Biomedical Engineering - Theses

Permanent URI for this collection

Search Results

Now showing 1 - 10 of 19
  • Item
    Thumbnail Image
    Computational Investigations into the Role of Membrane Tension in Cell Migration
    Collette, Jared ( 2021)
    Migratory cells are well understood to experience changes in membrane tension in response to chemical and mechanical stimuli. Less understood are the direct changes to intracellular signalling processes and intracellular mechanical processes involved in cell migration resulting from changes in the mechanical state of the cell. These mechanical states are important in furthering our understanding how diseases like cancer metastasis hijack these processes. This thesis explores the relationship between membrane tension and signalling processes through mechanobiological feedback. In this work, a novel mathematical technique is developed for spatiotemporal signal transduction pathways described by reaction diffusion in moving domains. Then, a new link is explored between the relationship of membrane tension and a migrating cell’s ability to adapt and maintain polarity in high and low chemo-stimulant environments. Finally, the relation between membrane tension and cell protrusion is explored in 3D through growth mechanics, allowing for a more thorough mechanical characterization of various types of 3D protrusions. This work culminates into furthering our understanding of the role of membrane tension in migratory cell processes and provides more insights and knowledge useful in the development of new therapeutics and treatments.
  • Item
    No Preview Available
    Tuning the Bioactivity of Cell-derived Extracellular Matrices through Control of Substrate Properties
    Yang, Michael ( 2021)
    Mesenchymal stromal cells (MSCs) are the subject of thousands of clinical trials for treatment of innumerable human pathologies. However, their widespread clinical use is still hampered by difficulties in retaining their stem cell-like properties during prolonged ex vivo expansion. The use of cell-derived extracellular matrix (ECM) has recently garnered much interest as a culture substrate because ECM significantly improves cell viability in ex vivo culture. However, these biomaterials are themselves produced in an environment which is not physiologically accurate, and therefore, in this study the aim was to optimise bioactivity of cell-derived ECM by controlling substrate properties to more accurately mimic the in vivo niche. MSCs are innately sensitive to substrate stiffness and their surrounding mechanical environment. The first results chapter (Chapter 2) describes investigation into whether manipulation of substrate stiffness increases bioactivity of cell-derived ECM. Polyacrylamide hydrogels with tunable stiffnesses were fabricated, and MSCs induced to deposit ECM on these surfaces. Subsequent culture and behavior of primary MSCs on the ECM was then observed. Primary cells cultured on ECM deposited on soft substrates demonstrated the highest levels of proliferation. On the other hand, there were significantly higher levels of osteogenesis when culturing MSCs on ECM deposited on stiff substrates. These results show that the bioactivity of ECM can be taken even further by controlling mechanical properties of substrates that mimic the cellular microenvironment. In addition to substrate stiffness, surface chemistry was controlled to determine whether this affected ECM bioactivity in a similar fashion, described in Chapter 3. The surface chemistry of glass coverslips was modified with various silanes, introducing amine, carboxylic acid, propyl, and octyl functional groups onto the surfaces, and studied how introduction of these moieties affected the ECM. ECM on these surfaces improved cellular proliferation and adipogenesis. However, the bioactivity of the ECM did not depend significantly on the surface chemistry for the range of chemistries and culture timescales tested. These results allow for greater flexibility in fabricating tissue engineering materials with a variety of surface chemistries for use as ECM scaffolds. Lastly, Chapter 4 describes an investigation to determine whether the beneficial effects observed by tuning substrate stiffness could be scaled up and applied to 3D culture systems. Porous polyacrylamide cryogels were fabricated for the culture of MSCs and subsequent ECM deposition. MSCs cultured in this manner did not demonstrate improved cell behavior. However, these 3D scaffolds were a suitable material for the deposition of cell-derived ECM, providing a future avenue of research for the scaleup and clinical translation of these matrix biomaterials. This dissertation presents findings which demonstrate that the effects of varying substrate stiffness can be combined with use of ECM to increase MSC proliferation and osteogenesis. Altering surface chemistry is also shown to not have a significant effect on dECM bioactivity for the range of functional groups tested. Fabrication of tunable porous cryogels with ~65 um pores also revealed that secretion of ECM by MSCs is not affected sufficiently to result in changes to its bioactivity. These findings contribute to the field by showing that parameters such as substrate stiffness can be manipulated and used with ECM in cell culture to result in improved cell behavior compared to using ECM alone, and also illustrates that strict control of other variables such as surface chemistry and three-dimensionality are not necessary in order to maintain the bioactivity of cell-derived ECM.
