Tyndall National Institute - Doctoral Theses

Permanent URI for this collection

https://hdl.handle.net/10468/3619

Browse

Now showing 1 - 5 of 229

The potential role of optical guidance for bone-related biomedical applications in orthopedics and neurosurgery
(University College Cork, 2024) Li, Li Yao; Andersson-Engels, Stefan; Burke, Ray; Science Foundation Ireland
Optical sensing technology was explored as a means of intraoperative guidance for bone-related procedures in orthopedics and neurosurgery. Specifically, the feasibility of diffuse reflectance spectroscopy (DRS), a non-invasive and real-time optical technique that measures diffusely reflected light off samples of interest, was investigated in the thesis to differentiate biological tissue types and inform tissue boundaries as an intraoperative safety measure for revision total hip arthroplasty. Feature selection (FS) frameworks based on DRS measurements were developed utilizing machine learning techniques to determine wavelength features of optimal discriminative power for bone-related surgical procedures. Four FS frameworks, incorporating principal component analysis (PCA), linear discriminant analysis (LDA), backward interval partial least squares and an ensemble approach (biPLS), were designed with high adaptability to facilitate modifications and applications to other clinical scenarios. A feature subset of 10 wavelengths was generated from each FS framework yielding promising balanced accuracy scores for the one-vs-rest binary classification task. For cortical bone versus the rest class labels, PCA, LDA, biPLS and ensemble -based FS framework computed balanced accuracy scores of 94.8 ± 3.47%, 98.2 ± 2.02%, 95.8 ± 3.04% and 95.8 ± 3.16, respectively. For bone cement versus the rest, 100% balanced accuracy scores were generated from all FS frameworks. Subsequently, an in-house designed optical probe integrating DRS sensing was engineered and examined in ex vivo experiments. The most discriminative DRS wavelengths, selected by the FS frameworks including 1200 and 1450 nm, were incorporated as the illumination light sources. Furthermore, the performance of DRS to predict drilling depths in cranial bones was evaluated for craniotomy. Two models including partial least squares (PLS) regression and feedforward neural networks (FNN) were examined for prediction of skull thickness ranging from 1 to 5 mm away from the brain, yielding a root mean squared error regression loss of 0.08 and 0.06 mm from PLS, and 0.2 and 0.1 mm from FNN by using all versus selected features as model inputs, respectively. The predicted depths served as a safety protocol to indicate lookahead distances. On the other hand, the potential of ultrafast lasers in bone-related surgical applications was reviewed and explored from multiple perspectives. The advantages offered by ultrafast lasers over conventional laser systems (continuous wave or long-pulse lasers) included superior precision and minimized collateral thermal damage to surrounding tissues. However, clinical translation of ultrafast lasers to surgical applications had been constrained by limitations in pulse average power and material removal rate. In contrast, the use in implant surface texturing had advanced substantially, effectively enhancing bioactivation and osteointegration within bone matrices. At the end, ambient mass spectrometry, which employed a picosecond laser system for plume generation, was additionally assessed for tissue differentiation in a preliminary study. The classification model employed PCA for dimensionality reduction and LDA for multi-class classification. By using the reduced mass spectra dataset, bone cement was distinguished from biological tissue types with 100% in different classification metrics (precision, recall, F1 score). The highest misclassification rate occurred between trabecular and cortical bone with 18 instances where trabecular bone was classified as cortical bone. Overall, the research presented in the thesis has demonstrated promising results to advance basic science and consequently set the foundation for translational study of integrating optical sensing into surgical tools in bone-related procedures with valuable insights. This work was supported by Science Foundation Ireland (SFI), Grant No. SFI/15/RP/2828 and Grant No. SFI/22/RP-2TF/10293.
Novel Vertical-External-Cavity Surface-Emitting Lasers (VECSEL) systems and applications
(University College Cork, 2024) Bondaz, Thibault Alain Georges; McInerney, John G.; Laurain, Alexandre; Jones, R. Jason; Moloney, Jerome V.
