Thesis (Ph. D. in Biomedical Engineering)--Harvard-MIT Program in Health Sciences and Technology, 2013. / Vita. Cataloged from PDF version of thesis. / Includes bibliographical references (pages 85-97). / Anterior cruciate ligament (ACL) injuries affect over 200,000 Americans yearly, and many occur in young athletes. Current treatment options include tendon autografts and cadaveric allografts. However, these approaches often lead to secondary medical problems, such as donor-site morbidity and immune rejection. Furthermore, in younger patients these grafts fail to grow, leading to additional complications and underlining the need for the development of new approaches that improve the healing and repair of ligaments and tendons. This thesis aims to develop a technique to engineer ACL from autologous mesenchymal stem cells (MSC) and primary ACL fibroblasts using the basic principles of Tissue Engineering. The first part of the thesis characterizes MSCs isolated from tibial bone marrow as an alternative to hip-derived marrow aspirates. The proximity of the tibia to the surgical site of ACL reconstructions makes it a viable source of marrow derived-MSCs for ligament repair, with less stress for the patient and increased flexibility in the operating room. Characterization was performed by fluorescenceactivated cell sorting for MSC-surface markers, and assays to differentiate MSCs towards adipogenic, osteogenic and chondrogenic lineages. The second part of the thesis describes the effects of in vitro co-cultures of ACL fibroblast and MSC on the expression of ligament-associated markers. The goal was to optimize the cell-cell ratio in order to maximize the positive effects of co-cultures on ligament regeneration. Co-cultures of ACL fibroblasts and MSCs were studied for 14 and 28 days in vitro, and the effects assessed with quantitative mRNA expression and immunofluorescence of ligament markers Collagen type I, Collagen type III and Tenascin-C. Finally, based on the enhancing effect observed in co-cultures, the thesis explores a method to regenerate ACL using a three-dimensional polyglyconate scaffold seeded with cell-hydrogel suspensions containing ACL fibroblasts and MSCs. Constructs were analyzed biochemically and by immunofluorescence after 4 weeks in culture with and without mechanical stimulation. Together, our results establish an experimental framework from which a new technique for ACL repair can be developed. The ultimate goal is to foster the design of a one-stage surgical procedure for improved primary ACL augmentation repair that can soon be translated into clinical practice. / by José Antonio Canseco. / Ph.D.in Biomedical Engineering

Thesis (Ph. D.)--Harvard-MIT Division of Health Sciences and Technology, 2011. / Cataloged from PDF version of thesis. / Includes bibliographical references (p. 177-179). / Traditional representations of speech are derived from short-time segments of the signal and result in time-frequency distributions of energy such as the short-time Fourier transform and spectrogram. Speech-signal models of such representations have had utility in a variety of applications such as speech analysis, recognition, and synthesis. Nonetheless, they do not capture spectral, temporal, and joint spectrotemporal energy fluctuations (or "modulations") present in local time-frequency regions of the time-frequency distribution. Inspired by principles from image processing and evidence from auditory neurophysiological models, a variety of twodimensional (2-D) processing techniques have been explored in the literature as alternative representations of speech; however, speech-based models are lacking in this framework. This thesis develops speech-signal models for a particular 2-D processing approach in which 2-D Fourier transforms are computed on local time-frequency regions of the canonical narrowband or wideband spectrogram; we refer to the resulting transformed space as the Grating Compression Transform (GCT). We argue for a 2-D sinusoidal-series amplitude modulation model of speech content in the spectrogram domain that relates to speech production characteristics such as pitch/noise of the source, pitch dynamics, formant structure and dynamics, and offset/onset content. Narrowband- and wideband-based models are shown to exhibit important distinctions in interpretation and oftentimes "dual" behavior. In the transformed GCT space, the modeling results in a novel taxonomy of signal behavior based on the distribution of formant and onset/offset content in the transformed space via source characteristics. Our formulation provides a speech-specific interpretation of the concept of "modulation" in 2-D processing in contrast to existing approaches that have done so either phenomenologically through qualitative analyses and/or implicitly through data-driven machine learning approaches. One implication of the proposed taxonomy is its potential for interpreting transformations of other time-frequency distributions such as the auditory spectrogram which is generally viewed as being "narrowband"/"wideband" in its low/high-frequency regions. The proposed signal model is evaluated in several ways. First, we perform analysis of synthetic speech signals to characterize its properties and limitations. Next, we develop an algorithm for analysis/synthesis of spectrograms using the model and demonstrate its ability to accurately represent real speech content. As an example application, we further apply the models in cochannel speaker separation, exploiting the GCT's ability to distribute speaker-specific content and often recover overlapping information through demodulation and interpolation in the 2-D GCT space. Specifically, in multi-pitch estimation, we demonstrate the GCT's ability to accurately estimate separate and crossing pitch tracks under certain conditions. Finally, we demonstrate the model's ability to separate mixtures of speech signals using both prior and estimated pitch information. Generalization to other speech-signal processing applications is proposed. / by Tianyu Tom Wang. / Ph.D.

