Electrical neuromodulation is an evolving therapeutic approach used to treat neurological conditions such as Parkinson’s disease, epilepsy, and vision loss. Early systems, such as cardiac pacemakers and deep brain stimulators, typically utilized low-channel-count stimulation. Recent technological progress has enabled large-scale multichannel systems supporting hundreds or thousands of electrodes.

As channel counts increase, power consumption becomes a critical constraint for the scalability of implantable neurostimulators. While small systems often rely on implanted batteries, the substantial power demands of large-scale systems make battery-powered operation impractical. Wireless power transfer (e.g. inductive coupling) offers an alternative, but is fundamentally limited in the amount of power that can be safely delivered. Consequently, optimizing energy efficiency in large-scale multichannel neurostimulators is essential for maximizing channel count within the available power budget.

While prior studies evaluated pulse shaping mainly from a neurophysiological perspective, this thesis is the first to systematically analyze the relationship between pulse shape, physiological effectiveness, and circuit-level power consumption to identify optimal stimulation strategies. The results challenge existing perspectives by demonstrating that rectangular pulses lead to fewer circuit-level losses, making them competitive compared to non-rectangular alternatives. Although non-rectangular pulses can reduce neural activation thresholds, when circuit losses are included, they require 14-51% more energy than rectangular pulses. This suggests that rectangular pulses may be preferable for practical neurostimulator implementations.

A second contribution is the introduction of a quantitative framework to capture the impact of channel-to-channel variability on power efficiency. Due to inherent variations in electrode impedance and required stimulation amplitudes, individual channels have different power requirements. Conventional power management techniques often neglect this variability, resulting in low energy efficiency. Although several strategies have been proposed to enhance efficiency, a quantification of their efficacy is lacking in the literature. This thesis introduces a systematic methodology to analyze power losses in multichannel neurostimulation systems, enabling consistent benchmarking of existing strategies and providing a foundation for new application-specific approaches. Applying this methodology to previously published experimental data demonstrates that the effectiveness of power management strategies varies across applications, underscoring the necessity for application-specific optimization.

Building on these insights, the thesis proposes an advanced power-management approach designed specifically for the varying power needs of individual stimulation channels. This strategy incorporates a channel-specific regulating rectifier optimized for current-mode stimulation, capable of dynamically adjusting its output voltage without compliance monitoring. The rectifier quickly adapts to changing load conditions, enabling efficient time-division multiplexing and improved scalability in multichannel neurostimulation systems. It achieves a median efficiency of 84% on a dataset of intracortical visual stimulation, representing a 74% improvement over conventional fixed supplies and 6% compared to an 8-rail stepped supply.

In conclusion, this thesis offers critical insights into enhancing energy efficiency in multichannel electrical stimulation, contributing to advancing next-generation large-scale neurostimulator technologies. A key observation is the interdependence of multiple system-level factors, emphasizing the importance of a holistic optimization approach. Additionally, the findings highlight that optimal pulse width for minimizing activation energy varies significantly with pulse shape, underscoring the necessity of co-optimizing stimulation parameters for both physiological effectiveness and energy efficiency. The methods developed provide new perspectives on energy-efficient stimulator designs, and the proposed power-management approach shows promising results for efficient channel-specific voltage regulation and reduced output losses.

