Dissertations / Theses on the topic 'Biomedical engineering'

The inherent design freedom of protein engineering and recombinant protein production enables specific tailoring of protein structure, function, and properties. Two areas of research where protein engineering has allowed for many advances in biomedical materials include the design of novel protein scaffolds for molecular recognition, as well as the use of recombinant proteins for production of next generation biomaterials. The main focus of my dissertation was to develop new biomedical materials using protein engineering. Chapters three and four discuss the engineering of repeat proteins as bio-recognition modules for biomedical sensing and imaging. Chapter three provides an overview of the most recent advances in engineering of repeat proteins in the aforementioned field. Chapter four discusses my contribution to this field. We have designed a de novo repeat protein scaffold based on the consensus sequence of the leucine rich repeat (LRR) domain of the NOD family of cytoplasmic innate immune system receptors. Innate immunity receptors have been described as pattern recognition receptors in that they recognize "global features" of a family of pathogens versus one specific antigen. In mammals, two main protein families of such receptors are: extracellular Toll-like receptors (TLRs) and cytoplasmic Nucletide-binding domain- and Leucine-rich Repeat-containing proteins (NLRs). NLRs are defined by their tripartite domain architecture that contains a C-terminal LRR (Leucine Rich Repeat) domain, the nucleotide-binding oligomerization (NACHT) domain, and the N-terminal effector domain. It is proposed that pathogen sensing in NLRs occurs through ligand binding by the LRR domain. Thus, we hypothesized that LRRs would be suitable for the design of alternative binding scaffolds for use in molecular recognition. The NOD protein family plays a very important role in innate immunity, and consequently serves as a promising scaffold for design of novel recognition motifs. However, engineering of de novo proteins based on the NOD family LRR domain has proven challenging due to problems arising from protein solubility and stability. Consensus sequence design is a protein design tool used to create novel proteins that capture sequence-structure relationships and interactions present in nature in order to create a stable protein scaffold. We implement a consensus sequence design approach to develop proteins based on the LRR domain of NLRs. Using a multiple sequence alignment we analyzed all individual LRRs found in mammalian NLRs. This design resulted in a consensus sequence protein containing two internal repeats and separate N- and C- capping repeats named CLRR2. Using biophysical characterization methods of size exclusion chromatography, circular dichroism, and fluorescence, CLRR2 was found to be a stable, monomeric, and cysteine free scaffold. Additionally, CLRR2, without any affinity maturation, displayed micromolar binding affinity for muramyl dipeptide (MDP), a bacterial cell wall fragment. To our knowledge, this is the first report of direct interaction of a NOD LRR with a physiologically relevant ligand. Furthermore, CLRR2 demonstrated selective recognition to the biologically active stereoisomer of MDP. Results of this study indicate that LRRs are indeed a useful scaffold for development of specific and selective proteins for molecular recognition, creating much potential for future engineering of alternative protein scaffolds for biomedical applications. My second research interest focused on the development of proteins for novel biomaterials. In the past two decades, keratin biomaterials have shown impressive results as scaffolds for tissue engineering, wound healing, and nerve regeneration. In addition to its intrinsic biocompatibility, keratin interacts with specific cell receptors eliciting beneficial biochemical cues, as well as participates in important regulatory functions such as cell migration and proliferation and protein signalling. The aforementioned properties along with keratins' inherent capacity for self-assembly poise it as a promising scaffold for regenerative medicine and tissue engineering applications. However, due to the extraction process used to obtain natural keratin proteins from natural sources, protein damage and formation of by-products that alter network self-assembly and bioactivity often occur as a result of the extensive processing conditions required. Furthermore, natural keratins require exogenous chemistry in order to modify their properties, which greatly limits sequence tunability. Recombinant keratin proteins have the potential to overcome the limitations associated with the use of natural keratins while also maintaining their desired structural and chemical characteristics. Thus, we have used recombinant DNA technology for the production of human hair keratins, keratin 31 (K31) and keratin 81 (K81). The production of recombinant human hair keratins resulted in isolated proteins of the correct sequence and molecular weight determined by sodium dodecyl sulfate polyacrylamide gel electrophoresis and mass spectrometry. Proteins with no unwanted sequence truncations, deletions, or mutations indicate recombinant DNA technology can be used to reliably generate full length keratin proteins. This allows for consistent starting materials with no observable impurities or undesired by-products, which combats a major challenge associated with natural keratins. Additionally, recombinant keratins must maintain the intrinsic propensity for self-assembly found in natural keratins. To test the propensity for self-assembly, we implemented size exclusion chromatography (SEC), dynamic light scattering (DLS), and transmission electron microscopy (TEM) to characterize K31, K81, and an equimolar mixture of K31 and K81. The results of the recombinant protein characterization reveal novel homo-polymerization of K31 and K81, not previously reported, and formation of characteristic keratin fibers for the K31 and K81 mixture. Therefore, recombinant K31 and K81 retain the intrinsic biological activity (i.e. self-assembly) of natural keratin proteins. We have also conducted a comparative study of recombinant and extracted heteropolymer K31/K81. Through solution characterization and TEM analysis it was found that use of the recombinant heteropolymer allows for increased purity of starting material while also maintaining self-assembly properties necessary for functional use in biomaterials design. However, under the processing condition implemented, extracted keratins demonstrated increased efficiency of assembly. Through each study we conclude that recombinant keratin proteins provide a promising solution to overcome the challenges associated with natural protein materials and present an exceptional design platform for generation of new biomaterials for regenerative medicine and tissue engineering.
Ph. D.

Analysis of infant body temperature, environmental temperature and respiratory behaviour has become an important aspect of Sudden Infant Death Syndrome research. The application of engineering techniques as a means of providing research tools has been found to be beneficial for medical research. Signal processing techniques have been developed and applied to the analysis of physiological signals that have been collected from infants in the home environment. These techniques allow physiological signals to be analysed and correlated with the use of both time and frequency domain algorithms. Signals of several days duration are manipulated so they may be easily viewed and studied without the loss of significant information. Parameter evaluation of the fundamental frequencies of periodic signals and statistical parameter estimation of random signals have been employed to tease out trends from within the data. Analysis of physiological signals from sleeping infants has revealed hourly oscillations in their body temperatures that are highly correlated with their breathing rate and breathing rate interquartile range (variability). The oscillations appear to have the highest magnitude when the infant rectal temperatures are near to the mean rectal temperature value. Although some form of relationship between temperature and respiration is evident, insufficient information has been yielded by these signal processing techniques to divulge exactly what the relationship is. A mathematical model of the human thermoregulatory control system has been developed to investigate the behaviour of temperature regulation in infants. The model has been used to test the hypothesis that infant thermal control is inherently unstable. In this model, heat flow through the body tissues is calculated and the effect of bedding on heat loss is also considered. Automatic temperature regulation is achieved by negative feedback control of the metabolic rate, sweat rate and blood flow distribution in the model. Under physiologically normal conditions, the model shows oscillatory behaviour with a period of approximately one hour. Therefore, the model indicates that the temperature oscillations that have been observed in infants in the home environment, may be a direct result of a marginally stable or unstable thermoregulatory control system. The oscillations occurred when the model was operating just below the thermoneutral point. If the mean infant rectal temperature is assumed to be close to the thermoneutral point, then the model behaviour agrees closely with the data collected from infants. Evidence gathered from the behaviour of the thermoregulation model and from the signals collected from infants suggests that thermoregulation may be a dominant control system within the body, therefore, temperature may directly influence respiration. A mathematical model of the human infant respiratory control system has been developed to investigate the effect of body temperature on respiratory system behaviour during sleep and to test the hypothesis that the respiratory system is influenced directly by temperature and indirectly by thermoregulation. A multi-compartment model configuration is used to represent the carbon dioxide and oxygen stores within the body and a controller, sensitive to carbon dioxide and oxygen, adjusts the ventilation rate to complete a negative feedback control loop. Small changes in body temperatures were found to affect the steady state response of the respiratory model while the stability remained relatively unaffected. However, the respiratory model is highly sensitive to small amounts of noise added to blood flows, metabolic rate and arterial gas partial pressures. Therefore, the observed oscillations in infant breathing rate may be a direct effect of thermoregulation while the infant breathing rate interquartile range oscillations are probably induced by another mechanism.