  • Item
    Thumbnail Image
    B1 insensitive techniques for ultra-high field magnetic resonance imaging
    Green, Edward Michael ( 2021)
    Magnetic resonance imaging (MRI) is a powerful technology that is widely used in medical imaging and science. MRI is a flexible imaging modality that can be used to acquire a range of structural and functional information non invasively and without the use of ionising radiation. Ultra-high field (UHF) MRI allows for improved outputs from a range of imaging sequences due to increased signal-to-noise ratio and increases in contrast. However, due to wavelength effects, UHF imaging suffers from inhomogeneity of the RF transmit field, B1+. Nonuniformity of the B1+ field results in spatial variation in image intensity and contrast which compromises image utility. Existing approaches for overcoming the effects of B1+ inhomogeneity include specialised pulse design, use of multiple element transmit arrays and correction to images in post-processing. As each approach has limitations, none is used universally. This thesis seeks to improve the capabilities of B1+ insensitive imaging at UHF with three major contributions. In this thesis a new class of pulses is described, termed Spin Lock Adiabatic Correction pulses, aimed at decreasing the required energy deposition to achieve a given level of B1+ insensitivity in excitation or inversion. Additionally, theory is developed to describe an approximation for adiabatic pulse behaviour. A Fourier relationship is demonstrated for a range of well known adiabatic pulses and the strengths and limitations of the approximation are explored with simulation. Finally, this thesis presents a novel imaging pipeline for production of B1+ artifact-free images. The method combines super resolution imaging with measurement and correction of B1+ field inhomogeneity in order to achieve B1+ insensitive imaging for 2D slice stacks with no time penalty over uncorrected slice stacks.
  • Item
    Thumbnail Image
    Quantitative micro-computed tomography for joint mechanobiological measurement
    Oliveira Silva, Mateus ( 2021)
    Osteoarthritis (OA) is associated with tissue damage and loss of function, which can be observed in preclinical studies using micro-computed tomography (microCT). MicroCT has been used to obtain and evaluate structural changes in the knee joint of OA models via bone morphometric analysis. However, bone morphometric analysis has yet to be used to evaluate early structural changes in a collagenase-induced OA mouse model at multiple time points. In this model, an acute inflammatory response starts immediately; however, the onset and progression of structural changes have not been fully described. Some of these changes are related to the osteochondral interface. The osteochondral interface is a transitional zone anchoring articular cartilage to subchondral bone. It has been shown that during OA, increased neoangiogenesis creates porous channels in this zone, allowing the transport of molecules linked to OA development. Importantly, the connection between these porous channels and the early stages of OA development is still not fully understood. While standard microCT is usually able to image and distinguish mineralised tissues (such as cortical bone), it is not enough to differentiate between unmineralised tissues (such as articular cartilage and surrounding soft tissues or the osteochondral interface). Therefore, a contrast agent is necessary to image the porous channels at the osteochondral interface and distinguish the different tissues (both mineralised and unmineralised) in this region. Thus, the aim of this doctorate work was to develop robust protocols and tools for acquiring accurate joint mechanobiological measurements with contrast-enhanced microCT. In order to develop this, three specific aims to be addressed in this work were formulated, namely; 1. Evaluate early structural changes in the knee joint over time in a collagenase induced OA mouse model with microCT, 2. Synthesise and optimise a lanthanide-based microCT contrast agent and evaluate it in sintered porous channel models and in mouse subchondral bone, 3. Couple the optimised lanthanide nanoparticles with dextran as a biocompatible tracer and evaluate their cytotoxicity. In response to these aims, this thesis implements a multidisciplinary approach focused on investigating creative methods to further investigate OA progression in the osteochondral interface with microCT. First, multiple structural parameters in the collagenase-induced