Thesis (Ph. D.)--Harvard-MIT Division of Health Sciences and Technology, 2008. / This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. / Includes bibliographical references. / Nature has evolved elegant strategies to temporally and spatially control protein activity, including the synthesis of subunits that require assembly for function, the incorporation of pro-domains that require cleavage for activation, and the formation of inhibitory and stimulatory networks that continually hold enzyme activity in balance. While nanomaterials have been developed with a myriad of functions for sensing or delivering biomolecules, the incorporation of Nature's more sophisticated control strategies in nanoparticle systems has yet to be fully realized. Here, inspired by the biological motifs that control protein function, a series of nanoparticle coatings are developed that utilize bio-inspired strategies for temporal and spatial control, including emergent function through self-assembly, unveiling of latent properties in response to protease activation, and the integration of inhibitory and stimulatory protein interactions. Specifically: 1) Protease removable polymer coatings are developed that veil and unveil complementary iron-oxide nanoparticles, leading to enhanced MRI contrast from protease-activated nanoparticle self-assembly. 2) Prodomain nanoparticles are developed using removable coatings that veil nanoparticle-cell interactions, lengthen particle circulation times, increase tumor accumulation, and unveil nanoparticle surface domains upon cleavage by proteases in the tumor. 3) Removable coatings are adapted to poly (β-amino ester) gene delivery vectors using electrostatically adsorbed peptide linkers, enabling localized gene delivery to cancer cells expressing tumor-associated proteases. 4) Electrostatic coatings for gene delivery vectors are characterized and developed in vivo, leading to the identification of certain peptide-particle formulations that integrate stimulatory and inhibitory blood and cell surface interactions to achieve specific and effective organ-directed delivery. / (cont) 5) Materials used for gene delivery are applied to siRNAs, demonstrating the potential use of these coatings strategies for gene silencing applications. In order to achieve effective cancer therapy nanoparticles must overcome a series of sequential obstacles including undisrupted passage through the blood, accumulation in the tumor, shuttling across intracellular boundaries, and therapeutic delivery inside the cell. In this work we demonstrate how spatially and temporally controlled nanoparticle coatings effectively surmount these multiple barriers to improve imaging and therapy of tumors. / by Todd J. Harris. / Ph.D.

Thesis (Ph. D. in Speech and Hearing Bioscience and Technology)--Harvard-MIT Program in Health Sciences and Technology, 2013. / Cataloged from PDF version of thesis. / Includes bibliographical references. / Laryngeal dystonia (LD) is the focal laryngeal form of the neurological movement disorder called dystonia, a condition that often changes in severity depending on the posture assumed and on voluntary activity of the affected body area. Pathophysiology of dystonia is unknown. This thesis employed a combination of diffusion tensor and functional magnetic resonance imaging (DTI and fMRI) studies to investigate the structure and function of the basal ganglia (BG) in dystonia patients. Fractional anisotropy (FA) and probabilistic diffusion tractography analyses were used to investigate the questions of whether LD patients exhibited altered connectivity between BG and brainstem regions and whether FA and tractography could be used to predict differences in clinical presentations of dystonia. Findings of this study support the hypothesis that connections between the BG and brainstem may play a role in dystonia pathophysiology and may be used to predict differences in clinical presentations of dystonia. An fMRI study was carried out to investigate whether abnormally sustained BG activity observed after performance of a finger tapping task in hand dystonia patients may represent an amplification of a normal motor control mechanism. As dystonia has been hypothesized to result from overactivation of normal postural programs, this study aimed to investigate the question of whether the sustained BG activity was a normal feature observed in motor control tasks requiring more precision. Results suggest that cerebellar cortex is recruited particularly during fine motor control. / by Miriam L. Makhlouf. / Ph.D.in Speech and Hearing Bioscience and Technology