A recently developed electrohydrodynamic direct-write printing method which can be applied to all types of materials and used to create ordered structures and complex patterns using coarse processing needles is described. Utilising co-axial flow of materials has been successful in enabling encapsulated structures to be generated by this technique. Topography is a crucial physical cue in influencing cellular responses and should be considered when designing biomedical architectures. Electrohydrodynamic printing is used in this work to generate ordered topographies with proven biomaterials. By coupling this method with solvent evaporation techniques, desirable scaffold properties can be achieved. These novel areas will offer much greater control over the forming of a plethora of micro- and nano-scaled structures and is essential for topographic studies (e.g. of living cells), novel particle preparation methods, coatings and direct writing of biomaterials. Few studies have evaluated the early stages of cell attachment and migration on the surface of biomaterials, partly due to a lack of suitable techniques. One of the major aims of this study was to use time-lapse microscopy to evaluate the behaviour of fibroblasts cultured with polycaprolactone microfibers and to assess spatially and temporally, the cell-microfiber interaction over a 24 hours period. Ordered polymeric structures were printed onto glass substrates using electrohydrodynamic printing to produce fine microfibers according to a predetermined architecture. Fibroblast attachment and migration was characterized as a function of distance from microfibers. The use of time-lapse microscopy revealed a gradual decrease in cell attachment as the distance from the structures was increased. The technique also revealed interesting cell behaviour once attached to the structures that would otherwise have been missed with standard microscopy techniques. The findings demonstrate time-lapse microscopy is a useful technique for evaluating early stage cell-biomaterial interaction that is capable of recording important events that might otherwise be overlooked.

Low temperature plasma surface alloying with nitrogen (nitriding), carbon (carburising) and both (carbonitriding) has been successfully employed in hardening medical grade ASTM F138, ASTM F1586 and ASTM F2581 as well as engineering grade AISI 316 by the formation of a modified layer better known as S-phase or expanded austenite. In this study, systematic plasma treatments and characterisation were performed on medical grade stainless steel in order to establish the optimised treatment conditions, especially temperature, which can maximise the hardened case depth without any detriment in corrosion resistance. The surface of a biomaterial must not adversely affect its biological environment and return the material surface must not be adversely affected by the surrounding host tissue and fluids. Experimental results have shown that this duality of concern can be addressed by creating S-phase. It has been shown that low-temperature nitriding (430°C), carburising (500°C) and carbonitriding (430°C) improved the localised corrosion, corrosion-wear and fretting-wear resistance of these medical grade stainless. Also biocompatibility studies have proved that these hardened surfaces were biocompatible under the realms of the tests conducted in this study therefore the use of hardened medical grade austenitic stainless steel might be suitable in implant applications.

Surfactant coated microbubbles are widely used as contrast agents (UCA) in medical ultrasound imaging, due to their high echogenicity and non-linear response to acoustic excitation. Controlling the stability of microbubbles in vivo represents a considerable challenge. Understanding the characteristics of the bubble surface and how they change with production method, composition and environment is key to addressing this problem. The aim of this thesis is to investigate viscosity, bubble dissolution, and acoustic response as functions of their composition, manufacturing method and environment. Bubbles were made using combinations of phospholipid and an emulsifier in different molar ratios. Adding the emulsifier decreased both the size and the surface viscosity of the bubbles and caused changes in the scattered pressure amplitude of bubbles under ultrasound. To increase microbubble stability, solid inorganic nanoparticles were adsorbed on to the microbubble surface. These particles behaved as Pickering stabilisers, and deterred Ostwald ripening. The nanoparticles also enhanced the nonlinear behaviour of bubbles at low acoustic pressures. Three manufacturing methods (sonication, cross-flow and flow focusing) were investigated in order to verify stability differences. Sonication produced bubbles with surface viscosities hundreds of centipoise greater than those produced by microfluidics. Both pressure amplitude and harmonic content for sonicated bubbles were found to be much larger due to a higher liposomal adhesion rate at the surface. Solution temperature and bubble age were also investigated. When the solutions were heated above the phospholipid gelling temperature, microfluidic bubbles showed an increased surface viscosity, due to increased liposome adhesion caused by the increased temperature. Bubble composition, manufacturing method and environment were found to vary the surface characteristics of the microbubbles. Further investigations into the affects of the filling gas, in vitro studies, and low temperature TEM characterisation should be conducted to produce a microbubble with the full range of desired characteristics.

Falls are common and burdensome accidents among the elderly. About one third of the population aged 65 years or more experience at least one fall each year. Fall risk assessment is believed to be beneficial for fall prevention. This thesis is about prognostic tools for falls for community-dwelling older adults. We provide an overview of the state of the art. We then take different approaches: we propose a theoretical probabilistic model to investigate some properties of prognostic tools for falls; we present a tool whose parameters were derived from data of the literature; we train and test a data-driven prognostic tool. Finally, we present some preliminary results on prediction of falls through features extracted from wearable inertial sensors. Heterogeneity in validation results are expected from theoretical considerations and are observed from empirical data. Differences in studies design hinder comparability and collaborative research. According to the multifactorial etiology of falls, assessment on multiple risk factors is needed in order to achieve good predictive accuracy.

Falls are common and burdensome accidents among the elderly. About one third of the population aged 65 years or more experience at least one fall each year. Fall risk assessment is believed to be beneficial for fall prevention. This thesis is about prognostic tools for falls for community-dwelling older adults. We provide an overview of the state of the art. We then take different approaches: we propose a theoretical probabilistic model to investigate some properties of prognostic tools for falls; we present a tool whose parameters were derived from data of the literature; we train and test a data-driven prognostic tool. Finally, we present some preliminary results on prediction of falls through features extracted from wearable inertial sensors. Heterogeneity in validation results are expected from theoretical considerations and are observed from empirical data. Differences in studies design hinder comparability and collaborative research. According to the multifactorial etiology of falls, assessment on multiple risk factors is needed in order to achieve good predictive accuracy.

