1.536
Arşivlenen Tez
0
DOI Atanmış
0%
DOI Oranı
Arşivlenen Tez
Centriolar satellites are required for efficient ciliogenesis and ciliary content regulation.
Centrosome is the main microtubule organizing center of the cell. It has the role in cell division, cell shape and migration. Other than these roles, centrosome has an important function which is forming primary cilium. