Treasures @ UT Dallas
Not a member yet
7697 research outputs found
Sort by
An Efficient Variational Inference Method for MRF Learning and Structured Prediction Tasks
No full textThe combination of deep neural networks and probabilistic graphical models (PGMs), especially conditional random fields (CRFs), has been studied extensively in recent years, due
to a large variety of real world applications that could benefit from combining these two
different modeling approaches, especially in the computer vision field, e.g. stereo matching,
semantic segmentation and image colorization, etc. However, the traditional methods are
either too slow to be applied on large scale PGMs (say grid models over high definition
images) or are too simple to yield significant performance improvements.
In this dissertation, we propose a highly parallelizeable inference method that is especially
suitable for combined CRF + neural network frameworks. We first apply this inference
method to general MRF/CRF learning problems, using neural networks to model the potential functions. We show that the resulting model not only yields better classification
performance on real-world tasks, but that it also yields a better generative model of the
data.
We then explain how to combine CRFs with pure deep neural networks, using our inference
method as the backbone of the learning process, to solve structured prediction problems in
computer vision tasks, e.g. stereo matching, image colorization, semantic segmentation, etc.
We show that this strategy is not only efficient on modern GPUs, but it can also achieve superior performance to pure neural network solutions in each problem domain, sometimes
dramatically so
Biocompatible Tuning of Zeolitic Imidazolate Framework-8 and Encapsulation of Vaccine Model for Controlled Release
No full textIncreased development of proteinaceous therapeutics and other biomolecules has signaled the
demand for robust drug delivery technologies. Metal-organic frameworks have shown to meet the
durability requirements but controlled delivery of their therapeutic-cargo and their long-term
release profiles remain under-investigated. In this thesis, zeolitic imidazolate framework-8 (ZIF8) is explored as a vaccine delivery vehicle in combination with the highly-investigated polymer,
poly(lactic-co-glycolic acid) (PLGA). Previous studies have shown their separate potential as
immunogenic vaccine carriers, but a combined system remains unexplored. Using FITC-tagged
ovalbumin as a model vaccine therapy, this combined vaccine delivery system is characterized and
investigated to determine its’ release profile. In a separate study, a biocompatible method of tuning
ZIF-8 micropores to create a hierarchically porous material is developed via defect formation using
bovine serum albumin. Here, the cystine residues of the protein are alkylated to control the pore
size and ultimately the release time of the encapsulated cargo. Together, these two research
explorations serve as fundamental steps towards facilitating the establishment of ZIF-8 as a viable
vaccine carrier platform
Dr. Dana Scully of The X-Files: A Feminist Scientist Navigating Patriarchies
No full textMedical doctor and federal agent Dana Scully (portrayed by Gillian Anderson) was one of the
two lead characters in the speculative television franchise The X-Files (created in 1993 by Chris
Carter) and has thus been a prominent figure in popular culture for almost three decades. This
research argues that her character has been a positive and feminist representation of a woman
scientist. Scully is indeed able to overcome significant systemic violence perpetrated by different
patriarchal systems, especially the bio-terrorist shadow organization particular to the mythology
of the series. She accomplishes this by reclaiming her threatened agency in both her professional
and her personal lives. By investigating her primary role as a medical doctor, this project traces
her development in a more comprehensive way than it would if solely focusing on her as an FBI
agent, for Scully always retains her medical expertise and puts it to various professional uses.
Exploring the intersection of real-life science, trauma, and feminism in Scully’s journey is the
primary goal of the research, since popular culture participates to the representation of society,
including science, and fuels discussion in the general public about the large and multifaceted
field. While fiction may seem incidental compared to actual scientific practice and policies, the
way a franchise such as The X-Files presents a prominent woman scientist remains important, for
art exists in dialogue with society and not in a proverbial vacuum.
This research unfolds in three sections. The first one, “Scully’s Medical Expertise,” investigates
the character’s lineage as a fictional woman scientist, as well as her medical expertise both
within the X-Files department and outside of it. The second section, “Scully’s Narrative
Journey,” focuses on her personal and professional agency, the issues of her codependent
relationship with her longtime partner Fox Mulder (the other lead character of the franchise,
portrayed by David Duchovny), and the creative process behind Scully and her longevity.
Finally, “Scully and the Bio-Terrorist Patriarchy” explores the general gendered violence
featured in The X-Files, the ideology driving the bio-terrorist patriarchy permeating a significant
part of the show’s narrative, and Scully’s trajectory from victim to potential savior.