Ultrasound (US) technology has emerged as a powerful modality in both medical imaging and therapy, offering non-invasive, real-time, and high-resolution capabilities. In neuromodulation, particularly for vagus nerve stimulation (VNS), US enables precise anatomical targeting and functional stimulation without surgical implantation of electrodes, reducing procedural risks and improving patient accessibility. Conventional dual-mode systems employ separate ultrasound transducers for imaging and therapy, each mechanically configured during fabrication for a fixed quality factor (Q-factor) through the presence or absence of a backing layer. This approach increases system cost, physical footprint, power consumption, and integration complexity while preventing seamless real-time switching between modes.
This thesis presents a novel electronically configurable Q-factor control architecture for 2D phased-array piezoelectric transducers, implemented in 180 nm CMOS technology, enabling dynamic switching between imaging and therapeutic modes with a single transducer array. The proposed method employs an active damping compensation technique to electronically reduce the Q-factor for imaging mode while preserving the high-Q state for therapy, eliminating the need for mechanical reconfiguration. The architecture integrates a mixed-signal signal chain comprising a fully differential operational transconductance amplifier, an 8-bit SAR ADC, an 8-bit DAC with programmable phase delays, and a high-voltage linear
amplifier, supporting phased-array beamforming for imaging mode of operation. System- and circuitlevel simulations using a PZT-5A air-backed transducer model demonstrate a reduction in the Q factor from 78.3 to 8.08, corresponding to a damping efficiency (FOMQ) of 9.7, with fewer than 1% variation between the process-voltage-temperature corners. These results place the Q-factor achieved in the imaging mode within the optimal range for high-resolution ultrasound imaging, while maintaining high-Q performance for therapy. The proposed architecture offers a compact, cost-effective, and reconfigurable solution for advanced ultrasound platforms in neuromodulation, including real-time image-guided VNS, as well as other biomedical imaging and therapeutic applications.

Focused ultrasound vagus nerve stimulation (FUVNS) promises neurostimulation without implants, yet practical deployment is constrained by the need for accurate, real-time nerve localization and the cost, power, and operator burden of conventional imaging pipelines. This thesis proposes a complete hardware-software co-design for direct, on-chip localization tailored to FUVNS.

On the algorithmic side, we construct a task-specific ultrasound RF corpus spanning simulation phantoms and in-vivo acquisitions, and train a lightweight convolutional regressor that estimates nerve position directly from raw RF streams, bypassing B-mode reconstruction. Inter-channel correlation analysis guides an aggressive receive-element reduction, while hardware-aware quantization and magnitude pruning align model precision and sparsity with analog compute constraints, preserving inference robustness under stringent I/O and memory budgets.

On the hardware side, we design and simulate an analog convolutional layer based on charge-domain MAC operations, accepting analog activations from the ultrasound front-end and executing energy-efficient multiply–accumulate with compact passive structures. System- and block-level evaluations validate functional compatibility with the network’s dataflow and demonstrate substantial reductions in downstream bandwidth and front-end power compared with imaging-centric workflows.

Together, these contributions deliver a proof-of-concept pathway from RF acquisition to neurostimulation control within a single, application-specific edge-AI stack. The results indicate that direct-RF inference coupled with analog in-memory computation can enable compact, low-power vagus nerve localization systems.

Neurostimulation has emerged as a transformative approach for the treatment and management of a broad range of neurological disorders, including depression and epilepsy. The increase in usage of neurostimulation also necessitates high spatial resolution and depth of penetration in order to inflict minimal damage to the surrounding tissue. Optical or light-based stimulation offers unique benefits in this regard, enabling highly spatially resolved neural activation and paving the way for innovative modalities in targeted neural modulation.
In this work, an electronic platform for wireless powering and data transmission is presented,
utilising organic P–N junctions as the core technology for mediating photo-electric stimulation. These junctions are selected for their capacity to provide non-genetic neural activation with excellent spatial resolution. Power delivery to the implantable device is achieved acoustically via ultrasound, taking advantage of ultrasound’s superior tissue penetration characteristics.
Comprehensive characterisation of the organic P–N junctions was performed, culminating in
the identification of the PDCBT/ITIC architecture as the most suitable P-N junction, based on its
favourable optical absorption profile and photocurrent generation capability. In parallel, a dedicated power management and stimulation platform was developed as the initial steps in incorporating acoustic energy harvesting and efficient signal demodulation. Validation experiments confirmed reliable device performance and established a functional basis for wireless optoelectronic neurostimulation.
The results of this thesis establish a promising system-level paradigm for minimally invasive,
wirelessly powered neurostimulation using organic photo-sensitive interfaces and acoustic power links. This approach contributes a viable pathway toward the development of next-generation neural therapies with enhanced implantability.

Ultrasound neuromodulation is a rapidly developing, minimally invasive technique for brain stimulation. It holds great promise for treating neurological disorders, psychiatric conditions, and enabling brain– computer interfacing. A key challenge is achieving fine-grained, dynamic control over the acoustic focus within tissue. This requires focusing mechanisms that are both electronically tunable and scalable, in order to support precise and addressable focal regions. Current beamforming strategies based on phased arrays face limitations in spatial resolution and power efficiency.

