Thesis (Ph. D.)--Harvard-MIT Division of Health Sciences and Technology, 2005. / Includes bibliographical references (p. 142-146). / This study focuses on the acoustic patterns of stop consonants and adjacent vowels as they develop in young children (ages 2;6-3;3) over a six month period. Speech is generated using a series of articulatory, laryngeal, and respiratory gestures that children must learn to reproduce. As a child's speech develops, the gestures become more precise and coordinated, and the resulting acoustic patterns are refined. To explore their development, over forty different acoustic measurements were made on each of 1049 recorded utterances from ten children, including durational, amplitude, spectral, formant, and harmonic measurements. These acoustic data are interpreted in terms of the supraglottal, laryngeal, and respiratory actions that give rise to them. Data show that some details of the child's gestures are still far from achieving the adult pattern. Children have acquired appropriate positioning of their primary articulator for producing a stop consonant, but are still learning to adjust the tongue body during the consonant production. At constriction release, children have a high incidence of multiple bursts and a short burst duration, interpreted as a reflection of increased articulator compliance, smaller articulator size, and high subglottal pressure. Children are also still acquiring correct adjustment of vocal fold stiffness and glottal spreading as well as intraoral pressure, as evidenced by long voice onset times and highly variable fundamental frequencies. Additionally, amplitude changes over the course of the utterance and high amplitude variability reveal that children have not yet gained full control over subglottal pressure. / (cont.) Overall, results indicate that children are less consistent than adults in controlling and coordinating various gestures and with finding the ideal respiration and vocal tract postures, including the stiffness of their articulators. Certain aspects of child speech are found to become more similar to adult values over the six month period of the study. / by Annika Karin Karlsson Imbrie. / Ph.D.

Thesis (Ph. D.)--Harvard-MIT Division of Health Sciences and Technology, 2007. / Includes bibliographical references (p. 223-237). / The ear represents a remarkable achievement in sensory physiology. It is very fast (timescales on the order of 1-100 kHz), has a large bandwidth (-10 octaves in the human), highly sensitive (threshold is ultimately determined by thermal noise), operates over an enormous dynamic range (factor of a trillion in terms of energy input), capable of sharp stimulus selectivity (e.g. frequency and intensity discrimination) and exhibits robust nonlinear behavior. As a sensor designed to detect acoustic sound pressure, surprisingly, the ear also emits sound as well. These otoacoustic emissions (OAEs) have been developed extensively for clinical applications (healthy ears emits while impaired ones do not), though their full potential has yet to be realized. Much of the effort gone into understanding OAEs has been developed within the context of mammals, where specific anatomical and physiological features (e.g. traveling waves and somatic motility) are thought to play an integral role in generation. This thesis approaches OAEs comparatively, systematically characterizing emissions in humans and an array of non-mammals (chickens, geckos and frogs) who lack these mammalian features and exhibit a diverse range of morphologies. / (cont.) First, our results show that for a fixed set of stimulus conditions (employing moderate intensities), emissions are relatively largest in the gecko and frog (the two species with the fewest number of sensory cells) and smallest in the human and chicken for frequencies below -2 kHz. At higher frequencies (3-5 kHz), emissions descend toward the noise floor for the non-mammals but remain relatively constant in human. Second, OAE phase behavior indicates that emissions are generated by multiple mechanisms in the human and chicken (and possibly gecko in certain stimulus conditions), but not the frog. OAEs in all species exhibit significant delays (-1 ms or longer), those being largest in humans. Tuning can explain these delays in all species except the frog, where some additional source of delay is present. Lastly, non-monotonic growth (relative to stimulus intensity) was found in all species, suggesting the presence of multiple level-dependent sources. We interpret the observed similarities and differences in emission properties across species within the context of anatomical/physiological comparisons. / by Christopher Bergevin. / Ph.D.

Thesis (Ph. D.)--Harvard-MIT Division of Health Sciences and Technology, 2008. / Includes bibliographical references (p. 134-139). / In multicellular organisms, cells do not exist in isolation but communicate with other cells via extracellular signaling molecules, many of which diffuse into the microenvironment. More than most cell types, embryonic stem cells (ESCs) are critically sensitive to their microenvironment, which plays an important role in determining whether an ESC will self-renew or differentiate. Although conventional methods exist for controlling the soluble microenvironment, they are able to exert only limited control over diffusible signaling. Especially in conventional static culture, the content of the cell culture media changes constantly over time as cells interact with and modify their surroundings. This thesis explores the use of microfluidic perfusion as a tool for modulating diffusible cell-cell signaling in mESC culture, thus enabling more control over the soluble microenvironment over time. Non-recirculating microfluidic perfusion culture can effect a more defined microenvironment by continuously controlling the supply and removal of soluble factors, with minimal use of expensive reagents. We describe development of the first successful protocol for culturing mouse ESCs in microfluidic perfusion over several days. To optimize flow-rate conditions such that proliferation is achieved while avoiding nutrient deprivation and high shear stress regimes, we developed novel logarithmic flow-rate devices for characterizing mESC behavior across a wide range of flow rates simultaneously. We observed both flow-rate and location-dependent proliferation and investigated the role of glucose depletion in generating these effects. Finally, we demonstrate that perfusion culture can significantly affect diffusible cell-cell signaling in the soluble microenvironment, and thus the cell's biological state. / (cont.) We first show that typical flow rates are able to remove secreted factors by collecting leukemia inhibitory factor (LIF) from the output of microfluidic ESC cultures perfused with LIF-free media. We then show that two well-established serum-free media: N2B27 (neuronal differentiation media) and N2B27 +LIF+BMP4 (selfrenewal media) are not sufficient for successful perfusion culture of mESCs at typical densities, although they are able to support cultures in static conditions. These results suggest the presence of autocrine/paracrine loops that support ESC propagation in serum-free media when cultured under static conditions. / by lily Y. Kim. / Ph.D.

Thesis (Ph. D.)--Harvard-MIT Division of Health Sciences and Technology, 2008. / Includes bibliographical references (p. 211-215). / Regions of phonation exhibiting nonmodal characteristics are likely to contain information about speaker identity, language, dialect, and vocal-fold health. As a basis for testing such dependencies, we develop a representation of patterns in the relative timing and height of nonmodal glottal pulses. To extract the timing and height of candidate pulses, we investigate a variety of inverse-filtering schemes including maximum-entropy deconvolution that minimizes predictability of a signal and minimum-entropy deconvolution that maximizes pulse-likeness. Hybrid formulations of these methods are also considered. we then derive a theoretical framework for understanding frequency- and time-domain properties of a pulse sequence, a process that sheds light on the transformation of nonmodal pulse trains into useful parameters. In the frequency domain, we introduce the first comprehensive mathematical derivation of the effect of deterministic and stochastic source perturbation on the short-time spectrum. We also propose a pitch representation of nonmodality that provides an alternative viewpoint on the frequency content that does not rely on Fourier bases. In developing time-domain properties, we use projected low-dimensional histograms of feature vectors derived from pulse timing and height parameters. For these features, we have found clusters of distinct pulse patterns, reflecting a wide variety of glottal-pulse phenomena including near-modal phonation, shimmer and jitter, diplophonia and triplophonia, and aperiodicity. Using temporal relationships between successive feature vectors, an algorithm by which to separate these different classes of glottal-pulse characteristics has also been developed. / (cont.) We have used our glottal-pulse-pattern representation to automatically test for one signal dependency: speaker dependence of glottal-pulse sequences. This choice is motivated by differences observed between talkers in our separated feature space. Using an automatic speaker verification experiment, we investigate tradeoffs in speaker dependency for short-time pulse patterns, reflecting local irregularity, as well as long-time patterns related to higher-level cyclic variations. Results, using speakers with a broad array of modal and nonmodal behaviors, indicate a high accuracy in speaker recognition performance, complementary to the use of conventional mel-cepstral features. These results suggest that there is rich structure to the source excitation that provides information about a particular speaker's identity. / by Nicolas Malyska. / Ph.D.

Thesis (Ph.D.)--Harvard--Massachusetts Institute of Technology Division of Health Sciences and Technology, February 1999. / "January 1999." / Includes bibliographical references (leaves 219-239). / by Steven Treppo. / Ph.D.