Each section addresses core aspects of Scully’s characterization as well as the narrative
environment in which she evolves. Regardless of the state of the FBI department and its ties to
paranormal investigations and seemingly unexplained cases that give the name to the franchise,
Scully remains a medical doctor in all installments. Even with younger generations of women on
screen, Scully’s longevity and unique narrative arc still offer ground for discussion about the
depiction of women, especially female scientists
Analysis of Heme Functions in Therapy Resistance and Tumorigenesis in Non-small Cell Lung Cancer
No full textLung cancer remains the leading cause of cancer-related death in the United States, 84% of them
being non-small cell lung cancer (NSCLC). Early-stage treatment includes surgery and
radiotherapy followed by periodic radiographic imaging for routine surveillance. These curative
treatments have shown promise in some lung cancer patients while being ineffective for majority
(30–60%) of the patients who are predicted to have advanced disease, particularly local or distant
metastasis. This limit overall survival rates in these patients to less than 60%. Moreover, despite
the advent of personalized therapy which includes various targeted therapies and
immunotherapies, there has not been a significant improvement in the 5-year survival rate.
Therefore, there is a pressing need to further optimize current strategies while continuing to
explore novel strategies to improve therapeutic outcomes for patients with lung cancer, based on
individual patient needs.
Numerous studies are now focusing on the importance of mitochondrial respiration or oxidative
phosphorylation (OXPHOS) in cancer progression. However, very little is known about its role
and potential as a therapeutic target in non-small cell lung cancer (NSCLC). Several studies in our
lab show that NSCLC cells display elevated levels of intracellular heme. This increased level of
heme is either through de novo heme synthesis or heme uptake. Our lab has also demonstrated
elevated mitochondrial respiration/oxidative phosphorylation (OXPHOS) in NSCLCs. Studies in
our lab show that elevated expression of enzymes involved in heme biosynthesis, uptake, and
degradation, as well as oxygen-utilizing hemoproteins in resistant cells post treatment with
vascular disrupting agents. Limiting oxidative functions using Cyclopamine tartrate (CycT),
inhibition of heme uptake with heme sequestering peptides (HSP2) and heme synthesis using
succinyl acetone, have all shown promise in delaying growth and progression of NSCLC cells and
tumor xenografts. Another feature of NSCLC tumor is its heterogeneity. NSCLCs exhibit
widespread inter- and intra-tumoral heterogeneity as well as incidences of subtype
transdifferentiation. This kind of plasticity enable them to develop drug resistance and pose great
challenges for their treatment. In this study I aimed to understand the comparative dependence of
the two major NSCLC subtypes, adenocarcinoma (ADC) and squamous cell carcinoma (SCC), on
heme and OXPHOS, for their growth and progression. I observed that both ADC and SCC have
similar demands for heme uptake and synthesis in cell culture. My results also suggest that
OXPHOS activities are elevated in both ADC and SCC to support tumorigenic functions in culture.
I used the Genetically Engineered Mouse Model (GEMM), KLLuc to study NSCLC tumor
heterogeneity since tumors developed in these mice consists of both ADC and SCC phenotypes.
My findings in vitro were corroborated in this model, using immunohistochemistry (IHC) and
histology. Bioluminescence imaging and histology studies demonstrate that heme sequestering
peptides successfully reduce tumor development and progression in KLLuc mice
Routing Methods for Transistor-level Programmable Fabrics
No full textYou may be surprised if someone tells you that chips are the new oil. Economic historian
Chris Miller, in his new book Chip War, states that chips are the world’s most critical
resource, and explains how the semiconductor came to play a critical role in modern life.
Today, military, economic, and geopolitical power are built on a foundation of chips. In
many ways, our world is “built” on semiconductors. As the impact of digital on lives and
businesses has accelerated, semiconductor markets have boomed, with sales growing by more
than 20 percent to about 1 trillion dollar industry by the end
of the decade, assuming average price increases of about 2 percent a year and a return to
balanced supply and demand after current volatility.
The semiconductor industry has evolved over the past few decades in all fields, specifically
chip design, manufacture, packaging and testing. The chip design methodology has also advanced with the continuous scaling of the feature size in Very Large Scale Integrated circuits
(VLSI). The tiny chips are one of the most difficult devices to design in the world, following
a fairly long chip design flow, all design flow steps are necessary and equally important;
if mistakes are introduced in any step, this may make the whole chip unable to work as
expected. Among those numerous steps, routing is to connect all components in the circuit
together properly and efficiently.