This thesis explores a fundamentally different approach: shaping ultrasound fields through magnetic microparticle-based acoustic holography. These systems can be fabricated at the microscale and offer potential for control with high spatiotemporal precision. To assess the feasibility of this concept, this work presents the design, fabrication, and acoustic characterization of a planar, soft-lithography-based Fresnel Zone Plate (FZP). The FZP incorporates integrated microfluidic channels and is developed to manipulate ultrasound fields using magnetic microparticle suspensions.

The fabricated FZP consists of alternating concentric rings of PDMS and fluid-filled microchannels, where the acoustic properties of the microfluidic zones can be tuned by varying particle types and concentrations. Multiple layouts and methods were evaluated to enable reliable injection and exchange of different media into the sub-millimeter-thick microfluidic device. The finalized design was successfully replicated via soft lithography, and experimental measurements were conducted using five media: air, deionized water, and suspensions of three types of magnetic microparticles. Acoustic measurements revealed that the microparticle suspensions exhibited reduced transparency compared to water, but greater than that of air. Pressure amplitude decreased with increasing particle concentration, consistent with enhanced acoustic attenuation. Phase analysis showed a non-monotonic relationship between concentration and the effective speed of sound, deviating from trends reported in bulk suspensions. This highlights the complex interplay between confinement, concentration, particle settling, and wave propagation in microfluidic geometries.

The results demonstrate that magnetic microparticles enable simultaneous modulation of both the phase and amplitude of ultrasound. Combined with earlier demonstrations of digitally controlled spatial distribution using microfabricated current lines, this approach offers a promising route toward dynamic acoustic field steering. As such, it marks a step toward programmable acoustic holography at the microscale.

This thesis investigates a steady-state visually evoked potential (SSVEP)–based brain-computer interface (BCI) for controlling a computer cursor. The system employs two frequency-detection methods: filter-bank canonical correlation analysis (FBCCA) and ensemble task-related component analysis (eTRCA). Performance was evaluated using EEG data from a public dataset (MAMEM) and a self-recorded dataset. Classification accuracy, information transfer rate (ITR), and signal-to-noise ratio (SNR) were analyzed.

Results indicate that the MAMEM dataset, featuring a 256-channel high-density electrode setup, yielded superior classification accuracy—averaging 70.3% with FBCCA and reaching up to 93.3% peak accuracy in some subjects. In contrast, the self-recorded dataset, acquired with an 8-channel OpenBCI headset, showed lower performance, with FBCCA average accuracies ranging from 33.3% to 71.1%, and a maximum observed peak accuracy of 93.3% in isolated trials. FBCCA proved particularly effective in real-time conditions due to its calibration-free design, achieving real-time classification accuracies of 67.86% and 62.5% in two separate test sessions. In addition, system latency was assessed, and it was found that the complete signal processing pipeline executes within 0.232 seconds. eTRCA achieved higher offline accuracy on the MAMEM data when sufficient training trials were available but underperformed on the self-recorded dataset due to limited calibration data. Real-time cursor control was successfully demonstrated using two target frequencies (8.57 Hz and 12.00 Hz), confirming the practical viability of SSVEP-based BCIs. Future work should aim to improve signal quality, enhance spatial resolution through better electrode placement, and reduce the need for user-specific calibration to enable more reliable and accessible real-world applications.