Thesis (S.M.)--Harvard-MIT Division of Health Sciences and Technology, 2009. / Vita. Cataloged from PDF version of thesis. / Includes bibliographical references (p. 94-100). / This thesis investigates the economic impact of stratified medicine on industry and the subsequent ethical implications for patients. Stratified medicine involves the use of clinical biomarkers to indicate differential response among patients in efficacy or potential side effects of therapeutic agents. The advent of stratified medicine should, in theory, result in the safer, more effective use of therapeutic agents to treat cancer. However, reluctance remains within the broader life sciences community, in particular within the pharmaceutical industry, to embrace stratified medicine. I hypothesize that this is due to economic concerns. Firstly, an historical analysis of the rate of market adoption of stratified therapeutics is conducted by comparing the adoption velocity and time to peak sales of stratified therapeutics relative to traditional chemotherapeutics. The aim is to analyze whether historically, stratified medicines have been more or less successful in terms of speed of market adoption. To supplement this analysis interviews are conducted with investment analysts who cover pharmaceutical and diagnostics companies to gauge their views on stratified medicine. This is important due to the fact that publicly traded companies have an obligation to their shareholders, and shareholder views are shaped by the analyses of these individuals. In order to assess the future economic impact of stratified medicine on industry, particularly given that clinical biomarkers are now being developed much earlier in the R&D timeline, a model was constructed to predict economic outcomes based on various parameters associated with biomarker development. / (cont.) The aim of this model is to investigate how factors such as pricing, drug efficacy and biomarker accuracy, amongst other factors, impact the patient population, and therefore market size and economic performance for a drug with an associated biomarker. This body of analysis is then used to conduct a second set of interviews with representatives from patient advocacy groups to gauge their opinions on the ethical implications arising out of the economic considerations discussed in the first half of this thesis. In summary, this thesis undertakes a comprehensive review of the history of the adoption of stratified medicine within oncological therapeutics, and a forward-looking analysis of the economic and ethical implications with the aim of clarifying the circumstances in which stratification may be appropriate. In doing so, this thesis provides a resource to pharmaceutical companies and patient advocates attempting to chart a viable path forward in this rapidly changing field. / by Sameer A. Sabir. / S.M.

Thesis (Ph.D.)--Harvard--Massachusetts Institute of Technology Division of Health Sciences and Technology, 1998. / Includes bibliographical references (p. 113-117). / The goal of this research is to determine how aspects of human speech processing can be utilized to improve the performance of Automatic Speech Recognition (ASR) systems. Three traditional ASR parameterizations matched with Hidden Markov Models (HMMs) are compared to humans on a consonant recognition task using Consonant Vowel- Consonant (CVC) nonsense syllables degraded by highpass filtering, lowpass filtering, or additive noise. Confusion matrices were determined by recognizing the syllabies using different ASR front ends, including Mel-Filter Bank (MFB) energies, Mel-F filtered Cepstral Coefficients (MFCCs), and the Ensemble Interval Histogram (EIH). For syllables degraded by lowpass and highpass filtering, automated systems trained on the degraded condition recognized the consonants roughly as well as humans. Moreover, all the ASR systems produce similar patterns of recognition errors for a given filtering condition. These patterns differ significantly from that characteristic of humans under the same filtering conditions. For syllables degraded by additive speech-shaped noise, none of the automated systems recognized consonants as well as humans. As with filtered conditions, confusion matrices revealed similar error patterns for all the ASR systems. While the error patterns of humans and machines was more similar for noise conditions than for filtered conditions, the similarities were not as great as between the ASR systems. The greatest difference between human and machine performances was in determining the correct voiced/unvoiced classification of consonants. Given these results, work was focused on recognition of the correct voicing classification in additive noise (0 dB SNR). The approach taken attempted to automatically extract attributes of the. speech signal, termed subphonetic features, which are useful in determining the distinctive feature voicing. Two subphonetic features, intervocal period ( the length of time between the onset of the vowel and any preceding vocalization) and delta fundamental (the average first difference of fundamental frequency over the first 90 msec of the vowel) proved particularly useful. When these two features were appended to traditional ASR parameters, th-3 deficit exhibited by automated systems was reduced substantially, though not eliminated. / by Jason Sroka. / 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 245-271). / Despite significant advances in the fields of biophotonics and oncology alike, several challenges persist in the study, assessment, and treatment of cancer, ranging from the accurate identification and examination of potential risk factors, early diagnosis of dysplastic lesions, and monitoring of the complex heterogeneity of cellular populations within tumors. To study such dynamics at the microscale, non-invasive optical toolkits offer the potential to identify, characterize, and visualize key molecules and their interactions in their native biological context, ranging from in vitro cell cultures to in vivo studies in both animal models and humans. In the present thesis, examples of such applications of optical tools will be presented, including: (1) the assessment of cellular oxidative stress in ex vivo human skin cultures by imaging endogenous and exogenous fluorescent compounds using two-photon excitation fluorescence (TPEF) and fluorescence lifetime imaging microscopy (FLIM); (2) visualizing water and lipid distribution as well as cellular morphology using coherent Raman scattering (CRS) imaging techniques in the stratum corneum, the most superficial layer of the epidermis; (3) using photoconvertible labels to optically tag cell sub-populations of interest in situ for long-term monitoring of heterogeneous cell cultures from in vitro monolayers to in vivo xenograft models; (4) visualizing melanin species in the context of melanoma with coherent anti-Stokes Raman scattering (CARS) and sum-frequency absorption (SFA) microscopies. Altogether, development of such advanced microscopy toolkits will serve to improve both understanding of cancer pathology, as well as to validate clinical diagnostic and therapeutic strategies. / by Sam Osseiran. / Ph. D. in Medical Engineering and Medical Physics

