Wearable ultrasound for continuous, noninvasive deep-tissue imaging is crucial for early disease detection and monitoring. Fast 3D volumetric imaging offers superior accuracy and reproducibility compared to 2D methods, particularly for complex structures like the cardiovascular system. However, current technologies struggle to combine continuous monitoring with fast volumetric acquisition. Conventional piezoelectric transducers are bulky and require high driving voltages (above 25V), limiting their use in wearable devices. The impedance mismatch between transducer materials and tissue causes significant acoustic energy loss. While piezoelectric composites improve electromechanical coupling, they still demand high driving voltages. MEMS-based MUTs (cMUTs and pMUTs) offer miniaturization and ease of array fabrication. cMUTs have advantages in bandwidth and coupling but require high DC bias voltages and exhibit nonlinear transduction, reducing acoustic power. pMUTs, operating in flexural vibration mode, require lower voltages, making them more suitable for wearable devices. However, driving high-density pMUT arrays efficiently is challenging due to coupling effects (crosstalk) between cells. Existing equivalent circuit (EQC) models for pMUT arrays account for acoustic coupling but lack extensive experimental validation for medical imaging. Grouping pMUT cells into elements reduces drive channels, enhancing scalability and enabling low-power design, but lacks comprehensive analytical models considering both intra-element and inter-element acoustic crosstalk. This study addresses these limitations by introducing a novel hierarchical EQC model for 2D phased-array pMUT transducers to analyze performance and crosstalk and experimentally validating it using a low-voltage-driven array for fast 3D volumetric imaging.

Existing ultrasound methods for deep tissue imaging typically offer either continuous monitoring or fast volumetric imaging, but not both. Conventional piezoelectric array transducers are unsuitable for continuous monitoring due to bulkiness and fabrication challenges. Impedance mismatch at interfaces leads to significant acoustic energy loss, requiring the addition of impedance matching layers. While piezoelectric composites improve electromechanical coupling, they still require high driving voltages (25V or more) to achieve sufficient penetration depth. MEMS-based micromachined ultrasonic transducers (MUTs) offer potential advantages, including capacitive micromachined ultrasound transducers (cMUTs) and piezoelectric micromachined ultrasound transducers (pMUTs). cMUTs have large bandwidths and high electromechanical coupling but require high DC bias voltages and exhibit non-linear transduction which limits acoustic output. pMUTs do not require high bias voltages and offer better acoustic coupling efficiency making them suitable for long term wearable application. Previous research on grouping PMUT cells into elements for excitation has been reported, primarily for applications other than medical imaging. While models exist for analyzing pMUT array performance, these often lack comprehensive experimental validation for 3D imaging applications at low voltages. No prior work has reported full volumetric imaging using low-voltage pMUT arrays for medical imaging applications, particularly targeting deep tissues.

This study designs, fabricates, and characterizes a 2D phased-array pMUT transducer driven by a low voltage (5V) for fast 3D volumetric imaging. The bilayer diaphragm pMUT cell operates in the d31 flexural-vibration mode. The design incorporates a ceramic lead zirconate titanate (PZT) layer sandwiched between two thin electrodes, with the top electrode optimized to maximize vibration energy. A novel multilevel cell-element-array equivalent circuit (EQC) model is developed to quantitatively characterize the electrical, mechanical, and acoustic interactions at the cell, element, and array levels. The model accounts for acoustic coupling (crosstalk) between cells within an element and between elements in the array using self-radiation and mutual radiation impedance calculations. The model utilizes equations derived from the Helmholtz equation for membrane vibration and considers the mutual radiation impedance as a function of effective radius, distance between cells, and wavenumber. Fabrication involves multilayer processing on a silicon-on-insulator (SOI) wafer, including PZT deposition, electrode patterning, and Si etching. The fabricated pMUT array consists of 8x8 elements, each comprising a 4x4 array of cells, driven individually using 5V pulses with programmable phases for beamforming. Acoustic characterization involves transmission, receiving, and pulse-echo measurements using a hydrophone in a water tank. Volumetric imaging experiments are performed using wire and vascular phantoms to evaluate penetration depth and temporal performance.

The proposed cell-element-array design and EQC model accurately predict the performance of the pMUT array, accounting for acoustic coupling effects. The low-voltage (5V) driven pMUT array successfully performs fast 3D volumetric imaging in wire and vascular phantoms. The array demonstrates a high temporal frame rate of 11 kHz and covers a volumetric range of 40 mm × 40 mm × 70 mm. The experimental results validate the EQC model's predictions of both intra-element and inter-element acoustic coupling, demonstrating the effectiveness of the cell-element-array design for low-voltage operation. High-quality 3D images are obtained, demonstrating the feasibility of using this technology for deep-tissue imaging. The study shows that grouping pMUT cells into elements significantly reduces the number of drive channels without compromising imaging performance. The achieved frame rate and imaging depth are significantly improved compared to prior pMUT array implementations and existing low-voltage solutions, proving the efficacy of the approach for wearable ultrasound imaging.

The successful demonstration of fast 3D volumetric imaging using a low-voltage-driven MEMS pMUT array addresses a significant challenge in wearable ultrasound technology. The novel cell-element-array design and the accompanying EQC model provide a powerful tool for designing and optimizing such arrays for various applications. The high temporal frame rate and substantial imaging depth achieved show the potential for real-time monitoring of dynamic processes in deep tissues. The low-voltage operation is particularly advantageous for wearable applications, addressing the power consumption and safety concerns associated with high-voltage transducers. The quantitative agreement between the experimental results and the EQC model predictions confirms the accuracy and usefulness of the model for future pMUT array designs. This approach paves the way for developing highly integrated and energy-efficient wearable ultrasound devices for a wide range of clinical applications.

This paper presents a novel low-voltage-driven MEMS ultrasonic phased-array transducer for fast 3D volumetric imaging. The design utilizes a cell-element-array architecture to minimize the number of driving channels while achieving high-quality 3D images with a 5V drive voltage. The experimental results validate the proposed equivalent circuit model and demonstrate the feasibility of this technology for wearable applications. Future work could focus on further miniaturization of the transducer, integration with advanced signal processing techniques for improved image quality, and in vivo studies to evaluate the system's performance in real clinical settings. Exploration of different materials and cell geometries could potentially further enhance the transducer's performance.

The current study was conducted using phantoms, and further in vivo studies are needed to fully evaluate the transducer's performance in real-world clinical settings. The EQC model makes certain assumptions about the acoustic environment and the behavior of the pMUT cells, which might not perfectly reflect the complexities of in vivo tissue. The size of the imaging area could be further improved by increasing the number of elements in the array. Finally, long-term stability and biocompatibility of the device need to be assessed before clinical translation.