- The paper demonstrates an integrated CMUT-based tFUS system achieving sub-millimeter precision in BBB opening via phase-inversion processing.
- It validated the system with in vivo rat experiments, showing pressure-dependent acoustic emission enhancements and MRI-confirmed BBB permeability.
- The PI processing effectively suppressed device nonlinearities, offering up to 20 dB SNR improvement for enhanced feedback control.
CMUT-Based Transcranial Focused Ultrasound for Precise Blood-Brain Barrier Opening in Rodent Models
Overview and Motivation
The blood-brain barrier (BBB) presents a formidable obstacle for pharmacological interventions targeting neurological diseases. Conventional transcranial focused ultrasound (tFUS) systems, predominantly utilizing piezoelectric transducers, have faced limitations in terms of operational frequency bandwidth and spatial matching between transmit and receive components, complicating real-time feedback and monitoring of microbubble (MB)-assisted BBB opening. This paper introduces an integrated, all-CMUT (capacitive micromachined ultrasonic transducer) platform, engineered to enable both the delivery and sensing of ultrasound-mediated BBB opening with high spectral agility and spatial precision in small animal models. The platform leverages the inherent wide bandwidth and high sensitivity of CMUT arrays and implements phase-inversion (PI) processing for the effective suppression of device-induced nonlinearities, maximizing the signal-to-noise ratio (SNR) in the detection of MB acoustic emissions.
System Architecture and Characterization
The authors designed and fabricated a cylindrical CMUT array with five transmitting and one receiving elements, achieving a geometric focus suitable for targeting distinct regions within the rat brain. Acoustic field simulations and measurements validated focal integrity and spatial alignment, with -6 dB beamwidths consistent between modeled and hydrophone-mapped data, demonstrating sub-millimeter targeting accuracy.
To overcome the intrinsic electromechanical nonlinearity of CMUTs—which otherwise impairs detection of therapeutically relevant harmonics—the study adopted PI excitation and processing. Consecutive phase-inverted transmission bursts (0°/180°) allowed the cancellation of odd-order harmonics originating from the transducer, confirmed by a 58 dB fundamental reduction and a 36 dB third harmonic reduction under water tank conditions, supporting robust isolation of MB-induced emissions across broad frequencies.
Broadband detection capability was further demonstrated in vitro in tube phantoms with MBs, revealing significant spectral enhancement (up to 20 dB in subharmonic, 15 dB in first ultraharmonic, and 3 dB in second ultraharmonic) relative to baseline, and marked SNR increases post-PI signal processing (37 dB at fundamental, 13 dB at third harmonic), critical for feedback applications.
In Vivo Validation and Quantitative Results
Three rats underwent bilateral sonications at distinct pressure regimes (800 kPa and 400 kPa). Acoustic emission monitoring pre- and post-MB infusion revealed pressure-dependent nonlinear activity, with pronounced spectral markers indicating BBB opening:
- Subharmonic: 35 dB increase
- First ultraharmonic: 15 dB increase
- Second ultraharmonic: 12 dB increase
- Fundamental: 1 dB increase (conventional), up to 20 dB (PI processing)
- Third harmonic: 4 dB increase (conventional), up to 10 dB (PI processing, high pressure)
Temporal tracking of acoustic emission amplitudes highlighted distinct rise-fall patterns, especially in the fundamental and third harmonics under high-pressure sonication, elucidating MB kinetics during BBB opening. Notably, PI processing uncovered MB dynamic signatures at both pressure levels, which conventional processing failed to resolve, offering up to 20 dB enhancement in effective dynamic range.
MRI validation using T1-weighted and dynamic contrast-enhanced (DCE) modalities quantitatively confirmed spatially confined, pressure-dependent BBB permeability:
- Ktrans values: 0.0852 min⁻¹ (high-pressure), 0.0352 min⁻¹ (low-pressure), 0.01 min⁻¹ (control)
These metrics showed strong correlation with PI-processed acoustic emission peaks, establishing the clinical relevance of real-time acoustic feedback for precise BBB modulation.
Implications and Future Directions
The fully integrated CMUT-based platform demonstrates capabilities for high-fidelity, broadband detection of MB acoustic emissions and spatially localized BBB opening, validated quantitatively by MRI permeability mapping. The adoption of PI processing not only suppresses device-induced nonlinearities but also normalizes baseline variations, enabling robust, pressure-dependent differentiation of MB kinetics—particularly in the nonlinear fundamental component, which may serve as a new metric for feedback-controlled sonication.
Theoretical implications include advancing CMUT-centric tFUS systems toward closed-loop, real-time modulation of BBB permeability, with feedback derived directly from acoustic spectral features. Practically, the platform’s wide tuning range and scalable array architecture enable multi-site targeting and adaptive sonication strategies, potentially translatable to larger animal or human-scale models.
Remaining technical challenges involve co-designing low-noise, CMUT-optimized front-end electronics, exploiting full array channel counts for electronic beamforming and spatial filtering, and integrating real-time PI processing with advanced modulation schemes. Further investigation correlating MB acoustic emission spectra with safety/efficacy metrics (MRI and histology) in expanded cohorts is needed for clinical translation.
Conclusion
This study establishes a high-performance CMUT-based transcranial focused ultrasound platform, capable of both driving and passively monitoring MB-assisted BBB opening in small animal models. With broadband sensitivity and dynamic range enhanced by PI processing, the system achieves spatially localized, pressure-dependent BBB modulation validated with quantitative MRI. The findings underpin the transition toward real-time, closed-loop, frequency-agile ultrasound-mediated brain drug delivery, with significant implications for preclinical and future clinical applications in neurotherapeutics (2604.22666).