Thesis: Ph. D. in Biomedical Engineering, Harvard-MIT Program in Health Sciences and Technology, 2016. / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 136-153). / The Hepatitis B virus (HBV) has, at one time or another, infected one third of the world's population, and over 300 million people worldwide are chronically infected, leading often to progressive liver damage, cirrhosis, and the development of hepatocellular carcinoma over decades. While an effective vaccine exists, imperfect vaccine penetrance and perinatal transmission result in the continuous establishment of chronic infection in new patients, for whom there is very little chance for a cure. This thesis sought to move closer toward the development of consistently curative treatments for HBV by developing novel model systems to study HBV infection, profiling the host response to this infection to nominate targets for therapeutic intervention, and repurposing novel genome editing tools to directly attack the long-lived viral form that is resistant to cure. First, we use micro-patterning tools and directed differentiation protocols to develop primary adult human hepatocyte and iPS-derived hepatocyte models of HBV infection, respectively, which enable us to test antiviral therapies with multiple mechanisms of action and identify a significant interferon-driven response to HBV infection. Second, we use global transcriptional profiling of this response in multiple human hepatocyte donors to identify candidate genes and pathways that are putatively important for viral infection, and demonstrate significant variability in this infection response between donors and virus sources. We then confirm the importance of nominated virus-regulated genes by showing that pharmacological inhibition of these targets - such as heat shock proteins - restricts viral infection in both primary hepatocytes and cell lines. Finally, we provide a proof of concept for using the CRISPR/Cas9 system to directly attack HBV. We show that carefully selected guide RNAs can direct CRISPR-based inhibition of HBV both in vitro and in vivo. We importantly show that Cas9 can directly cleave and direct degradation of the long-lived, episomal HBV cccDNA, resulting in >1 log-fold reductions in cccDNA in both constitutively HBV-producing and newly HBV infected cells. Overall, this work takes steps toward both elucidating HBV biology, and accelerating the path toward novel treatments for this disease. / by Vyas Ramanan. / Ph. D. in Biomedical Engineering

Thesis (Ph. D. in Medical Engineering)--Harvard-MIT Program in Health Sciences and Technology, 2013. / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 102-113). / The requirement that individual cells be able to communicate with one another over a range of length scales is a fundamental prerequisite for the evolution of multicellular organisms. Often diffusible chemical molecules originate from a source and span the distance between cells in order to establish a line of communication - where the meaning of the signal is a function of both spatial and temporal chemical concentrations. In the case of chemotaxis, cells respond to concentration gradients to establish directionality. In the case of morphogenesis, cells respond to the magnitude of the local concentration field to regulate gene expression. Presented here is an in vitro platform, applicable in the contexts of chemotaxis and morphogenesis, where cells may be exposed to dynamic chemical concentration fields while cultured in a 3-dimensional macromolecular matrix. In the first generation system, cells are exposed to a one-dimensional gradient - constant along the two orthogonal axes. The second-generation system produces two orthogonally oriented gradients intersecting in a 2-dimensional field. These platforms were able to stimulate chemotaxis - both of cultured mammalian cells and emanating from murine skeletal muscle explants. Further, as a developmental tool, we were able to probe the role of Wnt signaling during Sonic Hedgehog based patterning of the vertebrate ventral neural tube. Using the presumptive enhancer for the p3 progenitor domain gene Nkx2.2, our findings indicate that such an enhancer would both negatively and positively regulate Nkx2.2 expression in response to Wnt signaling. However we found that the net effect of positive Wnt signaling - in the context of the cross-repressive interactions between various neural tube transcription factors (Nkx2.2, Olig2, and Pax6) - is inhibition of Nkx2.2 expression and p3 progenitor domain specification. On the basis of our new model, we postulate that the two opposing influences of Wnt on Sonic hedgehog signaling have distinct but dependent functions: first to inhibit Sonic Hedgehog signaling in the dorsal neural tube and secondly to prevent oscillatory behavior at the dorsal p3 boundary. / by Ovid C. Amadi. / Ph.D.in Medical Engineering