This project explores the development of a real-time Brain-Computer Interface (BCI) system based on Steady-State Visually Evoked Potentials (SSVEPs). The aim is to enable hands-free computer control by detecting brain responses to visual stimuli flickering at specific frequencies. Our focus was on the acquisition of clean and stable EEG signals using the Unicorn Hybrid Black headset. Key challenges included minimizing noise from motion and eye artifacts, selecting the right channels for SSVEP detection, and applying appropriate filters. While the hardware setup and real-time streaming via Lab Streaming Layer (LSL) were successfully established, consistent frequency detection during visual stimulus trials remains an unresolved issue. Controlled tests with simulated signals confirmed that the pipeline can detect known inputs, indicating that future work should focus on improving artifact removal and stimulus reliability. The current system provides a solid foundation for further development toward robust, real-time brain-controlled interfaces

This thesis presents the design and implementation of a graphical user interface (GUI) for a real-time brain-computer interface (BCI) system based on steady-state visual evoked potentials (SSVEP). Developed as part of a larger collaborative project alongside Data Acquisition and Signal Processing subgroups, the GUI enables users to control a computer cursor using brain activity. Core functionalities include real-time EEG visualization, configurable data recording trials, visual stimulus presentation, and a cursor control interface that provides feedback based on live classification results. The system architecture allows the GUI to connect to the EEG headset via the Lab Streaming Layer (LSL) protocol, stream EEG data to the back-end pipeline, and translate classification output into interactive visual feedback. The integrated system successfully demonstrated real-time SSVEP-based control, validating both functional and technical requirements of the GUI subsystem. This work contributes to the development of more intuitive and accessible GUI designs for future BCI applications.

This master thesis investigates the potential of vector field stimulation as an innovative technique for enhancing visual cortex stimulation [1]. An algorithm was designed and implemented to evaluate this concept, utilizing COMSOL simulations to model the generated electric fields. The algorithm was validated using Green functions, which serve as analytical approximations to the Poisson equation solved in COMSOL. The interior point method was employed in the algorithm to minimize visual error [2], a concept introduced in the literature and adjusted to our problem. Results indicate that vector field stimulation improves targeted stimulation precision noticeably compared to traditional bipolar stimulation methods. Additionally, it was observed that incorporating the minimization of input current alongside visual error further enhanced the algorithm’s performance. Despite challenges such as numerical errors and the algorithm getting trapped in local minima, the findings highlight the need for future research to improve microstimulation techniques. This report emphasises the promising potential of vector field stimulation and the supporting algorithm in improving visual cortex stimulation.

In recent years, many therapeutic applications for medical ultrasound have arisen, such as ablation of tumors, breaking up kidney stones, and neuromodulation. To localize the target area, often ultrasound imaging methods are used. Conventionally, a different transducer array is used for imaging and therapeutic applications because of their conflicting requirements of transducer damping. If this damping would be done electronically, it can be switched on or off on demand, allowing a single transducer array to be used for both therapeutic and imaging ultrasound. This would allow medical ultrasound devices to be smaller and therapeutic treatments to have reduced overall system complexity. This thesis develops a PCB to compare a combination of existing and novel methods of electronic damping on a single transducer, also called electronic Q-factor control. A maximum damping of 23\% is achieved relative to the undamped transducer mounted on the PCB. This is achieved by a novel method based on a feedback loop around a second 'dummy' transducer. Methods of electronic Q-factor control should eventually be extended to be applied in an Integrated Circuit for damping of entire ultrasound transducer arrays.

The development of neural prostheses, especially those directly targeting the brain, requires extensive research and modelling before clinical trials can be performed. Currently, the resolution of artificial vision is not sufficient for everyday tasks. By studying the expected spatial extent of stimulation, we aim to provide insights to researchers that can be used to improve the resolution of artificial vision. The goal of this MSc thesis was to visualise the shape and intensity of electric fields in the cortex as a response to intracortical microelectrode stimulation to observe the expected regions of neuronal activation considering electrode design parameters. To do this, parameters that can influence the generated electric field and the regions of activated tissue have been defined. Implementing these parameters in a Finite Element Model (FEM) allows the computation of the generated electric field in 3D of a stimulating electrode to observe the spatial extent of activated tissue. The spatial extent of activated tissue can be estimated using simplified methods such as the activating function (AF) or current density threshold. The result is a parameterised framework that creates a Visualisation of Neuronal Activation (VoNA) that can be used to assess activated tissue regions for varying scenarios by defining material properties and dimensions of the model, and allows for the adjustment of the stimulation configuration, electrode contact spacing, customisation of electrode size and modulation of stimulation current. It enables the user to tune the model settings towards their specific needs and explore the possibilities by visualising the results from different angles by defining subsets of the entire solution. In line with expectations, the presented models show that the model parameters can influence factors such as the generated electric field, the current density and electric potential, which are indicators of neuronal activation. The findings support the hypothesis that these parameters should be considered during electrode design to achieve accurate stimulation.