The development of laser technologies represents a pivotal milestone in human technological history, with semiconductor lasers playing a particularly transformative role due to their versatility, efficiency, and tunability. This thesis focuses on Vertical-External-Cavity Surface-Emitting Lasers (VECSELs), a subset of semiconductor lasers that offer unique advantages for high-power, single-mode operation and the integration of intracavity optical elements. These properties enable their application across a broad spectrum of fields, including terahertz (THz) generation and nonlinear imaging. This work presents two novel VECSEL-based systems. The first is a bicolor VECSEL designed for room-temperature THz emission via difference frequency generation (DFG). This system achieves milliwatt-level THz power and employs a two-stage stabilization method to enhance operational stability. The second system is a 1.5 GHz broadband pulsed laser developed for multiphoton microscopy. This high-repetition-rate laser minimizes photodamage and photobleaching. By using photonic crystal fibers and a Multiphoton Intrapulse Interference Phase Scan (MIIPS) scheme, the system generates a wide supercontinuum spectrum, enabling advanced pulse shaping for nonlinear imaging.
An investigation of materials and printing methods on the response of 3D printed supercapacitors
(University College Cork, 2024) Ferguson, Matthew; O'Dwyer, Colm
The primary objective of this thesis was to examine in detail the behaviour and electrochemical performance of symmetric supercapacitors made using two different 3D-printing methods (Vat-P and FDM). Two cell types are made in this study, one with metallized Vat-P-printed current collectors, the other with conductive PLA (polylactic acid) FDM-printed current collectors in a similarly designed printed coin cell. In Chapter 3, an array of carbon-based active materials was made and tested in both cell types. Chapter 4 focuses on the FDM printed current collectors, examining how solvent pre-soaking effects the electrochemical performance and cycle life of the 3D-printed supercapacitor cells containing these treated current collectors. Chapter 5 introduces pseudo-capacitive Mn3O4-inverse opal electrodes to the 3D-printed supercapacitor cells and highlights the enhanced performance of these electrodes when composited with conductive nanocarbons.
Nanostructured magnetic materials for integrated magnonic devices
(University College Cork, 2024) Samanta, Arindam; Roy, Saibal
This PhD thesis explores the fabrication, characterization, and application of advanced exchange spring (ES)/exchange coupled (EC) nanoheterostructures and their magnetic properties. Current challenges in the field of magnetics include the development of materials that demonstrate tunable magnetic properties, particularly in terms of controlling anisotropy and spin dynamics at the nanoscale. This thesis addresses these challenges by utilizing an electrodeposition technique, we have been able to develop for the first time in situ amorphous/nanocrystalline cobalt-phosphorus (CoP) thin films at room temperature. These films exhibit a unique transverse exchange spring structure due to the interplay between in-plane (IP) and out-of-plane (OOP) anisotropies. The inherent IP anisotropy of the amorphous phase competes with the OOP anisotropy of the nanocrystalline structure, producing characteristic stripe domain structures that evolve into novel corrugated stripe domain shapes. Systematic investigations reveal the evolution of hysteresis loops in these thin films, showing a transition from low coercivity non-ES loops to staircase-ES loops with multiple coercivities in thicker films. The First Order Reversal Curve (FORC) distributions demonstrate various reversal mechanisms within the samples, confirming the transition between non-ES and ES states based on the prevalent interfacial exchange coupling. Additionally, the field-dependent Brillouin Light Scattering (BLS) spectra unveil distinct spin wave modes, with ES films showing well-resolved bulk and Damon-Eshbach surface spin wave modes, while non-ES films exhibit mode doublets below a certain applied field threshold. These findings indicate a linear dependence of mode frequencies on magnetic field intensity and enhanced exchange coupling in thicker ES films. On the other hand, ultrafast magnetization dynamics studies highlight the in-plane magnetic orientation (φ) dependent ultrafast demagnetization and precessional dynamics of electrodeposited non-exchange spring nanostructured CoP alloys. The precession frequency shows dominant two-fold anisotropy superposed with moderate four-fold anisotropy, while the Gilbert damping coefficient exhibits four-fold anisotropy. The