Thesis (Ph.D.)--Harvard--Massachusetts Institute of Technology Division of Health Sciences and Technology, 2002. / Includes bibliographical references (leaves 142-154). / Poly-l-lysine and hexadimethrine bromide (polybrene) are cationic polymers which are frequently used to enhance the transduction efficiency of recombinant retroviruses in gene therapy experiments. Conversely, chondroitin sulfate proteoglycan is an anionic polymer endogenously present in retrovirus stocks which inhibits transduction efficiency. An experimental study was performed to determine the mechanisms of retrovirus transduction modulation by these charged polymers, and it was found that they were capable of increasing or decreasing the flux of virus particles onto the cell surface. These effects, and adsorption in general, were independent of the cellular receptor-virus envelope interaction which was believed to provide the driving force for initial virus attachment. In order to consider the feasibility of alternative driving forces for virus attachment, a mathematical model of adsorption was constructed taking into account the electrostatic properties of the system. The model predicted that either cationic polymer-mediated virus aggregation or membrane charge shielding could yield adsorption phenomenon consistent with the previous observations. An experimental study was undertaken to distinguish between these two potential mechanisms, and it was found that both were at work depending on the physicochemical characteristics of the cationic polymer. All cationic polymers were capable of charge shielding, however, only high molecular weight polymers (> 15 kD) could aggregate the virus. Anionic polymers, conversely, were found to inhibit transduction and adsorption via preventing cationic polymers from performing these functions. / by Howard E. Davis, Jr. / Ph.D.

Thesis: Ph. D. in Medical Engineering and Medical Physics, Harvard-MIT Program in Health Sciences and Technology, 2018. / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 94-103). / Invasive electrical brain stimulation has been increasingly used to treat an ever wide range of neuropsychiatric disorders from Parkinson's disease to epilepsy and depression. In addition, single pulse electrical stimulation (SPES) is increasingly used to map connections between cortical areas using cortico-cortical evoked potentials (CCEPs). However, the properties and mechanisms underlying brain stimulation remain mostly unknown, hindering the application of stimulation to new neurological disorders and the development of adaptive stimulation technologies that could improve clinical outcome. To improve understanding of SPES, we systematically explored the effects of cortical electrical stimulation in human epilepsy patients. These patients have intracranial electrodes implanted for intractable epilepsy as part of their clinical course, creating a unique opportunity to simultaneously stimulate and record the human brain in multiple locations. Single pulses of electrical current were delivered across pairs of electrodes in the human cortex, and the neurophysiological responses are recorded. Examining some fundamental properties of CCEPs, we show that the brain's response to less than a millisecond pulse of stimulation can be detected up to one second post-stimulus. This response has two peaks with distinct properties; compared to the second peak, the first is less variable, and its timing is less delayed by distance, while its magnitude is more diminished by distance. Looking at the spatial distribution of CCEPs, we show that stimulation-derived networks are more closely related to structural connectivity than functional connectivity. However, correcting for distance eliminates this difference. Monitoring CCEPs across different brain states, we show that the second peak of the CCEP is significantly diminished during anesthesia. Taken together, these results provide important insight into the basic neurophysiological properties of CCEPs, their spatial distribution, and how they are modulated by the state of the brain itself. These characteristics can inform experimental design, provide input parameters for modeling studies, and be applied towards the development of adaptive closed-loop stimulation paradigms. / by Britni Crocker. / Ph. D. in Medical Engineering and Medical Physics