Background
Damage dealt to the central nervous system (CNS) caused by trauma or disease can have detrimental effects on human quality of life because the CNS has limited regenerative capabilities. Efforts to replace lost neural cells require improved knowledge and methods for differentiation of neural stem cells (NSCs).

Objective
In this thesis, I aim to chart our current scientific knowledge and progression of neural differentiation and explore practically the feasibility of continuous ultrasound (US) stimulation on neural progenitor cells (NPCs) differentiating into neurons in vitro.

Methods
I performed a literature study examining previous studies that investigated electrical stimulation, nanoparticles, or ultrasound to improve in vitro or in vivo differentiation of neural stem cells. Using finite element method (FEM) frequency analyses with COMSOL Multiphysics, I investigated the use of a 24-wells plate with piezoelectric lead zirconate titanate [Pb(ZrxTi1–x)O3] (PZT) US transducers. Additionally, I validated in vitro previous findings on the feasibility of differentiation of NSCs to NPCs.

Results
With the knowledge gained from literature and findings from the experiments, I created a mold for the fabrication of a custom variant of a 24-wells plate made with polydimethylsiloxane (PDMS) to which 2.03 mm thick PZT can be mounted.

Conclusion
Future research efforts should focus on further developing this technique, specifically into electrical schemes to optimize US transmission to NPCs.

