Transcranial focused ultrasound precise neuromodulation: a review of focal size regulation, treatment efficiency and mechanisms

Neurological disorders impose a major global health burden and there is a need for effective, noninvasive interventions. Existing noninvasive neuromodulation methods such as TMS, tACS, and tDCS are limited by spatial resolution and stimulation depth. Transcranial focused ultrasound (tFUS) delivers mechanical energy through the skull and can target deep brain regions with millimeter-sized focal spots, offering high resolution, controllability, and noninvasiveness. Achieving precise neuromodulation requires improving spatial resolution and maximizing modulation efficiency at the focal spot while minimizing off-target effects. The review investigates three pillars for precision tFUS: metasurface-based approaches to reduce focal size and enhance focusing, optimization strategies (including parameters and microbubbles) to increase modulation efficiency within the target region, and mechanistic insights focusing on mechanosensitive ion channels as mediators of ultrasound neuromodulation.

The review synthesizes prior work on focal size regulation and modulation efficiency, and mechanisms of tFUS neuromodulation. For focusing, conventional methods include geometrical curvature, lenses, and phased arrays, where focal size depends on transducer aperture and frequency (e.g., a 3.4 cm aperture, 1.7 cm focal length transducer yields ~1 mm axial focus at 1 MHz; at 5 MHz axial focus ~0.25 mm, albeit with higher attenuation and thermal effects limiting transcranial use). Acoustic metasurfaces enable subwavelength focusing through refractive, reflective, and diffractive designs: solid phononic crystal lenses, gradient-index metasurfaces, multilateral metasurfaces for energy confinement, broadband reflective metasurfaces, metagratings and Fresnel zone plate (FZP) lenses achieving narrowed focal spots (e.g., FWHM ~0.65 λ; focal spot ~0.364 λ). Time-reversal with open-cavity structures and Soret-type FZP demonstrate robust subwavelength focusing over broad bandwidths; sparse broadband metalenses and metasurface ring arrays have achieved needle-like focuses. Airy-beam holographic metasurfaces further expand precise and flexible neuromodulation capabilities. Regarding modulation efficiency, microbubbles under low acoustic pressure undergo stable cavitation, generating scattered forces that amplify radiation force without tissue damage, thereby increasing stimulation success. Studies show US+microbubbles can activate neurons otherwise unresponsive to US alone (e.g., C. elegans TRP-4), enhance c-Fos expression in mouse motor cortex after tail vein injection, and enable targeted inhibition or stimulation using nanodroplets or gas vesicles; Piezo1-targeted microbubbles (PTMB) lower activation thresholds by binding to Piezo channels to increase calcium influx. However, intravenous delivery has low targetability, and direct brain injection requires craniotomy; shell modifications and molecular targeting are proposed to improve specificity. Parameter optimization (FF, PRF, DC, SD, AP) critically shapes excitatory/inhibitory outcomes. Studies report DC-dependent modulation of MEP amplitude (e.g., 10% DC inhibiting, 30% with no effect), frequency-dependent pressure requirements, AP and SD scaling of response success, and protocol-specific excitatory vs inhibitory regimes (e.g., higher DC with shorter SD favor activation; low DC with long SD favor inhibition). Mechanistically, tFUS can gate mechanosensitive ion channels: MscL (including mutants like G22S) responds across 0.5–30 MHz enabling low-intensity activation; TRP channels (TRPA1, TRPM2, TRPV1) can be gated via mechanical or thermal pathways but often require higher pressure; K2P channels (TREK-1/2, TRAAK) are activated via membrane tension under higher FF/AP; Piezo1 is highly sensitive and mediates US-induced neuronal activation, with knockouts reducing EMG responses and movement. Advances in AAV capsid engineering suggest future systemic, cell-specific gene delivery to enable targeted sonogenetics without craniotomy.

This article is a narrative review synthesizing published experimental and theoretical studies on transcranial focused ultrasound neuromodulation. The authors organize the evidence into three domains: (1) focal size regulation via conventional focusing (curved geometries, lenses, phased arrays) and acoustic metasurfaces (refractive, reflective, diffractive metalenses, FZP designs, metagratings, Airy-beam holography), summarizing reported focal spot sizes, bandwidths, and subwavelength achievements; (2) modulation efficiency enhancement through microbubble-assisted sonication and systematic parameter optimization (FF, PRF, DC, SD, AP), collating findings on excitatory and inhibitory regimes, success rates, and biomarker responses (e.g., c-Fos, EMG); and (3) mechanisms emphasizing mechanosensitive ion channels (MscL, TRP family, K2P channels, Piezo1/2), including expression systems (HEK cells, oocytes, rodent neurons), activation thresholds, frequency and pressure dependencies, and in vivo behavioral/electrophysiological outcomes. The review integrates quantitative examples (e.g., focal widths, duty cycles, pressures) and tabulates representative ultrasound parameters for different channels to guide precise neuromodulation design.