Programmable logic devices, such as Field Programmable Gate Arrays (FPGA) and Complex
Programmable Logic Devices (CPLDs), have grown in popularity in a myriad of applications
since their inception due to their reconfigurability and lower non-recurrent engineering costs
when compared to Application Specific Integrated Circuits (ASICs). To keep pace with
growing application needs and process technology improvements, FPGAs have traditionally
chosen full custom chip design approaches. However, embedded FPGAs (eFPGAs) have
been introduced to enable ASICs to be less application specific, thereby producing the need
for an agile design approach to accelerate the eFPGA design process. A TRAnsistor-level
Programmable fabric (TRAP) has received interest recently as a more compact eFPGA for
hardware obfuscation, in which a selected sensitive portion of the design is implemented
in the eFPGA, and the residue is implemented as ASIC. Unfortunately, state-of-the-art
routing tools are not fully compatible with the new architecture. In this work, we develop
routing methods customized for the TRAP fabric, to address all the unique architecture
requirements. Experimental results corroborate that the proposed routing methods for the
transistor-level programmable fabric are working as needed and are fully automated with a
single push-button solution
Surface Electromyography Based Control of a Prosthetic Hand
No full textAn ideal prosthetic hand should provide individual finger actuation in various states of flexion to
achieve improved dexterity and functionality, instead of either fully open or closed positions as
seen in many conventional prosthetic hands. In order to control fingers, a suitable control input
such as electromyography signal is needed. Electromyograms or electromyography signals are the
measured electric potentials that are generated during skeletal muscle contraction. These
electromyography signals have been used for the control of many prosthetic hands. Surface
electromyography involves the placement of non-invasive adhesive electrodes on the surface of
skin covering the targeted muscle, which in this case is the forearm. This technique is safer, simpler
and also facilitates the use for prolonged periods of time when compared to the invasive methods.
Electromyography signals are weak in nature and they require amplification and filtering. This
thesis presents a fully-fledged integrated system to measure and process the electromyography
signals, and then use these signals to control a unique prosthetic hand. The system also has flex
sensors and processing boards for measuring the position of the fingers which provides feedback
for the controller. A driver board is used for actuating the artificial muscles in the robotic hand,
which receives power form a rechargeable battery. All the boards are controlled by NVIDIA
Jetson TX2 Module.
Important features of the prosthetic hand presented here are lightweight structure, low cost and
silent actuation when compared with others. It uses the relatively new polymer artificial muscles,
Twisted and Coiled polymer (TCP) muscles, based on silver-coated nylon. Prosthetic hands
typically use electromechanical actuators or pneumatic actuators, which are heavy and bulky. TCP
muscles have a high power to weight ratio and can be electrothermally-actuated by Joule heating
effect. In this thesis, the performance of the TCP muscle-actuated prosthetic hand through the use
of EMG signals is presented, mainly focusing on classical control systems such as proportional
(P) controllers, and proportional and integral (PI ) controllers
Nonparametric Regression With the Scale Depending on Auxiliary Covariates and Missing Data
No full textNonparametric curve estimation is a powerful statistical methodology which allows estimation of curves with no assumption about their shape. It provides useful insight into the
nature of data and may guide further inference for specific parametric models. Considered
statistical problem is a nonparametric heteroscedastic regression with auxiliary covariates
and missing data. In this regression a univariate component is of the primary interest while
the scale function is allowed to be dependent on both the predictor and auxiliary covariates. Missing mechanism is the missing at random (MAR), and two settings with missing
responses or missing predictors are considered. The assumed MAR means that the probability of missing may depend on observed variables but not on missing variables. Developed
asymptotic theory shows how the heteroscedasticity and MAR mechanism affect the constant of minimax convergence under the mean integrated squared error criterion. Further,
it is shown that a procedure ignoring the scale function is not efficient and does not attain a
sharp constant in the minimax lower bound. Models of missing responses and predictors are
considered separately because their theory and methodology are different. For the case of
missing responses, a sharp minimax and data-driven procedure is developed which is based
on estimation of an unknown nuisance scale function. The estimator adapts to the MAR
response mechanism and unknown smoothness of an underlying regression function. Further,
efficiency is still preserved for a more general additive model with auxiliary covariates. A
model with MAR predictors is dramatically more involved, and here classic regression estimators are no longer even consistent. For a model with MAR predictors a novel data-driven
estimator is suggested which takes into account a scale function. This estimator is adaptive