Neurological disorders profoundly affect quality of life, impairing essential functions like memory, cognition, and movement. Traditional treatments are often limited due to the complexity of the brain, the invasiveness of procedures, or the poor precision of the treatment. Ultrasound (US) neuromodulation offers a promising and minimally invasive alternative, using ultrasound waves to modulate neural activity in deep brain regions with high spatial resolution. Despite its potential, a technological gap limits preclinical ultrasound neuromodulation studies, as current ultrasound devices are too bulky and imprecise for flexible and continuous neuromodulation in freely moving rodents. Moreover, precise targeting is essential for effective neuromodulation, necessitating fine control over the focal spot location. This thesis aims to discover new system-level and circuit-level approaches to designing a miniaturized, power-efficient, and high-spatial-resolution 2D phased array transmitter application-specific integrated circuit (ASIC). These approaches are guided by the vision of implementing a miniaturized ultrasound transducer device for precise ultrasound neuromodulation in freely moving rodents, ultimately laying the foundation for future clinical applications in neurological disorder treatment.
The thesis begins with a literature review to determine the acoustic parameters required for effective ultrasound neuromodulation. Then, a comprehensive system-level study using the k-Wave Matlab toolbox is conducted, examining factors critical to the design of a 2D phased array transducer. This study highlights how parameters such as driving voltage, phase quantization resolution, frequency, and array length significantly impact the performance of the 2D phased array ultrasound transducer by influencing key acoustic and electrical attributes. Optimizing these parameters is essential to achieving high focal gain and spatial resolution while minimizing power consumption. The study finds that a three-bit phase resolution is optimal for accurate ultrasound beam steering. Additionally, while increasing array length improves spatial resolution and phased array gain, it also increases the overall size of the ultrasound transducer, constrained by the small brain sizes of rodents in preclinical studies. This study demonstrates that higher frequencies enhance spatial resolution and phased array gain. However, due to kerf limitations in the dicing manufacturing process of piezoelectric transducers, increasing frequency reduces the effective area of the transducers. Moreover, the pitch size should be reduced with increasing frequency in order to avoid the appearance of the grating lobes in the beam profile, forming a tradeoff between the frequency and the available area for the beamforming channels. Since higher frequencies are desired, particularly for applications in which higher spatial resolution is needed, high-voltage beamforming channels have been developed to investigate the highest frequency that allows for implementing a high-voltage beamforming channel. In this regard, two beamforming channels have been implemented and validated in TSMC 180-nm BCD technology, allowing for driving the ultrasound transducers with 12-MHz 36-V pulses. On the other hand, the power consumption of the beamforming channels has a crucial impact on the overall power efficiency of the ASIC. Therefore, a power-efficient HV pulser is developed, leading to 40.9% power consumption reduction with respect to the conventional HV pulser.
Utilizing the proposed beamforming channels, a prototype 2D phased array transmitter ASIC is developed to validate the findings of the initial system-level study. The transmitter ASIC includes a 12*12 array of beamforming channels that generate 12-MHz pulses with 20-V amplitude. The phasing circuitry in the ASIC generates the required delay for each beamforming channel, hence allowing the ASIC to steer the ultrasound beam. The measurement results proved the functionality of the phased-array transmitter ASIC in terms of phase quantization, showing a maximum DNL of 0.35 LSB. Furthermore, the ASIC allows for controlling the driving voltage and the burst duration, as needed for US neuromodulation applications. Based on the prototype 2D phased array transmitter ASIC, a US-guided US transmitter ASIC is developed for preclinical US neuromodulation experiments in freely moving rats. Considering the size of the rat's brain, the size of the ASIC is limited to 5*5 mm2. Given the prior discussion, the frequency was set to 12 MHz to achieve higher spatial resolution and phased array gain. Taking into account the proposed 20-V beamforming channel, the size of the 2D array was set to 66*66. According to the author's knowledge, this is the first transmitter ASIC that allows for preclinical US neuromodulation experiments on freely moving animals. Moreover, a temperature sensor was added to the ASIC to monitor the temperature rise in the ASIC. In order to monitor the location of the focal spot during the US neuromodulation, synthetic aperture US imaging capabilities are added to the ASIC to track the tissue displacement caused by the US neuromodulation push. In this regard, the ASIC includes eight groups of receiver channels that will be connected to 96 piezoelectric transducers using an analog multiplexer. To prevent side lobes in the beam profile, a system-level study was conducted to determine the optimal placement of receiver channels within the 2D array. The study revealed that a random distribution of receiver elements among the 2D transmitter elements is the most effective arrangement for minimizing side lobe appearance. To the best of the author's knowledge, this is the first ultrasound neuromodulation ASIC with integrated imaging capabilities. The measurement results proved the functionality of the ASIC in generating 20-V 12-MHz pulses with a programmable burst duration. The simulation results have shown that the receiver channels are capable of recording the echoes required for US imaging to monitor the location of the focal spot.
In summary, this thesis describes innovations in system-level and circuit-level integrated circuit design for developing an ultrasound ASIC that can be integrated with a 2D array of piezoelectric transducers, forming a miniaturized phased array US transducer that allows 3D US beam steering and focusing with the highest combination of spatial resolution and penetration depth in the context of pre-clinical studies. The ASIC includes the receiver channels that can be used for US imaging, allowing for monitoring real-time monitoring of the location of the focal spot during the US neuromodulation. This US ASIC leads to the emergence of miniaturized US transducers compatible with preclinical behavioral experiments in rodents, paving the way for the development of therapies to treat neurological disorders in humans.

In the emerging research field of bioelectronic medicine, it has been indicated that neuromodulation of the Vagus Nerve (VN) has the potential to treat various conditions such as epilepsy, depression, and autoimmune diseases. In order to reduce side effects, as well as to increase the effectiveness of the delivered therapy, sub-fascicle stimulation specificity is required. In the electrical domain, increasing spatial selectivity can only be achieved using invasive and potentially damaging approaches like compressive forces or nerve penetration. To avoid these invasive methods, while obtaining a high spatial selectivity, a 2 mm diameter extraneural cuff-shaped proof-of-concept design with integrated Lead Zirconate Titanate (PZT) piezoelectric ultrasound transducers is proposed. For the development of the proposed concept, wafer-level microfabrication techniques are employed. Moreover, acoustic measurements are performed on the device, in order to characterize the ultrasonic beam profiles of the integrated PZT-based piezoelectric ultrasound transducers. A focal spot size of 200 μm by 200 μm is measured for the proposed cuff. Moreover, the curvature of the device leads to constructive interference of the ultrasound waves originating from multiple piezoelectric ultrasound transducers, which in turn leads to an increase of 45% in focal pressure compared to the focal pressure of a single piezoelectric ultrasound transducer. Integrating piezoelectric ultrasound transducers in an extraneural cuff-shaped design has the potential to achieve high-precision ultrasound neuromodulation of the Vagus Nerve without requiring intraneural implantation.