ultrafast demagnetization remains nearly isotropic with φ, suggesting a significant role of spin-orbit coupling (SOC) in anisotropic precessional dynamics and isotropic spin-fiip scattering processes. These detailed studies of ultrafast spin dynamics reveal crucial dynamical properties for potential applications in high-frequency integrated magnetic passives for future monolithic on-chip power supplies. Furthermore, this thesis introduces novel "Magnon Microwave Antennas" (MMAs) for generating tunable microwave frequencies without external bias magnetic fields. The MMAs, comprising patterned arrays of magnetostrictive nanomagnets embedded in piezoelectric heterostructures, generate multimode microwave frequencies through the phonon-magnon coupling. Static magnetic studies elucidate various magnetization reversal processes within the nanowire and nanodot arrays, unveiling the critical role of demagnetization energy distribution in tuning the domain configuration and power-phase distributions of these MMAs. Functional tuneability has been proposed to be achieved through amplitude-dependent training using different combinations of nanowire and nanodot dimensions, topologies, material properties, and array configurations. The non-volatile nature of the spin textures generated in MMAs under bias-free conditions holds promise for energy-efficient logic and low-power computing applications. Thus, the comprehensive research presented in this thesis paves the way for the development and exploitation of next-generation nano-heterostructures for various cutting-edge magnetic vis-à-vis magnonic applications, including on-chip reservoir computing, leveraging the unique magnetic vis-à-vis magnonic properties and their tuneability in these advanced materials/devices.
AI-enabled chipless RFID sensing system for reliable IoT applications
(University College Cork, 2024) Rather, Nadeem; O'Flynn, Brendan; Buckley, John; Tedesco, Salvatore; Simorangkir, Roy B.V.B.; Science Foundation Ireland
The Internet of Things (IoT) is growing rapidly, driving the need for innovative and sustainable solutions for wireless identification and environmental monitoring. Passive Radio Frequency Identification technology (RFID) has been a key wireless communication technology enabling IoT. Recent advances have paved the way for battery-less, chipless RFID (CRFID), which eliminates the need for an integrated circuit (IC) component on the tag. This PhD thesis presents a new design strategy for developing concentric rings-based polarization-insensitive CRFID sensing tags. The proposed tag design approach of exponential spacing results in an 88.2% higher tag data encoding capacity than conventional designs which incorporate uniform spacing of the resonant rings. This is coupled with the idea of using the innermost ring for capacitive sensing. The concept of using RCS nulls for data encoding is implemented to enable convenient and accurate sensing by the innermost ring. This is made possible by adding an extra ring at the tag’s outermost edge. To enable robust detection of these tags, Artificial Intelligence (AI) is integrated on the reader side, employing both machine learning (ML) and deep learning (DL) techniques for decoding RCS EM signatures. In this research, ML and DL regression modelling techniques are applied to a dataset of measured RCS data derived from large-scale automated measurements of custom-designed, 4-bit CRFID sensor tags. The robotic measurement system is implemented using the first-of-its-kind automated data acquisition method using an industry-standard robot. The results show that all the ML/DL models were able to generalize well, that is, the ability of a model to perform accurately on new, previously unseen data. However, the 1D-CNN DL models outperformed the conventional ML models in the detection of ID and sensing values. In another contribution, a 3-bit depolarizing CRFID tag is developed and enabled for surface and shape robust detection using AI. For the first time reported, the system was trained on a dataset of 12,600 EM signatures, capturing varying surface permittivity, tilt angles, read ranges, and tag bend scenarios. The AI models using 1D-CNN are trained and validated, resulting in a low RMSE of 0.040 (0.66%) for tag ID detection. On the same dataset, for the first time, DL models were evaluated with Bidirectional Long Short-Term Memory (Bi-LSTM) and attention mechanism, further reducing the RMSE to 0.029 (0.48%). The outcomes of this thesis contribute significantly towards state-of-the-art of AI-enabled CRFID systems for robust and reliable real-world IoT applications.

Browse

Recent Submissions