OA mouse model were obtained from two subchondral bone compartments, e.g., cortical, and epiphyseal trabecular bone. Statistically significant differences between OA and control samples were observed since early disease progression stages in several bone morphometric measures, e.g., bone volume (BV), bone volume density (BV/TV), trabecular thickness (Tb.Th), trabecular separation (Tb.Sp), trabecular number (Tb.N), cortical thickness (Ct.Th) and cortical porosity (Ct.Po). These findings showed that bone remodelling activity occurs simultaneously with the acute inflammatory response in this model. These findings also stimulated further investigation of early structural changes occurring specifically in the osteochondral interface. Due to the interconnected nature of the osteochondral interface, the following aim was to explore the optimisation of nanoparticle-based microCT contrast agents for potentially imaging the porous channels at the osteochondral interface. BaYbF5 nanoparticles were synthesised and coated with a biocompatible silica shell (SiO2), generating an optimised microCT contrast agent in order to obtain an appropriate contrast attenuation for subsequent segmentation of structures of interest (i.e., porous channel models and subchondral bone). For this purpose, a proof-of-concept was successfully demonstrated in porous channel models and mouse subchondral bone, in which the optimised nanoparticles successfully increased their X-ray attenuation. Finally, this work describes a series of chemical reactions to couple the optimised nanoparticles with a biocompatible tracer, dextran, that can be transported via the vascular system and then diffuse through the osteochondral interface. These reactions generated a novel compound, NP-Dextran, able to provide efficient contrast in microCT and potentially target the osteochondral interface in animal studies. Extensive characterisation techniques were performed to evaluate the NP-Dextran chemical structure. Further, an in vitro toxicity test demonstrated the low toxicity of the novel compound NP-Dextran, evidencing a high potential for its use in vivo animal experiments for further investigating OA progression.
  • Item
    Thumbnail Image
    Computational Investigation into Cadherin Clustering and Patterning at Cell-Cell Adhesions
    Yu, Qilin ( 2021)
    Cadherin is a transmembrane protein at adherens junctions (AJs), which physically link adjacent cells through trans and cis extracellular binding and cytoplasmic interactions with the actin cytoskeleton. As the adhesiveness of individual cadherin molecules is negligible, assembly of cadherin clusters is necessary to maintain junction integrity. Cadherin clusters are observed in different sizes and distributions, but limited studies have examined the mechanisms that regulate cadherin clustering and patterning systematically. In this dissertation, firstly a lattice-based model was implemented to explore how cadherin-cadherin interactions and membrane environments affect cadherin clustering at cell-cell adhesions. Then a novel model was constructed by coupling the lattice-based model of cadherin dynamics with a reaction-diffusion model of actin cytoskeletal remodeling. The model recapitulated key characteristics of the process of cell-cell contact maturation in a cell doublet. The model study garnered a novel insight that a positive feedback loop between cadherin and F-actin is essential for the formation of a characteristic ring distribution of cadherin and actin on the contact edge. As we know that actin can physically manipulate the dynamics of cadherin molecules through tethering and corralling, a new coarse-grained model was then created to further explore the effects of factors such as actin density, actin structure and cadherin/F-actin binding affinity. The model showed that actin tethering facilitates cadherin clustering, by localizing cadherin molecules near F-actin. However, when there are more actin binding sites than the number of cadherin molecules, actin binding can also compete with cis binding for cadherin molecules and destabilize cadherin clusters. Finally, in vitro experiments were performed using the 3D stochastic optical reconstruction microscopy (3D-STORM), which captured a thorough profile of E-cadherin clusters on the apical, lateral and basal junctions of the A431 cells. The experimental observations suggest that our model provides plausible mechanistic observations of cadherin clustering patterns in a more complex cell culture system to the cell doublet.