Thesis (Ph. D.)--Harvard-MIT Division of Health Sciences and Technology, 2005. / Vita. / Includes bibliographical references. / Developing materials that interact effectively with surrounding tissue is a major obstacle in sensor and drug delivery research. The body's natural immune response prevents foreign objects from easily integrating with an organism. Without an intimate link between a biomedical device and the proximate environment, reliable measurements or delivery of molecules is not possible. Many of the current materials used for biomedical applications are centered on inert substances and polymers that degrade in the body but have limited functional capabilities. This thesis work addresses the need to develop materials that are capable of interacting in biological environments. Polypyrrole (PPy) is a conducting polymer that is a promising biomaterial for drug delivery and sensing applications. Because PPy is a polymer that can be made in degradable forms and because it can be stimulated electrically, it is an interactive platform for biomedical applications. By accomplishing the following research objectives, this thesis work could help develop an effective polymeric paradigm for tissue interactions: 1) Develop a new method to effectively micro-pattern electrodeposited polymers and metals for in vivo devices 2) Determine the optimal synthesis conditions of the conductive polymer, PPy, for sensor and implant applications. / (cont.) 3) Fabricate PPy tubes to be used as nerve guides to promote nerve regeneration 4) Modify PPy for neurotrophic factor drug delivery devices and antibody-based sensing applications Through the use of standard microfabrication techniques, the patterning template upon which PPy is electrodeposited can be controlled precisely. By utilizing the growth mechanism of PPy on these templates, three-dimensional polymer objects can be created. Being able to micropattern the PPy and release the polymer generates the ability to create implants and devices that are completely erodible in the body. To develop the optimum conditions for sensor and drug delivery applications, PPy implants were fabricated and implanted into rat cortical tissue. Compared to similar Teflon implants, the electrically conductive PPy had preferable characteristics for material integration in the cortex. Additionally, PPy tubes have been designed and promoted peripheral nerve growth after tissue injury. By controlling the shape and morphology of PPy, the polymer implants formed an interactive bridge with their biological environment. By incorporating bioactive molecules into the PPy matrix, materials for externally controlled drug release and sensing devices can be designed. / (cont.) Drug delivery was demonstrated through the integration of nerve growth factor (NGF), a neurotrophic factor, into the PPy followed by triggered pulsatile release. Such neurotrophic factors can be used to promote neural growth in peripheral and central nervous system injury. Because PPy is easily modifiable through the use of dopants and control of its shape, PPy provides a flexible platform for novel polymeric-tissue interactions. / by Paul M. George. / Ph.D.

Thesis (Ph. D.)--Harvard-MIT Division of Health Sciences and Technology, 2007. / Includes bibliographical references (p. 107-119). / Epidemiology, the study of disease risk factors in populations, emerged between the 16th and 19th centuries in response to terrifying epidemics of infectious diseases such as yellow fever, cholera and bubonic plague. Traditional epidemiological studies have led to modifications in hygiene, diet, and many other practices that have profoundly altered the dynamic between humans and diseases. In this thesis, we develop mathematical techniques to address modern challenges, including emerging diseases such as SARS and West Nile virus, the threat of bioterrorism, and stringent legislation protecting patient privacy. Within spatial epidemiology, one problem is to map the risk of disease across space (i.e., disease mapping), and another is to analyze the data for clustering. We propose a general technique, cartograms created from exact patient location data, that can address both of these problems. We also develop a graph-theoretical method to detect spatial clusters of any shape based on Euclidean minimum spanning trees. For mapping applications, we present an optimal strategy for mapping patient locations that preserves both privacy and spatial patterns within the data. For real-time disease surveillance, in which the goal is early detection of outbreaks based on time-series data, we introduce a generalized additive model that maintains constant specificity on various time scales. / by Shannon Christine Wieland. / Ph.D.