and matches performance of an oracle that knows all underlying nuisance functions. The
asymptotic theory is extended to the case of a general additive model as well. The theory
and methodology are tested using Monte Carlo simulation studies and real examples. The
results favor the proposed methodology and support practical feasibility of the proposed
methods for heteroscedastic regressions with missing data
Multidimensional Evaluation of Daily Device Use, Communication, and the Family System in Pediatric Cochlear Implant Users
No full textThis dissertation investigates samples of children with hearing loss who use cochlear implants
(CIs) through a family systems lens. The aim is to understand the impact of pediatric hearing loss
on the parents and siblings of the affected child and the impact that families have on their
children (by way of facilitating daily device use). A series of interconnected manuscripts
centered on parents, siblings, and children with CIs combine to illuminate the multidimensional
and bidirectional effects of pediatric hearing loss and the family system. Chapters 2 and 3
address the state of parents and siblings, respectively, of school-age and adolescent children with
CIs, to understand the impact of hearing loss on other members of the family. Chapter 2 (Study
1) compares general and condition-specific stress (via the Family Stress Scale) in 31 parents of
CI users (8-16 years) to previously published samples of children with HL, finding similarities
and differences across samples. Child temperament significantly predicted parental stress after
controlling for other variables. Chapter 3 (Study 2) examines quantitative and qualitative
perspectives of 36 children and adolescents with typical hearing (age 6-17 years) who have a
sibling with CIs (age 7-17). Quantitative results indicated that siblings with TH express positive
perspectives of their brother or sister with CIs and report having a CI user in the family does not
affect them much, particularly if the CI user has good speech understanding and intelligibility.
Qualitative responses diverge from quantitative data, with siblings expressing more negative
feelings surrounding differential attention from parents and the CI user’s social communication
skills. Chapter 4 (Study 3) considers an aspect of parental involvement - daily device use – in 65
young children with CIs (< 5 years), exploring the impact of device use on of emerging
communication skills in 65 young children. Results of this retrospective chart review indicate
better early auditory skills, speech recognition in quiet skills, and expressive/receptive language
outcomes in children who wear their CIs more hours per day (on average). Chapter 5 provides
pilot data for a prospectively examination of daily device use, family-related variables (e.g.,
parental involvement, socials support, family hardiness), and communication outcomes (i.e.,
speech recognition, spoken language) in a small group of young children who predominantly use
CIs (n = 9, age < 6 years), representing the intersection of topics in Studies 1-3. These data
demonstrate variability in daily device use and communication outcomes, but little difference in
family variables (e.g., low parental stress, high parental involvement), revealing both feasibility
and challenges of data collection moving forward. These studies collectively highlight the
importance of considering the entirety of the family system in cases of pediatric hearing loss to
maximize communication outcomes in children with hearing loss and to optimize well-being in
all members of the family
B–type Catalan States of Lattice Crossing
No full textM. K. Dabkowski and J. H. Przytycki defined for any realizable Catalan state C with no
bottom returns, the rooted plane tree with a delay function, (TC, f), and the partially
ordered set (B(C), 4) of some Kauffman states that realize C. In this dissertation, we
study the properties of (B(C), 4) and establish an important relation between its rank
generating function and the plucking polynomial of (TC, f). Furthermore, we show that the
rank generating function of (B(C), 4) is unimodal for any realizable A–type Catalan state
with no bottom returns of an A–type lattice crossing LA(m, n), where n ≤ 4. In the last part
of this dissertation, we study B–type Catalan states. We show which crossingless connection
between 2(m + n) outer boundary points of an annulus can be realized as Kauffman states
of the B–type Lattice crossing LB (m, n). Furthermore, we give a closed-form formula for
the number of realizable B–type Catalan states, and find coefficients of those obtained as
Kauffman states of LB(m, 1) and LB(m, 2)
Can I Get a Large Cup of Asian American Identi-tea? With Boba, Aesthetics, and Instagram
No full textRacializing Asian people, food, and culture stemmed from concepts like yellow peril and the
model minority myth, thus impacting the perception of Asian people in the United States. As
Web 2.0 develops, our digital media, smartphone technologies, and social media apps like
Instagram, all work as a playground for current Asian American identity and representation. Due
to the proliferation of East Asian food and culture in the United States, bubble tea (also known as
boba) and the aesthetics of bubble tea shops are analyzed and visually read to determine their
impact and affect on Asian American identity building. The aesthetics of these shops are studied
and elaborated in relation to smartphone photography, Instagram aesthetics, and food
photography