The use of technology in health-care today is increasing dramatically with a corresponding increase in cost and complexity to provide and support it. The degree to which a hospital manages this technology affects its ability to treat patients, to perform research, to teach and to attract competent staff. This thesis project has identified the role that clinical engineering could play in health-care technology provision and support in South Africa. A system synthesis technique was employed to develop an idealized clinical engineering model (ICE) that would satisfy South African technological requirements. An extensive literature survey of the current status of clinical engineering in both developed and developing countries was undertaken to provide input to the synthesis process. Surveys were then conducted to determine the actual current status of clinical engineering and its environment in the RSA. To enable such an idealised department to function as defined, it must be supported by appropriate and timeous information. The information needs of the idealised clinical engineering model were analysed and a corresponding decision support system (DSS) defined. Further surveys were conducted to test the applicability and acceptability of the idealised clinical engineering model. The feasibility of implementing the idealised clinical engineering model in South Africa was investigated and recommendations were made based on the research results of this thesis to bring the actual status of clinical engineering closer to the idealised model. ii

Recently, the interest in implantable devices for biomedical telemetry has significantly increased. Amongst the different components of the implantable device, the antenna plays the most significant role in the wireless data transmission. However, the human body around the antenna alters its overall characteristics and absorbs most of its radiation. Therefore, this thesis is mainly focused on improving the antenna characteristics (bandwidth and radiation efficiency) to overcome the human body effect and investigating new structures that reduce the power absorption by the human body tissues. A novel antenna design methodology is developed and used to design new flexible implantable antennas of much lighter weight, larger radiation efficiency, and wider bandwidth than existing embedded antennas. These antennas work for multiple ((401-406 MHz) MedRadio, 433 MHz and 2.45 GHz ISM) bands which satisfy the requirements of low power consumption and wireless power transfer. This has been combined with thorough investigations of the antenna performance in the anatomical human body. New effective evaluation parameters such as the antenna orientation are investigated for the first time. New structures inspired by complementary and multiple split ring resonators (CSRRs and MSRRs) are designed. The structures are found to reduce the electric near field and hence the absorbed power which increases the radiated power accordingly. This new promising function of metamaterial based structures for implantable applications is investigated for the first time. The path loss (between pacemaker and glucose monitoring implantable antennas inside the anatomical body model) and (between an implantable and external antennas for a wireless power channel at 433 MHz) are estimated. Moreover, the optimum antenna type for on-in body communication is investigated. Loop antennas are found to outperform patch antennas in close proximity to the human body.

This thesis treats biomedical ultrasonics, cavitation and sonoporation. Focussed ultrasound surgery can heat tissue to a temperature that causes protein denaturation and coagulative necrosis. For high-resolution focused ultrasound microsurgery, high working frequencies are necessary. We manufactured a highfrequency, high-intensity focussed ultrasound transducer, using lithium niobate as the active element. The transducer was capable of creating 2.5×3.4 (mm)2 lesions without affecting surrounding tissue. Such disruptive effects of ultrasound also have applications outside medicine. Since cyanobacteria contain gas vesicles, we hypothesised that these can be disrupted with the aid of ultrasound. During 1-hour sonication in the clinical diagnostic range, we forced blue-green algae to sink, thus promoting natural decay. In medical diagnostics, ultrasound contrast agents are added to the blood stream to differentiate between blood and other tissue types. We injected such lipid-shelled microbubbles into a synthetic capillary and sonicated using continuous ultrasound. The microbubbles formed clusters at a quarter wavelength apart owing to radiation forces. We observed cluster coalescence and translation towards the capillary wall. To study acoustic cavitation, we designed and built a scientific instrument combining a pulsed laser and a high-intensity focussed ultrasound transducer, capable of nucleating at precise locations. The cavitation dynamics were recorded using highspeed cameras. At high acoustic intensities, interacting cavitation clouds were formed. Microbubbles under sonication have been observed to create transient pores in adjacent cell membranes. This so called sonoporation has been associated with highly non-linear bubble phenomena. We observed lipid-shelled microbubbles near cancer cells under quasi-continuous low-amplitude sonication. Typically within a second of sonication, microbubbles were seen to enter the cells and dissolve. This new explanation of sonoporation was verified using high-speed photography and confocal fluorescence microscopy. If drug and genes can be successfully coupled to acoustically active vehicles, sonoporation might revolutionise non-invasive therapy as we know it.

This thesis addresses the problem of information hiding in low dimensional digital data focussing on issues of privacy and security in Electronic Patient Health Records (EPHRs). The thesis proposes a new security protocol based on data hiding techniques for EPHRs. This thesis contends that embedding of sensitive patient information inside the EPHR is the most appropriate solution currently available to resolve the issues of security in EPHRs. Watermarking techniques are applied to one-dimensional time series data such as the electroencephalogram (EEG) to show that they add a level of confidence (in terms of privacy and security) in an individual’s diverse bio-profile (the digital fingerprint of an individual’s medical history), ensure belief that the data being analysed does indeed belong to the correct person, and also that it is not being accessed by unauthorised personnel. Embedding information inside single channel biomedical time series data is more difficult than the standard application for images due to the reduced redundancy. A data hiding approach which has an in built capability to protect against illegal data snooping is developed. The capability of this secure method is enhanced by embedding not just a single message but multiple messages into an example one-dimensional EEG signal. Embedding multiple messages of similar characteristics, for example identities of clinicians accessing the medical record helps in creating a log of access while embedding multiple messages of dissimilar characteristics into an EPHR enhances confidence in the use of the EPHR. The novel method of embedding multiple messages of both similar and dissimilar characteristics into a single channel EEG demonstrated in this thesis shows how this embedding of data boosts the implementation and use of the EPHR securely.

The aim of this project was to develop three-dimensional (3-D) constructs of phosphate-based glass fibres (PGF) incorporated dense collagen matrices for biomedical and tissue engineering applications. For this, a novel method of "plastic compression" (PC) was used which rapidly removes fluid from hyper-hydrated collagen gels through the application of unconfined compressive load. The project objectives were: the understanding of structure-property relationship of PGF the understanding of the mechanisms of PC to produce dense collagenous matrices, and the application of PC to produce cellular 3-D constructs of PGF reinforced collagen matrices. PGF are unique glasses as they are degradable and biocompatible, and their degradation can be controlled through their chemistry. Two different quaternary glass systems incorporating CuO and Fe2O3 into the ternary glass system (in molar percentage) 50 P2O5- 30 CaO-20 Na2O were developed for either antibacterial or tissue engineering applications. These additional oxides were incorporated into the glass structure by partially substituting Na20. The rate of degradation was significantly decreased by the incorporation of both oxides possibly due to increased cross-link density, which correlated with an increase in the density and glass transition temperature. There was a further decrease in degradation with increasing fibre diameter. The amount of Cu2+ release increased with increasing CuO content, and 10 mol % was the most effective in killing Staphylococcus epidermidis. YqjOt, had a much more significant effect on rate of degradation, and the rate of Fe3+ release decreased with increasing Fe203 content. From the compositions and fibre diameters investigated, fibres containing 3-5 mol % Fe203 with a diameter of 30 urn were more durable, and therefore suitable for use as scaffolds. Furthermore, upon long term degradation, the iron containing glass systems showed the potential for tube formation. PC depends mainly on the ability of collagen to undergo creep deformation and no recovery upon load removal. Using this principle, a dense collagen matrix with improved mechanical properties was produced. PC was also successful in producing PGF-PC collagen constructs with different compositions. It was anticipated that PGF would initially further enhance the mechanical properties of the constructs. Moreover, PGF also provided the intriguing possibility of capillary-like channels within the collagen for cell and nutrient transportations. The effect of PGF incorporation was assessed morphologically, mechanically, and biologically using live/dead staining. Increasing the proportion of PGF yielded significantly stiffer, stronger constructs while compromising their compliance. At greatest, only 20 % cell death due to either PC or PGF incorporation occurred, however, a significant increase in cell viability after 24 hours was observed. The findings suggested that PC is effective for engineering composite, biomimetic collagen matrices with controllable properties.