  • Item
    No Preview Available
    Mechanical Consequence of Glycosaminoglycan Content and Arrangement in Native Cartilage for Tissue Engineering Applications
    Rathnayake Mudiyanselage, Manula Saubagya Bandara Rathnayake ( 2021)
    Cartilage is a connective tissue found in the body which performs a variety of mechanical and protective functions according to its location. Cartilage is classified into three types: hyaline, elastic, and fibrocartilage. Cartilage related injuries and congenital conditions can have long-term effects on the quality of life of affected individuals. Tissue engineering approaches to repair degenerated cartilage have been investigated since the early 1990s. Producing tissue-engineered cartilage with mechanical properties similar to native cartilage is a key issue reported in literature. The basic building blocks of cartilage include chondrocytes, water, collagen, elastin, proteoglycans, and glycosaminoglycans (GAGs). Proportions of these components vary according to cartilage type. Among those components, the specific GAGs present in each type of cartilage are not well investigated. The structural interactions between GAGs and other extracellular matrix (ECM) macromolecules are overlooked in literature. The differences in mechanobiological environments of different cartilage types are under-reported, and there is a necessity to develop methods that can evaluate the mechanobiological environment of native cartilage, to facilitate better replication of native cartilage in tissue engineering approaches. Taking these gaps into account, the following research aims were formulated. -Explore the glycosaminoglycan mediated interactions between tissue components in different types of cartilage. -Identify the intrinsic glycosaminoglycan expression in native cartilage and investigate how this can be reproduced in tissue-engineered models. -Develop an image-guided micromechanical evaluation (IGME) approach to investigate the mechanobiological environment of the cartilage. Selective GAG depletion was used to explore the variations in GAG mediated interactions in different cartilage. Articular, auricular, meniscal, and nasal cartilage plugs were treated with chondroitinase ABC, guanidine hydrochloride, and hyaluronidase, and remaining GAG content was quantified. The ability of these reagents to deplete GAG depends on molecular and structural interactions between GAGs and other macromolecules in the ECM. GAGs in auricular cartilage showed strong interactions with the rest of the ECM when compared to the other cartilage. Hyaluronidase treatment removed over 99% GAG content from other cartilage but only 76% from auricular, indicating a strong interaction between GAGs in auricular cartilage and ECM or elastin fibres present in the tissue. Overall, this showed GAG-ECM interactions vary according to cartilage type and location, indicating specific structural roles for GAG types present in cartilage. Intrinsic GAG expression of cartilage was then investigated with immunohistochemistry to identify the spatial localisation of different GAG populations in cartilage. An extensive comparison of specific sulphated glycosaminoglycans (sGAG) types (chondroitin sulphate-CS, dermatan sulphate-DS, keratan sulphate-KS, and heparan sulphate-HS) present in high load bearing cartilage (articular and meniscal) and low load bearing cartilage (auricular and nasal) was carried out. Articular and nasal cartilage showed significantly different GAG expression despite being hyaline cartilage. CS and KS were expressed in every cartilage in both extracellular and pericellular areas, but in auricular cartilage middle zone, CS and KS were expressed only in pericellular areas where chondrocytes are bounded by elastin fibres. Expression of DS was expressed only in tensile load-bearing areas of articular cartilage (superficial zone) and meniscal cartilage (outer zone). Cartilaginous regions of other cartilage did not express DS. Differences in GAG expression between low load-bearing cartilage (auricular and nasal) and high load-bearing cartilage (articular and meniscal) indicates that GAG expression may depend on the local mechanobiological environment of cartilage. Immunohistochemistry results showed that GAG in auricular cartilage is associated with elastin fibres. Such spatial interactions were not seen in other cartilage types, which has collagen as the main fibrous component. Therefore, the effect of ECM components on intrinsic GAG production was investigated in tissue engineering models. Bovine articular chondrocytes were encapsulated in three groups of hydrogel beads containing 1) alginate, 2) alginate and collagen, 3) alginate, collagen, and elastin. Then CS, DS, and KS expression were investigated with immunohistochemistry at day 0, 7, 21, and 35 of culture. Expression and distribution of GAG types were compared between the different groups. CS was more concentrated in pericellular areas. With time, CS staining spread over a wider area in the beads with collagen and elastin. DS showed a more uniform distribution in the matrix. DS staining intensity in beads with collagen and elastin was higher when compared to alginate only beads. KS was only expressed by few cells. Results indicate that collagen and elastin have an effect on distribution patterns of GAG in 3D in vitro environments. Previous studies showed the GAG expression and their interactions in cartilage could depend on the local mechanobiological environment. Therefore, to visualise the mechanics involved in the local mechanobiological environment a novel image-guided micromechanical evaluation (IGME) technique capable of evaluating the local collagen and elastin deformation in cartilage was developed. Mechanical compression, multiphoton imaging, and digital volume correlation (DVC) were combined to analyse the displacement fields of collagen and elastin in auricular cartilage. Preliminary results suggest the feasibility of this method to analyse soft tissue mechanical deformation in 3D space. A custom-built compression device that can be placed under an objective of a multiphoton microscope was used to image cartilage under different strain levels (5%, 10%, 15%, and 20%). An image preprocessing pipeline was developed to segment collagen and elastin from the multiphoton images. 