- Increasing ultrasound frequency or transducer aperture reduces focal size (e.g., axial focus ~1 mm at 1 MHz vs ~0.25 mm at 5 MHz for a 3.4 cm aperture, 1.7 cm focal length transducer), but higher frequencies increase attenuation, scattering, and thermal effects in transcranial settings. - Acoustic metasurfaces (refractive, reflective, diffractive) can achieve subwavelength focusing: FZP and metagrating-based lenses narrowed focal spots to ~0.364 λ and FWHM ~0.65 λ; broadband sparse metalenses and metasurface ring arrays produced needle-like focuses. - Microbubble-assisted stimulation enhances neuromodulation efficiency via stable cavitation and scattered forces, increasing success rates without tissue damage; targeted actuators (gas vesicles, Piezo1-targeted microbubbles) reduce activation thresholds and improve spatial specificity. - Parameter optimization shapes excitatory/inhibitory outcomes: DC and SD strongly modulate MEP amplitude (e.g., 10% DC inhibitory; 30% DC no effect); higher AP and longer SD increase motor response success; higher FF requires greater AP for equivalent effects; excitatory regimes favor higher DC (>30%) with shorter SD, while inhibitory regimes favor lower DC (<10%) with longer SD (>1 min). - Mechanosensitive ion channels mediate tFUS effects: MscL (including G22S, I92L) enables low-intensity activation across 0.5–30 MHz; TRP channels (TRPA1, TRPM2, TRPV1) respond to US via mechanical or thermal gating but often require higher pressures; K2P channels (TREK-1/2, TRAAK) are activated via membrane tension under higher FF/AP; Piezo1 is a key mediator for neuronal activation, with knockout reducing US-induced EMG and limb movements. - Sonogenetic targeting via AAV-mediated expression of mechanosensitive channels provides cell-type specificity; evolving AAV capsids may enable brain-wide transgene expression via intravenous delivery, improving safety and translatability. - Overall, combining metasurface-enabled subwavelength focusing, microbubble or targeted actuator strategies, and parameter tuning yields improved spatial precision and efficiency of tFUS neuromodulation.

The review addresses how focal size reduction, modulation efficiency enhancement, and mechanistic targeting together enable more precise tFUS neuromodulation. Metasurfaces offer controllable, compact, and potentially cost-effective means to realize subwavelength focusing, overcoming the resolution limits of conventional transducers without resorting to impractically high frequencies. Microbubble-assisted approaches and careful selection of ultrasound parameters concentrate neuromodulatory effects within the focal spot and adjust excitatory versus inhibitory outcomes, thereby improving functional specificity. Mechanistic insights implicating mechanosensitive ion channels, particularly Piezo1, provide biological targets to lower stimulus thresholds and tailor responses via sonogenetics. These advances collectively enhance the feasibility of precise, deep, and noninvasive neuromodulation for both experimental neuroscience and potential clinical therapies. Remaining challenges include transcranial attenuation at higher frequencies, safe and targeted delivery of actuators or genetic tools, and mitigating off-target pathways (e.g., auditory activation) while ensuring reproducible outcomes across species and paradigms.

Transcranial focused ultrasound offers unique advantages of noninvasiveness, high spatial resolution, and deep penetration. Precision can be substantially improved by: (1) reducing focal spot size using acoustic metasurfaces and advanced lens designs to achieve subwavelength focusing; (2) increasing modulation efficiency via microbubble-assisted actuation and optimized ultrasound parameters to maximize on-target effects; and (3) leveraging mechanosensitive ion channels for targeted neuromodulation through sonogenetics. Future work should integrate metasurface-based focusing with channel targeting and parameter regimes to create robust, safe, and clinically translatable protocols. Key directions include optimizing metasurface designs for transcranial use, developing molecularly targeted and circulatory-stable actuators, engineering more US-sensitive channel variants, and advancing noninvasive gene delivery (e.g., AAV capsids crossing the blood–brain barrier) to enable precise, cell-specific neuromodulation.

The review highlights several constraints: high-frequency ultrasound improves resolution but suffers from increased transcranial attenuation, scattering, and thermal effects, limiting practical use. Microbubble delivery via intravenous routes has low target specificity, while direct brain injection is invasive and clinically impractical. Many mechanosensitive channels (e.g., TREK/TRAAK) require higher frequencies or pressures, challenging safety and translation. Sonogenetic targeting currently relies on AAV vectors often delivered by stereotactic injection (craniotomy), raising safety concerns; although newer capsids may mitigate this, widespread clinical application remains uncertain. Off-target stimulation of non-targeted neurons and potential auditory pathway confounds necessitate careful parameter selection. Overall precision and safety require further refinement and validation in transcranial scenarios.