Cette étude présente des contributions en traitement statistique du signal avec des applications biomédicales. La thèse est divisée en deux parties. La première partie traite de la détection des hotspots à l'interface des protéines. Les hotspots sont les résidus dont les contributions énergétiques sont les plus importantes dans l'interaction entre protéines. Les forêts aléatoires (Random Forests) sont utilisées pour la classification. Une nouvelle famille de descripteurs de hotspot est également introduite. Ces descripteurs sont basés seulement sur la séquence primaire unidimensionnelle d'acides aminés constituant la protéine. Aucune information sur la structure tridimensionnelle de la protéine ou le complexe n'est nécessaire. Ces descripteurs, capitalisant les caractéristiques fréquentielle des protéines, nous permettent de savoir la façon dont la séquence primaire d'une protéine peut déterminer sa structure tridimensionnelle et sa fonction. Dans la deuxième partie, le RDT (Random Distortion Testing), un test robuste d'hypothèse, est considéré. Son application en détection du signal a montré que le RDT peut résister aux imperfections du modèle d'observation. Nous avons également proposé une extension séquentielle du RDT. Cette extension s'appelle le RDT Séquentiel. Trois problèmes classiques de détection d'écart/distorsion du signal sont reformulés et résolus dans le cadre du RDT. En utilisant le RDT et le RDT Séquentiel, nous étudions la détection d'AutoPEEP (auto-Positive End Expiratory Pressure), une anomalie fréquente en ventilation mécanique. C'est la première étude de ce type dans la littérature. L'extension à la détection d'autres types d'asynchronie est également étudiée et discutée. Ces détecteurs d'AutoPEEP et d'asynchronies sont les éléments principaux de la plateforme de suivi de manière automatique et continue l'interface patient-ventilateur en ventilation mécanique
This PhD thesis presents some contributions to Statistical Signal Processing with applications in biomedical engineering. The thesis is separated into two parts. In the first part, the detection of protein interface hotspots ¿ the residues that play the most important role in protein interaction ¿ is considered in the Machine Learning framework. The Random Forests is used as the classifier. A new family of protein hotspot descriptors is also introduced. These descriptors are based exclusively on the primary one-dimensional amino acid sequence. No information on the three dimensional structure of the protein or the complex is required. These descriptors, capturing the protein frequency characteristics, make it possible to get an insight into how the protein primary sequence can determine its higher structure and its function. In the second part, the RDT (Random Distortion Testing) robust hypothesis testing is considered. Its application to signal detection is shown to be resilient to model mismatch. We propose an extension of RDT in the sequential decision framework, namely Sequential RDT. Three classical signal deviation/distortion detection problems are reformulated and cast into the RDT framework. Using RDT and Sequential RDT, we investigate the detection of AutoPEEP (auto-Positive End Expiratory Pressure), a common ventilatory abnormality during mechanical ventilation. This is the first work of that kind in the state-of-the-art. Extension to the detection of other types of asynchrony is also studied and discussed. These early detectors of AutoPEEP and asynchrony are key elements of an automatic and continuous patient-ventilator interface monitoring framework

Thesis: Ph. D., Massachusetts Institute of Technology, Department of Materials Science and Engineering, 2017.
This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.
Cataloged from student-submitted PDF version of thesis.
Includes bibliographical references (pages 170-176).
Research on biomedical applications of magnetic nanoparticles (MNPs) has increasingly sought to demonstrate noninvasive actuation of cellular processes and material responses using heat dissipated in the presence of an alternating magnetic field (AMF). By modeling the dependence of hysteresis losses on AMF amplitude and constraining AMF conditions to be physiologically suitable, it can be shown that MNPs exhibit uniquely optimal driving conditions that depend on controllable material properties such as magnetic anisotropy, magnetization, and particle volume. "Magnetothermal multiplexing," which relies on selecting materials with substantially distinct optimal AMF conditions, enables the selective heating of different kinds of collocated MNPs by applying different AMF parameters. This effect has the potential to extend the functionality of a variety of emerging techniques with mechanisms that rely on bulk or nanoscale heating of MNPs. Experimental investigations on methods for actuating deep brain stimulation, drug release, and shape memory polymer response are summarized, with discussion of the feasibility and utility of applying magnetothermal multiplexing to similar systems. The possibility of selective heating is motivated by a discussion of various models for heat dissipation by MNPs in AMFs, and then corroborated with experimental calorimetry measurements. A heuristic method for identifying materials and AMF conditions suitable for multiplexing is demonstrated on a set of iron oxide nanoparticles doped with various concentrations of cobalt. Design principles for producing AMFs with high amplitude and ranging in frequency from 15kHz to 2.5MHz are explained in detail, accompanied by a discussion of the outlook for scalability to clinically relevant dimensions. The thesis concludes with a discussion of the state of the field and the broader lessons that can be drawn from the work it describes.
by Michael G. Christiansen.
Ph. D.

Thesis: Ph. D., Massachusetts Institute of Technology, Department of Chemical Engineering, 2014.
Cataloged from PDF version of thesis.
Includes bibliographical references (pages 153-164).
Functional polymeric thin films are often stratified with nanometer level structure and distinct purposes for each layer. These nanostructured polymeric materials are useful in a wide variety of applications including drug delivery, tissue engineering, controlling condensation and polymeric batteries; all of which will be discussed in this work. The first area of my thesis will detail the use of C₆₀ cluster-ion depth profiling X-ray Photoelectron Spectroscopy (XPS) to fundamentally understand how thin film structure and function relate. This method has the unique capability to determine the atomic composition and chemical state of polymeric thin films with <10nm nanometer depth resolution without any chemical labeling or modification. Using this technique, I probed the nanostructure of functional thin films to quantify the interlayer diffusion of the biopolymer chitosan as well as demonstrate methods to stop this diffusion. I also explored the role of interlayer diffusion in the design of hydrophobic yet antifogging 'zwitter-wettable' surfaces. Additionally, I probed the lithium triflate salt distribution in solid block copolymer battery electrolytes (PS-b-POEM) to understand the lithium-ion distribution within the POEM block. In the second area of my thesis, I show how the nanostructure of materials control the function of polymeric particles in vitro and in vivo. One example is a 'Cellular Backpack' which is a flat, anisotropic, stratified polymeric particle that is hundreds of nanometers thick and microns wide. In partnership with the Mitragotri group at UCSB, we show that cellular backpacks are phagocytosis resistant, and when attached to a cell, the cell maintains native functions. These capabilities uniquely position backpacks for cell-mediated therapeutic delivery and we show in vivo that immune cells attached to backpacks maintain their ability to home to sites of inflammation. In addition, we have designed polymeric microtubes that can control their orientation on the surface of living cells. Inspired by chemically non-uniform Janus particles, we designed tube-shaped, chemically non-uniform microparticles with cell-adhesive ligands on the ends of the tubes and a cell-resistant surface on the sides. Our results show that by altering the surface chemistry on the end versus the side, we can control the orientation of tubes on living cells. This advance opens the capability to control phagocytosis and design cellular materials from the bottom up.
by Jonathan Brian Gilbert.
Ph. D.