3D images were then used to perform DVC with TomoWarp2 software. This is the first study to evaluate the feasibility of analysing collagen and elastin deformation in cartilage without using markers. This method also allows assessment of individual components without altering the tissue composition. Therefore, this approach can be used to investigate mechanics in cartilage mechanobiological environment without altering it. In this work, it was shown that GAG-ECM interactions are unique to both the cartilage type and anatomical location. This variation of GAG also suggests the presence of different GAG populations in different cartilage. Therefore, spatial localisation of different GAG populations in different articular, auricular, meniscal, and nasal cartilage were investigated. GAGs present in auricular cartilage showed pericellular localisation pattern associated with elastin fibres present. This specific localisation pattern was not seen in other cartilage which had collagen as the main fibrous component. Therefore, effect of collagen and elastin on the production of different GAG types was investigated in 3D in vitro environments. In vitro culture of chondrocytes showed that GAG dispersion patterns are significantly different from the native GAG dispersion patterns. Finally, an IGME technique that would facilitate observing of the behaviour of individual ECM components in cartilage mechanobiological environment was developed and feasibility of that method was assessed.
  • Item
    Thumbnail Image
    Neuronal networks in absence epilepsy : dynamics of seizure transitions
    O'Brien, Patrick Pierre ( 2021)
    Epilepsy is the most common recurrent neurological condition, affecting a little over 1\% of the world's population. It is often debilitating because of significant psychological, social and cognitive burdens apart from the seizures themselves. Advancing treatment options for generalised epilepsy syndromes has proven a challenging problem, since they are often refractory to treatment with anti-epileptic medication, and are currently not amenable to surgical treatment, so that it is rarely ethical to conduct invasive intracranial EEG studies in human patients with these conditions in order to study them further. Although there has been controversy over the origin and networks involved in absence seizures, previous work by our group has shown that alternative areas of somatosensory cortex are important parts of the thalamocortical circuit in the generation of seizures. The work of this thesis aims to extend those studies by showing that the coupling between cortex and thalamus is altered during seizures. It is established that there is a process of resonance ocurring prior to and during seizures in relevant deep layers of cortex, and the mechanisms behind this are studied. Novel aspects of the work presented in this thesis include the examination of the EEG in the GAERS rat, a whole animal model of absence epilepsy, at the scales of both individual cortical microcircuits and whole cortical areas using simultaneous depth electrodes, and also at both very fine and long time scales. The studies described in this thesis show that the somatosensory cortex leads the thalamus in the first second of seizure onset, in agreement with the results of previous work. To account for the mechanisms behind this increased coupling at seizure onset, an analysis of the properties of resonance in different parts of cortex is presented. This indicates altered phase coupling prior to seizures, in the form of increased phase entrainment, phase coherence, and cross-frequency coupling. These changes are most marked in the junction between primary and secondary somatosensory cortex, with earlier, larger increases prior to seizures, and earlier decoupling at seizure termination. Dynamical analysis points to a number of processes underlying such increases in cortical resonance. Alterations in scaling behaviour are observed in different cortical regions, and also prior to seizures. Analysis of the Lyapunov spectrum of cortical depth and tetrode EEG data shows evidence for a separation in timescales of dynamics at seizure onset, and changes consistent with critical slowing. Further, gradual changes in the statistics of the largest exponents are seen with approaching seizures, suggesting an alteration in dynamical properties giving rise to critical slowing. The results of this thesis support the recently proposed cortical focus theory of absence seizure generation in GAERS rats, and provide new insight into cortical network changes relevant to seizures. This work provides a basis to explore the mode of action of antiepileptic drugs, and more targeted therapy for generalised absence seizures.