Vagus nerve stimulation serves as an approach to manage drug-resistant epilepsy. Focused ultrasound stands out as a unique tool, offering a non-invasive modality for VNS. With its excellent depth of penetration and sub-millimeter focusing capabilities in soft tissue, ultrasound allows for precise targeting of the vagus nerve without the need for surgical intervention. This technology relies on piezoelectric transducers that convert electrical energy into ultrasonic waves. However, the high costs and technical difficulties associated with the production of these arrays have led to limited commercialization of piezoelectric transducer arrays.

This thesis investigates the design, fabrication, and optimization of a small, low-cost phased-array piezoelectric ultrasonic transducer. The transducer is specially designed for non-invasive UVNS applications in the form of a small patch for the neck. The research is dedicated to the fabrication of the transducer array, designing a transducer capable of both transmitter (TX) and receiver (RX) functions, which in the future can be used for simultaneous image-guided neural regulation. Two prototypes were fabricated: one is a 256-element 2MHz rectangular transducer array integrated into an 8-layer stacked rigid printed circuit board (PCB), and the other is a dual-frequency array integrated into a 4-layer flexible PCB substrate, explicitly designed for simultaneous imaging and neural modulation.

The study utilized two key materials, the well-known PZT and the lesser-known but promising <001> oriented PMN-0.3PT piezoelectric single crystal. Through extensive optimization, air-filled PZT transducer arrays were realized, and the cutting parameters for PMN-PT were significantly improved. A unique aspect of the fabrication was the use of flip-chip technology to directly integrate the air-filled 2D array onto the PCB, a method seldom detailed in existing research.

While exhaustive testing on phase-delayed beam-forming and imaging capabilities could not be conducted due to equipment limitations, preliminary evaluations of the capacity to generate plane waves were made. The properties of the bulk PZT and PMN-PT transducers were characterized. It was observed that the PMN-PT transducer generated significantly lower ultrasonic pressure compared to the PZT transducer, which aligns with expectations due to the fragile nature of PMN-PT. The first prototype of the PZT array transducer was successfully fabricated and assessed, achieving a pressure of 176 kPa at its 10 mm natural focal point. Subsequently, the internal array second of the prototype was fabricated and evaluated, registering a pressure of 1.14 MPa at the 1.2 mm natural focus point.

The results demonstrate the feasibility of building complex piezoelectric arrays using optimized fabrication methods. This lays the foundation for future prototypes to test beam-forming capabilities and for fabrication of the complete equipment afterwards.

Micro-Doppler (µDoppler) ultrasound imaging is a high frame rate ultrasound imaging modality that provides high spatiotemporal resolution ultrasound images of blood flow. It is sensitive to slow blood flow and particularly suitable for capturing fast-changing phenomena like rapid blood flow. Clutter filtering is an essential step in µDoppler data processing to reject tissue clutter signals and keep blood flow information as much as possible. 3D freehand µDoppler imaging is an emerging ultrasound technique that can construct full spatial vasculature images with a panoramic view that conventional 2D ultrasound is not able to provide. As freehand implies the continuous and nonuniform movement of the probe, it becomes more challenging for clutter filtering to acquire high-quality images.

This thesis explores and compares different state-of-art clutter filtering techniques on freehand in-vivo µDoppler imaging of the human brain. Specifically, Singular Value Decomposition (SVD), Robust Principle Component Analysis (Robust PCA), and Independent Component Analysis (ICA) clutter filtering techniques have been investigated. The aim is to test and compare their performance on in-vivo µDoppler ultrasound data with freehand probe movement and understand how freehand motion affects the threshold selection criteria. Besides that, a newly proposed method that combines ICA clutter filtering and clustering is included in this thesis to bring another perspective for sorting independent components corresponding to blood flow and rejecting unwanted ones consisting mostly of tissue clutter signals.