The growing rate of diabetes and undiagnosed diabetes related diseases is becoming a worldwide major health concern. The motivation of this thesis was to make use of a technology called the ‘electronic nose’ (eNose) for diagnosing diseases. It presents a comprehensive study on metabolic and gastro-intestinal disorders, choosing diabetes as a target disease. Using eNose technology with urinary volatile organic compounds (VOCs) is attractive as it allows non-invasive monitoring of various molecular constituents in urine. Trace gases in urine are linked to metabolic reactions and diseases. Therefore, urinary volatile compounds were used for diagnosis purposes in this thesis. The literature on existing eNose technologies, their pros and cons and applications in biomedical field was thoroughly reviewed, especially in detecting headspace of urine. Since the thesis investigates urinary VOCs, it is important to discover the stability of urine samples and their VOCs in time. It was discovered that urine samples lose their stability and VOCs emission after 9 months. A comprehensive study with 137 diabetic and healthy control urine samples was done to access the capability of commercially available eNose instruments for discrimination between these two groups. Metal oxide gas sensor based commercial eNose (Fox 4000, AlphaMOS Ltd) and field asymmetric ion mobility spectrometer (Lonestar, Owlstone Ltd) were used to analyse volatiles in urinary headspace. Both technologies were able to distinguish both groups with sensitivity and specificity of more than 90%. Then the project moved onto developing a Non-dispersive infrared (NDIR) sensor system that is non-invasive, low-cost, precise, rapid, simple and patient friendly, and can be used at both hospitals and homes. NDIR gas sensing is one of the most widely used optical gas detection techniques. NDIR system was used for diagnosing diabetes and gastro related diseases from patient’s wastes. To the best of the authors’ knowledge, this is the first and only developed tuneable NDIR eNose system. The developed optical eNose is able to scan the whole infrared range between 3.1μm and 10.5 μm with step size of 20 nm. To simulate the effect of background humidity and temperature on the sensor response, a gas test rig system that includes gas mixture, VOC generator, humidity generator and gas analyser was designed to enable the user to have control of gas flow, humidity and temperature. This also helps to find out system’s sensitivity and selectivity. Finally, after evaluating the sensitivity and selectivity of optical eNose, it was tested on simple and complex odours. The results were promising in discriminating the odours. Due to insufficient sample batches received from the hospital, synthetic urine samples were purchased, and diabetic samples were artificially made. The optical eNose was able to successfully separate artificial diabetic samples from non-diabetic ones.

The research conducted in this thesis aims to develop an efficient microwave delivery system employing miniature resonant microwave cavities, targeted at compact, flexible and ideally field-deployable microwave-assisted diagnostic healthcare applications. The system comprises a power amplifier as a solid-state microwave source and a load - as a single mode cavity resonator to hold the sample. The compactness of the practical microwave delivery system relies on the direct integration of the sample-holding cavity resonator to the power amplifier and inclusion of the built-in directional coupler for power measurements. The solid state power transistors used in this research (10W-LDMOS, 10W-GaN) were provided by the sponsoring company NXP Inc. In practical microwave delivery applications, the impedance environment of the cavity resonators change significantly, and this thesis shows how this can be systematically utilized to present the optimal loading conditions to the transistor by simply designing the series delay lines. This load transfer technique, which critically can be achieved without employing bulky, lossy and physically larger output matching networks, allows high performance of the power amplifier to be achieved through waveform engineering at the intrinsic plane of the transistor. Starting with the impedance observation of a rectangular cavity, using only series delay lines allowed the practical demonstration of the high power and high efficiency fully integrated inverse class-F (F-1) power amplifier. Temperature is an important factor in a microwave heating and delivery system as it changes the impedance environment of the cavity resonator. This natural change in both cavity and sample temperature can be accommodated through simplified series matching lines and the microwave heating system capable of working over substantial bandwidth was again practically demonstrated. The inclusion of the coupler maintained the compactness of the system. In the practical situations envisaged, the microwave delivery system needs to accommodate natural variation between sample volumes and consistencies for heating. The experimental work considered the heating of different sample volumes ii of water, and characterizing the change in the natural impedance environment of the cavity as a result. It was shown how the natural impedance variation can not only be accommodated, but also exploited, allowing ‘continuous’, high-efficiency performance to be achieved while processing a wide range of sample volumes. Specifically, using only transistor package parasitic, the impedance of the cavity itself together with a single series microstrip transmission line allows a continuous class-F-1 mode loading condition to be identified. Through different experiments, the microwave delivery systems with high-performance are demonstrated which are compact, flexible and efficient over operational bandwidth of the cavity resonators.

AbstractManual classification of text is both time consuming and expensive. However, it is anecessity within the field of biomedicine, for example to be able to quantify biomedical data.In this study, two different approaches were researched regarding the possibility of usingsmall amounts of training data, in order to create text classification models that are able tounderstand and classify biomedical texts. The study researched whether a specialized modelshould be considered a requirement for this purpose, or if a generic model might suffice. Thetwo models were based on publicly available versions, one specialized to understand Englishbiomedical texts, and the other to understand ordinary Swedish texts. The Swedish modelwas introduced to a new field of texts while the English model had to work on translatedSwedish texts.The results were quite low, but did however indicate that the method with the Swedish modelwas more reliable, performing almost twice as well as the English correspondence. The studyconcluded that there was potential in using general models as a base, and then tuning theminto more specialized fields, even with small amounts of data.KeywordsNLP, text-mining, biomedical texts, classification, labelling, models, BERT, machinelearning, FIC, ICF.
Sammanfattning Manuell klassificering av text är tidskonsumerande och kostsamt, däremot är det en nödvändighet inom exempelvis biomedicinska områden för att kunna kvantifierabehandlingen av data. I denna studie undersöktes två alternativa sätt att utan tillgång till stora mängder data, kunna framställa textklassificeringsmodeller som kan förstå och klassificerabiomedicinsk text. Studien undersökte ifall om en specialiserad modell borde anses som ettkrav för detta, eller ifall om en generisk modell kan räcka till. Båda modellerna som användesvar baserade på allmänt tillgängliga versioner, en som var tränad att förstå engelskbiomedicinsk text och en annan som var tränad att förstå vanlig svensk text. Den svenskamodellen introducerades till ett nytt område av text medan den engelska modellen arbetade påöversatta svenska texter. Resultatet visade att den svenska modellen kunde förstå och klassificera texten nästan dubbeltså effektivt som den engelska, däremot med en relativt låg grad av träffsäkerhet. Slutligenkunde slutsatsen dras att den använda metoden visade potential vid träning av modeller, ochvid brist på större datamängder borde generellt tränade modeller kunna nyttjas som bas för attsedan kunna specialiseras till andra områden. Nyckelord NLP, textbrytning, biomedicinska texter, klassificering, märkning, modeller, BERT,maskininlärning, FIC, ICF.