  • Item
    Thumbnail Image
    The design of seizure forecasters
    Payne, Daniel Eric ( 2021)
    Thirty percent of people with epilepsy have seizures refractory to all available medical and surgical treatment options. The uncertainty caused by the constant threat of seizures is the most significant factor impacting people's quality of life with refractory epilepsy. Seizure forecasters aim to alleviate this uncertainty by advising when seizures will occur. This thesis explores the design of seizure forecasters with the aim of improving seizure forecaster performance and clinical utility. Constraints on the seizure forecasting problem are considered, and algorithms for seizure forecasting are explored. Often after a seizure, the total energy in an EEG recording is suppressed, taking seconds to minutes to return to typical values. This period is associated with impaired cognition, drowsiness, and possibly Sudden Unexplained Death in Epilepsy (SUDEP). The postictal period and its relationship to seizure length is explored, motivated by the recent discovery of bimodal seizure length distributions within some patients. Postictal length is defined as the time from seizure end to when average EEG values pass average interictal levels. The findings showed that postictal suppression length is bimodal, with the short seizure population having noticeably shorter postictal periods. For seizure forecasting, this has important implications for how seizures should be labelled. Separately classifying different seizure populations may improve seizure forecasting performance. Forecasters may soon be effective enough to be clinically useful. To optimize clinical utility, the user requirements for a seizure forecasting device need to be determined. Requirements include forecasting performance and the preferred format and timing of the information provided to assist patients. User requirements of a seizure forecaster are explored, providing constraints to the seizure forecasting task. An app for use by the public was made to reproduce the uncertainty of having epilepsy to model the seizure forecaster experience. The app also provided forecasts for ‘events’ to model the seizure forecaster experience. Using this model had two potential benefits: a larger population could be asked about their views on forecaster requirements, and the people being asked had experience using a forecaster. The app asked questions relating to forecaster design, including how forecast information was portrayed, how valuable the forecasts were, and how it alerted the user. The insights gained from user feedback will be helpful to guide the design of future seizure forecasters. Many factors, both exogenous and endogenous, are considered seizure precipitants and may have utility in seizure forecasting. The use of weather, sleep, and temporal features to forecast seizures was explored in this thesis. These features were chosen based on the availability of data. The weather features considered were temperature, maximum daily temperature, minimum daily temperature, humidity, wind speed, rainfall, pressure, and pressure change. The sleep features were current sleep stage, hours asleep, hours awake, hours in stage 1/2/3, the number of sleep transitions, and the time since waking. The temporal features explored were the hour of the day, the time of the week, and the lunar phase. Forecasts from features were combined using a Bayesian approach. Weather, sleep, and temporal factors all performed better than chance in some but not all patients, with temporal features performing the best and weather feature performing the worst. A combination of all features produced forecasts that outperformed individual features on average. Finally, the use of Long short-term Memory (LSTM) neural networks to forecast seizures over long time scales was explored. LSTM networks have memory and, thus, an advantage over most other classifiers such as support vector machines or convolutional neural networks (CNNs). The network included CNN and LSTM components. The CNN component was able to produce short-term forecasts, and the LSTM component was able to use this information to forecast over minutes, hours, or days. The algorithm performed nearly as well as the best performing algorithm on the same dataset, surpassing the previous best performance for some patients. The LSTM networks were also able to perform better than chance when forecasting seizures more than a day in advance.
  • Item
    Thumbnail Image
    Decoding Sensorimotor Rhythms for Brain-Computer Interfaces
    Bennett, James David ( 2021)
    Brain-computer interfaces (BCIs) have great potential to improve the quality of life for people with severe motor disabilities. By measuring and interpreting neural activity, BCIs can predict and express intention through an external computerised device. This creates an alternative mechanism for communication and control for people with paralysis. Volitional changes in oscillatory activity near the sensorimotor cortex, known as sensorimotor rhythms (SMRs), can be measured with invasive techniques, such as electrocorticography (ECoG), or with non-invasive techniques, such as electroencephalography (EEG). This work explored novel approaches for decoding SMRs from EEG with the aim of utilising neurophysiological principles. This thesis also investigated the viability of vascular electrocorticography (vECoG) as a SMR-based BCI modality. The common spatial patterns (CSP) algorithm is used to linearly combine information from multiple EEG electrodes in order to accentuate SMR activity. Typical patterns of SMR activity derived by this method were characterised using publicly available EEG datasets. It was found that both neurophysiologically probable and improbable patterns of