Thesis (Ph. D.)--Harvard-MIT Program in Health Sciences and Technology, 2012. / 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 (p. 165-[181]). / Humans live in association with trillions of microbes and yet we know remarkably little about their symbiotic relationship. The role these microorganisms have in humans has been characterized only in the case of few bacteria and much less is understood about the dynamic of this relationship. Lately, the mass sequencing efforts accompanying the Human Microbiome Project have begun to uncover the composition of these different microbial niches, and shed light on some the effects they have on their host. The immune system largely determines the composition of bacterial populations living in association with humans. It lights off pathogens while allowing specific bacteria to colonize the body. However, immune system and microbiota appear even more intimately connected than previously imagined. Recent evidence shows that interaction with the associated microbiota is necessary for the proper development of the immune response throughout life. The interface with commensal microbes is notoriously difficult to probe experimentally, due to the diversity of its composition, which makes differentiating the individual ramifications of each associated microbe a much harder task. To understand the complex relationship between the human immune system and microbiome, we need methodologies that can simultaneously probe both in a high throughput fashion, as well as analysis tools to cope with the large amount of resulting data. Herein I present the development of immune mass screening tools capable of comprehensively profiling the antibody-mediated and cell-mediated immune response to microbes. I employ microfluidics techniques to describe the response of single immune cells at high-resolution and in a physiologically relevant environment. I also present the application of machine learning to gut microbiome data and demonstrate how it can be used to differentiate between diseased and healthy individuals in an IBD patient cohort and to allows to deal with the complexity of microbial community data. Moving forward, the goal is to combine these approaches to map how changes in the immune response affects microbiome composition and vice versa. In turn, characterizing this interplay will contribute to our understanding of how bacteria shape our homeostasis and health, facilitating the prediction of which imbalances may lead to disease. / by Eliseo Papa. / Ph.D.

Thesis (Ph. D. in Biomedical Engineering)--Harvard-MIT Program in Health Sciences and Technology, 2013. / Cataloged from PDF version of thesis. / Includes bibliographical references (p. 108-126). / Chemical exchange saturation transfer (CEST) is a relatively new magnetic resonance imaging (MRI) acquisition technique that generates contrast dependent on tissue microenvironment, such as protein concentration and intracellular pH. CEST imaging has the potential to become an important biomarker in a wide range of disorders. As an indicator of tissue pH, CEST imaging may allow the identification of the ischemic penumbra in stroke, and predict chemo- and radiation therapy outcomes in cancer. As a marker of protein concentration, CEST may be able to delineate tumor margins without contrast enhancement, identify disease onset in Alzheimer's disease, and monitor cartilage repair therapies. Despite several promising pilot studies, CEST imaging has had limited clinical application due to two main technical challenges. First, CEST imaging is extremely sensitive to magnetic field inhomogeneity. Images suffer from large susceptibility artifacts unless specialized BO inhomogeneity correction methods are employed that tremendously increase scan time. Second, the CEST contrast cannot be separated from the intrinsic macromolecular magnetization transfer (MT) asymmetry and brain images reflect the MT properties of white and gray matter rather than the desired protein and pH contrast. We have developed a novel CEST imaging acquisition scheme, dubbed saturation with frequency alternating RF irradiation (SAFARI), designed to be insensitive to Bo inhomogeneity and MT asymmetry. Studies in healthy volunteers demonstrate that SAFARI is robust in the presence of BO inhomogeneity and eliminates the need for specialized BO correction, thereby reducing scan time. In addition, results show that SAFARI removes the confounding MT asymmetry. We applied SAFARI imaging towards the study of the saturation transfer contrast in patients with high grade glioma. Results show that the contrast in brain tumors, which was previously attributed to an increase in the CEST signal from amide protons due to an elevated protein concentration, is instead the result of the loss of MT asymmetry found in the normal brain. Therefore, our work has lead to a new understanding of the different sources of signal in saturation transfer images of the brain with important implications for the design and analysis of future CEST studies of brain tumors. / by Rachel Nora Scheidegger. / Ph.D.in Biomedical Engineering