Three-dimensional (3D) bioprinting is a rapidly growing field that is particularly well suited for “bottom up” tissue engineering, largely due to its ability to control the 3D shape of the engineered construct, as well as its constituents (e.g., cells and/or material) and their spatial distribution. A variety of nozzle-based techniques have emerged for tissue engineering, and while these excel at building large 3D architectures, they suffer from moderate print resolution and limited printable materials, making them less attractive for smaller, high-resolution constructs. This is due in part to shearing effects and clogging of the nozzle. Thus, alternative printing methods are needed to create smaller constructs requiring high-spatial pattern resolution and size control.

Our laboratory has previously developed a laser-based biofabrication platform, gelatin-based laser direct-write (LDW) as a technique for bioprinting highly viable cells with spatial resolution unmatched by other printing techniques in 2D. In this thesis, a novel single-step technique was developed to extend this platform to fabricate and spatially pattern 3D alginate microbeads. With this new method, we demonstrate excellent size-control of fabricated microbeads by manipulating the beam diameter used for deposition. We further show that deposited beads have excellent pattern registry, and cells within LDW microbeads maintain high cellular viability. Additionally, we demonstrate that this technique is compatible with our laboratory’s 2D laser direct-writing of cells, illustrating the ability to fabricate spatially-precise, hybrid, 2D/3D cultures of cells and cell-loaded microbeads. Within cellular applications, the mechanical properties of the extracellular matrix have become an important feature for regulating behavior. To further develop our control over the cellular microenvironment, we demonstrate our ability to mechanically tune the stiffness of LDW-printed microbeads, by varying the crosslinking divalent cation and cation concentration used in the LDW microbead fabrication process. Microbead mechanical properties were determined using large printed arrays of microbeads (12 × 12 array) to amplify the resistance generated during traditional compression testing. Using this method, we demonstrated microbead mechanical properties could be tuned by adjusting fabrication and crosslinking parameters, to achieve a wide range of elastic moduli, from physiologic to pathologic values. While this was a valuable step to demonstrate our ability to control aspects of the engineered cellular microenvironment, our alginate structures were still largely limited for cellular interaction due to the lack of adhesion ligands. The inability for cells to interact with the alginate prevents migration within the matrix.

To overcome the limitations of the inert alginate of our microbeads, we used an established materials processing approach to produce core-shelled microcapsules. This technique consists of coating the printed microbead with a positively charged polymer (e.g., chitosan or poly-L-lysine), to produce a polyelectrolyte membrane around the bead, then chelating the calcium crosslinking the interior. This resulted in a polymeric shell with an aqueous core entrapping the cellular payload. We found that core-shelled microcapsules from LDW microbeads maintained their pattern fidelity through processing, and encapsulated cells retained high viability. Cancer cells and stem cells encapsulated within these structures were observed to self-assemble to form size-controlled 3D aggregates; tumor spheroids and embryoid bodies, respectively.

In addition to creating conventional core-shelled microcapsules, we demonstrate that LDW’s spatial precision can be leveraged to produce advanced core-shelled structures of customizable planar geometries, by utilizing single microbeads as voxels, and patterning these in overlapping arrays. Using this technique, we were able to create custom geometries, such as microstrands, bifurcations, rectangular mats, and rings, wherein aggregating cells self-assembled to make continuous three-dimensional aggregates that conform to the shape of the structure. Overall, this doctoral thesis research developed a powerful, laser-based method for engineering custom 3D microenvironments, with applications in tumor modeling and regenerative medicine. These advances hold great promise for fabricating the next generation in vitro diagnostics.

To date, only engineered tissues of planar geometry, such as epidermal and dermal layer substitutes, have successfully reached the market, mainly due to their relative low complexity and simple geometry. In contrast, the mechanical and functional requirements of tubular tissues are more stringent compared to planar tissues. Tubular tissues, which are the main components of several biological systems (e.g. circulatory, urinary or respiratory), not only present an increased complexity in geometry and tissue architecture, they are also populated by mixed cell types. In addition, these are continuously exposed to cyclic mechanical stimuli, which modulate cellular responses and ultimately the functionality of the tissues. Therefore, the understanding and the ability to reproduce physiologically equivalent environments are critical to generate mechanically and biologically functional neo-tissues or tissue models. The aim of this doctoral research was to produce and characterize 3D DC-based tubular constructs as tissue models for airway tissue engineering in physiologically relevant culture conditions. The first objective was to develop DC-based constructs and evaluate, in real-time, the responses of seeded fibroblasts to PC and to culturing with the DC environment; the fabrication and characterization of mesenchymal stem cell (MSC) seeded multilayered DC-SF-DC hybrids; and to evaluate the differentiation of MSCs cultured within multilayered DC-SF-DC hybrids.The second objective was to develop and characterize cell-seeded tubular dense collagen constructs (TDCCs) with bioinspired mechanical properties.The third objective was to implement tubular dense collagen-based constructs as an airway tissue model through the evaluation of airway smooth muscle cell (ASMC) responses within TDCC under physiological mechanical stimuli, and the development of a multilayered tubular dense collagen-silk fibroin construct (TDC-SFC) that mimicked airway tract architecture in order to study MSC responses under physiological mechanical stimulation.By providing ASMCs with a physiologically equivalent niche, and through pulsatile flow stimulation, in vitro, ASMCs exhibited their native orientation, maintained their contractile phenotype and enhanced the mechanical properties of the TDCC through matrix remodelling. The ability of TDC-SFC to transfer physiological pulsatile stimulation to resident MSCs resulted in native-like cell orientation (i.e. parallel to circumferential strain), and induced MSC contractile phenotype expression.In conclusion, the tubular dense collagen-based constructs developed and implemented, in this doctoral dissertation, effectively provided an in vitro airway tissue model for potential preclinical studies to mimic physiological and pathological conditions (e.g. inflammatory and degenerative diseases) in a relevant biomechanical environment, as alternatives to simple tissue culture techniques or complex animal models.
À ce jour, seuls les tissus synthétisés de forme plane, comme les substituts dermiques et épidermiques, ont réussi à percer le marché, surtout en raison de leur complexité relativement faible et de leur géométrie simple. À l'opposé, les exigences mécaniques et fonctionnelles des tissus tubulaires imposent un plus grand nombre de contraintes que les tissus planaires. Principales composantes de plusieurs systèmes biologiques (circulatoire, urinaire ou respiratoire), les tissus tubulaires sont non seulement plus complexes sur le plan de la géométrie et de l'architecture tissulaire, mais ils sont aussi composés de cellules de différents types. De plus, ils sont continuellement exposés à des stimuli mécaniques cycliques. Voilà pourquoi il est essentiel de comprendre les milieux physiologiquement équivalents et de pouvoir les reproduire si on veut obtenir des néotissus ou des modèles tissulaires fonctionnels sur le plan mécanique et biologique.La présente recherche de doctorat visait donc à produire et à caractériser des constructions tubulaires 3D à base de CD, les tissus des voies respiratoires dans des conditions de culture physiologiquement pertinentes. Le premier objectif était de concevoir des constructions à base de CD et d'évaluer la réaction des fibroblastes ensemencés à la CP et à la culture dans un milieu à base de CD; de fabriquer et de caractériser des hybrides multicouches CD-fibroïne-CD ensemencés de cellules souches mésenchymateuses (CSM); et d'évaluer la différenciation.Le deuxième objectif de la présente recherche était de concevoir et de caractériser des constructions tubulaires faites de collagène dense (CTCD). Le troisième objectif était d'implanter des constructions tubulaires à base de CD comme modèle tissulaire des voies respiratoires par l'évaluation de la réponse des cellules musculaires lisses (CML) des voies respiratoires dans les CTCD en présence de stimuli mécaniques physiologiques.En leur fournissant une niche physiologiquement équivalente, et grâce à la stimulation de l'écoulement pulsatoire, in vitro, les CML des voies respiratoires ont pris leur orientation naturelle, maintenu leur phénotype contractile et amélioré les propriétés mécaniques de la CTCD grâce au remodelage matriciel. La capacité de la CTCD à transférer la stimulation physiologique pulsatile aux CSM résidentes a donné une orientation des cellules s'apparentant à leur orientation naturelle et induit l'expression phénotypique.En conclusion, les constructions tubulaires à base de collagène dense qui ont été développées et implantées sont parvenues à fournir in vitro un modèle tissulaire des voies respiratoires pour d'éventuelles études précliniques visant à reproduire les conditions physiologiques et pathologiques.

