPA was escalated to 8 W/cm 2 with 18 stroke participants without meeting the stopping rules. Compared to the LOW, more participants in the HIGH performed better on MSL (6/9 vs. 0/9, p = 0.009) and showed a sign of greater corticospinal excitability (7/9 vs. 5/9, p = 0.62). Interpretation Our phase-I safety study suggests that one session of LIFUS up to 8 W/cm 2 I SPPA is safe and feasible in stroke patients, and LIFUS at high intensity induces positive changes in both MSL and corticospinal excitability. The next logical step is to conduct a phase-II trial testing the efficacy of LIFUS and continuously monitoring its safety profiles.

Abstract via europepmc.

Specieshuman
Subjects18 participants
Sessions per subject1
Randomisednot reported
Blindingnot reported
Sham / controlinactive transducer
Auditory controlnot reported
Readout timingboth
Anaesthesianot applicable
Readoutsemg mep, behaviour, other mrimotor sequence learning (MSL) response-time task; TMS-elicited motor evoked potentials (MEP) from paretic APB; diffusion-weighted imaging (DWI) / apparent diffusion coefficient (ADC)
Direction of effectexcitatoryHigher-intensity LIFUS (HIGH: 4, 6, 8 W/cm2 ISPPA) produced greater motor sequence learning improvement (6/9 vs 0/9 participants improving >=20%, p=0.009) and a non-significant trend toward greater corticospinal excitability (7/9 vs 5/9, p=0.62) compared to LOW (0, 1, 2 W/cm2).
Adverse eventsobservedNo participant met the pre-defined stopping rules (2nd-degree scalp burn, seizure, new DWI lesion/major ADC reduction, discontinuation); one participant at 8 W/cm2 had a mild first-degree scalp burn with pain that resolved within 24 h without treatment, attributed to transducer heat and insufficient coupling gel.

Exposures

Exposure 1: 3+3 intensity-escalation LIFUS over the ipsilesional motor cortex during motor sequence learning

Target: motor cortex — “ipsilesional motor cortex
Device: Blatek · Blatek Industries, Inc. · AT32080

Pulse timing
Waveformpulsed
Fundamental frequency (kHz)500✓✓
Pulse duration (ms)0.2✓✓
Pulse repetition frequency (Hz)1,000✓✓
Duty cycle (%)20pulse duration × PRF gives 20%✓✓
Sonication duration (s)0.5✓✓
Pressure and intensity, by domain
Free-field pressure (kPa)not reported
Free-field Isppa (W/cm²)40✓✓
Free-field Ispta (W/cm²)not reported
In-situ estimatederatingsingle value
In-situ pressure (kPa)490✓✓
In-situ Isppa (W/cm²)0, 1, 2, 4, 6, 8swept✓✓
In-situ Ispta (W/cm²)0, 0.05, 0.1, 0.2, 0.3, 0.4swept✓✓
Protocol, in the paper’s words

We applied 12-min LIFUS over the ipsilesional motor cortex while participants were concurrently practicing 3 blocks of a motor sequence learning (MSL) task using the affected hand. Function generator (FG) A delivered 500 pulses at a 1000 Hz pulse repetition frequency (PRF) to trigger FG B, while FG B provided 100 cycles of a 500 kHz sine wave per pulse; the electrical waveform consists of a 200 us tone burst duration, a 500 ms sonication duration (SD), and a 1 s inter-stimulus interval. Intensity was escalated across participants using a 3+3 dose-escalation design (0, 1, 2, 4, 6, 8 W/cm2 ISPPA, estimated in-vivo transcranial value), with three participants enrolled per dose level.

Flags from extraction

  • sham_typeThe 0 W/cm2 dose level (transducer applied, zero acoustic output) functions as a sham but is not explicitly labelled 'sham' by the authors.
  • exposures[0].in_situ.methodThe estimated transcranial intensity is derived from a measured free-water intensity combined with a skull attenuation/transmission ratio informed by prior simulations; could be read as either 'derating' or 'simulation'.
  • exposures[0].in_situ.pressure_kpaThe 0.49 MPa pressure value in Table 2 is given only for the top of the escalation (matching the 8 W/cm2 ISPPA / 533 mW/cm2 ISPTA row); per-dose pressures for the lower intensity levels are not stated.
  • exposures[0].free_field.isppa_w_cm2The 40 W/cm2 free-water ISPPA is stated only as the setting used to produce the top (8 W/cm2) estimated transcranial dose; free-field values for the lower doses in the escalation are not stated.
  • randomisedDose level was assigned by a fixed 3+3 escalation scheme, not randomisation, but the paper never explicitly states whether allocation was randomised.