activity were commonly extracted and that selecting for, and adapting to, neurophysiologically probable activity could improve decoding performance. These findings highlight the importance of considering spatial filter adaptation in EEG BCI decoder design. The range of decoding algorithms available to BCI practitioners is extensive and diverse. A universal method for explaining the predictions of EEG decoders in terms of neurophysiologically relevant factors was investigated. The validity of the explanations was demonstrated using simulated EEG data. The method was also employed to compare the behaviour of four categories of decoders using real, pre-recorded EEG data. The results indicated that all decoders were able to harness neurophysiologically plausible electrodes and cortical sources to make accurate predictions. However, the influence of artifactual activity was also found to contribute to high decoder accuracy. These findings emphasise the need to understand the predictive behaviour of decoders and the proposed method may be useful a tool to help BCI researchers achieve this understanding. Vascular electrocorticography (vECoG) measures intracranial neural activity by chronically implanting a stent-electrode array within the brain vasculature. By omitting the need to penetrate the skull, this minimally invasive technique has the potential to safely record high fidelity neural information and be used as a long-term, clinically useful BCI. Data from the first-in-human clinical trial of the Stentrode was used to characterise vECoG signal quality. Participant-specific re-referencing was shown to improve signal quality and the maximum bandwidth of the signal was found to be consistent with previous animal studies. A multiclass, online SMR decoder was also implemented and tested with a single participant. Discriminatory activity from multiple motor imagery classes could be observed, however, signal non-stationarity affected online decoding performance. Together, the signal quality and multiclass decoding results suggest that vECoG has significant potential as a recording modality for a clinically useful BCI. Overall, the findings presented in this thesis contribute to two key facets of SMR BCI research. In terms of decoding SMRs from EEG, they demonstrate the need to harness neurophysiological phenomena, avoid contamination from artifactual activity and improve the interpretability of complex decoding algorithms. Furthermore, the vECoG results add to the growing body of evidence implicating this modality as a beneficial way for interfacing with the brain in a minimally invasive manner.
  • Item
    Thumbnail Image
    The dynamics and characterization of self-assembly in biopolymers and biosystems
    Jativa, Fernando ( 2020)
    Biological self-assembly is the foundation of the strength and elasticity which characterizes nature-derived biomaterials, including the cell’s architecture. Thus, an in-depth understanding of the mechanics behind this process can open the doors to various biomedical applications. For this purpose, this research employed novel experimental and characterization techniques to study self-assembly in biopolymers and biosystems. Initially, a droplet dissolution technique in liquid crystalline stages was used to analyze the process of self-assembly in two important biopolymers, silk and cellulose. Here, we report an effective and simple approach based on droplet dissolution in a liquid binary phase for the formation of silk fibroin transparent spheres as well as cellulose microbeads, both of which can span several hundred micrometers in diameter. The microstructure of the spheres formed at different ethanol concentrations was characterized by electron microscopy. High concentrations of ethanol caused droplets to be encased in a thin shell which collapses once it is taken out of the liquid phase. Generally, low ethanol concentrations produce transparent silk spheres and solid cellulose microbeads. This work on biopolymers demonstrates that controlled droplet dissolution self-assembling may be explored as a novel and effective way to tailor the microstructures of nature-derived biomaterials. The spheres generated in this manner have several different characteristics which can have multiple potential uses, such as templates for scaffolds, microcarriers, as well as photonics and nano-technological applications. The second part of this thesis investigated self-assembling in biosystems. Cell aggregates are an important tool in studying tissue remodelling, extracellular matrix formation, cell-cell interaction, and last but not the least, tissue-like biomechanical properties. A medium-throughput method was designed to characterize the mechanical properties of mesenchymal cell aggregates. This study was the first to present a precise and fast method to determine the Young’s modulus of mesenchymal cell aggregates, utilizing a step-by step aspiration technique. We were also able to recreate conditions that very closely resemble the in vivo environment, where the cells were found to be stretched, and the spheroids are soft and elastic Finally, potential applications of the self-assembled cell aggregates were explored in lung disease study and drug screening, specifically for Idiopathic Pulmonary Fibrosis (IPF). We demonstrated that the cell aggregates from IPF patients show an increase in stiffness, therefore mechanical testing of spheroids is an effective technique to study this disease. It was also found that a novel compound, PF670462, modulates the effect of TGF-beta and inhibits the fibrotic response of IPF cell aggregates. That is, this drug softens IPF spheroids and downregulates fibrogenic gene expression, therefore providing basis for the potential use of PF670462 in IPF treatment.