A new optical method, for the assessment of the optic nerve-head topography, was studied theoretically and experimentally. The motivation of this work is in relation with glaucoma, an eye disease that can lead to blindness. Glaucoma manifests itself by a destruction (atrophy) of the retinal axons, which in turn affects the shape (topography) of the optic nerve head (disc). Topographic measurement of the optic nerve head is used for diagnosis and tracking (progression) of glaucoma. The new technique consists of a dual-fringe projection system and a single-view imaging system. The dual-fringe projection system produces a 3-D irradiance distribution in the measurement volume (optic nerve location) and the modulation of the fringes is used to profile the optic nerve head. Since the optic nerve disc is optically accessible only through the ocular media, the new method is conceived to access the optic nerve through the eye pupil and to minimize the effects of ocular variations such as the optical power and third order aberrations of the eye. The method can use a low power broad-band source and is therefore safe for the eye. Two prototypes were built (coherent and incoherent source) and evaluated using an optical system that simulates the eye. The new method can be applied to the topographic measurement of any diffuse object, whether it is directly accessible or not.

This study reports the design, development, and characterization of 85/15 poly (dl-lactide-co-glycolide) acid (PLGA 85/15) scaffolds for tissue engineering applications. In this respect the effects of different processing parameters on the PLGA 85/15 scaffold's physical and mechanical properties were investigated. Porous PLGA 85/15 scaffolds were prepared using a gas foaming/salt leaching technique. The processing parameters under examination included gas saturation pressure, gas saturation time, and NaCl/polymer mass ratio. The physical properties of the scaffold considered were the density, the porosity, the average pore size, and the pore density. The mechanical property studied was the Young's modulus in compression. The results demonstrated that all the processing parameters worked in concert to produce a scaffold with a high level of interconnectivity. From the parameteric study, key processing parameters were identified and selected parameters led to optimal physical and mechanical properties of the scaffold. The experimental results obtained from the mechanical properties of scaffolds were compared with a theoretical model from Gibson and Ashby relating the scaffold's mechanical properties to the density. The predicted elastic responses of opened pores structure from the theoretical model showed agreement with the experimental results. Subsequently, the effect of the optimized PLGA 85/15 scaffolds on the cell growth and the cell viability of the human promyelocytic leukemia cell line (HL-60) were reported. The investigation showed that the cell growth and viability were not impaired by the presence of PLGA 85/15 scaffolds for the time period under investigation. Finally, the effects of different degradation media on the optimized PLGA 85/15 scaffold's physical and mechanical properties were also elucidated. The three different media were distilled water (dH2O), a phosphate buffered saline (PBS) solution, and HL-60 cells. In general, the average macropore size and the average molecular weight decreased as the degradation time increased in each medium. However, the scaffolds maintained mechanical and structural integrity throughout the study in all three media over the degradation period studied. Overall, PBS solution most strongly affected physical and mechanical properties, followed by dH2O and HL-60 cells. The distinct variations of the scaffold's properties using different media, demonstrated the importance of carefully selecting the medium to perform in vitro studies. The medium must replicate the actual environment where the scaffold would be used, in order to represent accurately the changes in properties that the scaffold would be undergoing.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2010.
Cataloged from PDF version of thesis.
Includes bibliographical references (p. 180-190).
Advances in mobile electronics are fueling new possibilities in a variety of applications, one of which is ambulatory medical monitoring with body-worn or implanted sensors. Digital processors on such sensors serve to analyze signals in real-time and extract key features for transmission or storage. To support diverse and evolving applications, the processor should be flexible, and to extend sensor operating lifetime, the processor should be energy-efficient. This thesis focuses on architectures and circuits for low power biomedical signal processing. A general-purpose processor is extended with custom hardware accelerators to reduce the cycle count and energy for common tasks, including FIR and median filtering as well as computing FFTs and mathematical functions. Improvements to classic architectures are proposed to reduce power and improve versatility: an FFT accelerator demonstrates a new control scheme to reduce datapath switching activity, and a modified CORDIC engine features increased input range and decreased quantization error over conventional designs. At the system level, the addition of accelerators increases leakage power and bus loading; strategies to mitigate these costs are analyzed in this thesis. A key strategy for improving energy efficiency is to aggressively scale the power supply voltage according to application performance demands. However, increased sensitivity to variation at low voltages must be mitigated in logic and SRAM design. For logic circuits, a design flow and a hold time verification methodology addressing local variation are proposed and demonstrated in a 65nm microcontroller functioning at 0.3V. For SRAMs, a model for the weak-cell read current is presented for near-V supply voltages, and a self-timed scheme for reducing internal bus glitches is employed with low leakage overhead. The above techniques are demonstrated in a 0.5-1.OV biomedical signal processing platform in 0.13p-Lm CMOS. The use of accelerators for key signal processing enabled greater than 10x energy reduction in two complete EEG and EKG analysis applications, as compared to implementations on a conventional processor.
by Joyce Y. S. Kwong.
Ph.D.

Aromatic and aliphatic diacid chlorides were used to condense naturally occurring diamino acids and their esterified derivatives. It was anticipated the resulting functional polyamides would biodegrade to physiologically acceptable compounds and show pH dependant solubility could be used for biomedical applications ranging from enteric coatings to hydrosoluble drug delivery vehicles capable of targeting areas of low physiological pH. With these applications in mind the polymers were characterised by infra red spectroscopy, gel permeation chromatography and in the case of aqueous soluble polymers by potentiometric titration. Thin films of poly (lysine ethyl ester isophthalamide) plasticised with poly (caprolactone) were cast from DMSO/chloroform solutions and their mechanical properties measured on a Hounsfield Hti tensiometer. Interfacial synthesis was investigated as a synthetic route for the production of linear functional polyamides. High molecular weight polymer was obtained only when esterified diamino acids were condensed with aromatic diacid chlorides. The method was unsuitable for the production of copolymers of free and esterified amino acids with a diacid chloride. A novel miscible mixed solvent single phase reaction was investigated for production of copolymers of esterified and non-esterified amino acids with diacid chlorides. Aliphatic diacid chlorides were unsuitable for condensing diamino acids using this technique because of high rates of hydrolysis. The technique gave high molecular weight homopolymers from esterified diamino acids and aromatic diacid chlorides.

Anterior cruciate ligament (ACL) injuries are one of most common musculoskeletal injuries and negatively affect mobility and quality of life. ACL rupture requires reconstruction to repair ligament at an estimated cost of $1.5 billion/year. Current surgical solutions invariably involve either donor site morbidity with the use of autografts or the risk of disease transmission and immune rejection with the use of allografts. Successful reconstruction requires the presence of an intact interface between ligament and bone, a transitional tissue called the enthesis. The enthesis is critical for the safe and effective transfer of force from the stiff bone to the more compliant ligament by providing a gradual transition of mechanical and biochemical properties to prevent the formation of stress concentrations. A tissue engineered ligament containing mature entheses is a promising alternative to autografts and allografts, especially since this interface does not normally regenerate. Toward this end, this dissertation sought to improve engineered fibrin-based bone-to-bone ligaments previously developed by our lab and to demonstrate their utility in understanding physiological processes through three specific aims: 1) optimize the environment for in vitro ligament function, 2) induce the formation of a fibrocartilaginous interface, and 3) demonstrate the utility of engineered ligaments as a physiological model.

In Aim 1, the in vitro culture environment was investigated for engineered ligaments formed using human ACL fibroblasts. Using a DOE approach, we identified significant effects and interactions of soluble factors on the maximal tensile load (MTL) and collagen content of engineered human ACL. The DOE model was used to predict a maximal growth media which significantly improved the MTL and collagen content of engineered ligaments and can be combined with increases in the initial construct volume for 77% further improvement in MTL. In addition to the improvements in tissue function, these data suggest that a DOE approach can more efficiently optimize in vitro parameters including the dosage and timing of chemical and mechanical stimuli as well as any interactions.

Aim 2 presented two strategies to improve of the engineered enthesis. First, the local release of bone morphogenetic protein (BMP)-4 at the enthesis of engineered ligaments demonstrated improved interface strength as well as the transition of cells at the enthesis towards an unmineralized fibrocartilage phenotype. Second, engineered ligaments formed in a modular fashion improved the mechanical function and the morphology of the engineered enthesis including the development of cell and soft tissue integration into the mineral phase, a tidemark between mineralized and unmineralized tissue, and the presence of a dense band of extracellular matrix (ECM) at the soft tissue-mineral interface. Importantly, this is the first demonstration of the in vitro formation of a functional interface between engineered ligament and mineral in a complete bone-to-bone ligament unit.

Aim 3 demonstrated the use of our engineered ligament model as a physiological tool. During the estrogen surge in the menstrual cycle, there is an associated increase in the incidence of ACL ruptures as well as knee laxity. Using physiological levels of estrogen mimicking the estrogen surge in vitro, we determined that estrogen decreases the activity of the collagen crosslinker lysyl oxidase (LOX) with a subsequent decrease in tissue stiffness providing insight into why women have greater incidences of ACL rupture. We also examined the role of the exercise-induced biochemical environment on connective tissue using our in vitro model. Engineered ligaments cultured with serum obtained from human donors after exercise had significantly better mechanical strength and collagen content than those treated with serum obtained at rest. In 2D culture, we determined that this effect was likely a result of greater mTOR and ERK signaling.

In summary, the work in this dissertation has made great strides in developing a more mature engineered bone-to-bone ligament. We have optimized a growth factor environment for their in vitro culture and created the most advanced engineered enthesis to date. We have also used these engineered tissues as a platform to mechanistically study the influence of hormones on connective tissue. With further advances in our understanding of the in vivo development of ligaments and their entheses, our bone-to-bone engineered ligaments can be improved making them more suited for clinical applications and for probing physiologically processes in a more controlled environment.

Mathematical and computational modeling allow for the rationalization of complex phenomenon observed in our reality. Through the careful selection of assumptions, the intractable task of simulating reality can be reduced to the simulation of a practical system whose behavior can be replicated. The development of computational models allow for the full comprehension of the defined system, and the model itself can be used to evaluate the results of thousands of simulate experiments to aid in the rational design process.

Biomedical engineering is the application of engineering principles to the field of medicine and biology. This discipline is composed of numerous diverse subdisciplines that span from genetic engineering to biomechanics. Each of these subdisciplines is concerned with its own complex and seemingly chaotic systems, whose behavior is difficult to characterize. The development and application of computational modeling to rationalize these systems is often necessary in this field and will be the focus of this thesis.

This thesis is centered on the development and application of mathematical and computational modeling in three diverse systems in biomedical engineering. First, computational modeling is employed to investigate the behavior of key proteins in the post-synapse centered around learning and memory. Second, computational modeling is utilized to characterize the drug release rate from implantable drug delivery depots, and produce a tool to aid in the tailoring of the release rate. Finally, computational modeling is utilized to understand the motion of particles through an inertial focusing microfluidics chip and optimize the size selective capture efficiency.

碩士
國立成功大學
機械工程學系碩博士班
92
Cuff electrode is an indispensable component in a neural prostheses system. It is used to stimulate the motor neuron and to record neural signals from the peripheral nerve. It is known that a pressure larger than 20mmHg has a harmful effect to the nerve trunk. Measuring the interface pressure between the cuff and the nerve trunk provides a means to monitor the health of nerve trunk. Therefore, the goals of this study are twofold: first, to improve the measurement performance of cuff electrodes and second, to develop a micro capacitive pressure sensor embedded into the cuff electrode. 　　In current study, the fabrication of a cuff electrode with a convex structure on the surface of gold electrode is suggested by using heat treatment. However, experiment result revealed that there is no significant difference in measurement performance between the electrodes with either the convex surface or the concave surface. In addition, a micro capacitive pressure sensor is successfully developed and tested. The dimension of the sensing electrode measured 7000�慆×7000�慆; the range of pressure measurement was from 0 to 50 mmHg; and the sensitivity was 0.01 pF/mmHg. It is found design the Poisson’s ratio and hysteresis properties of the dielectric layer has significant effect on sensor design and a thickness of the insulated layer affects the initial capacitance of the pressure sensor.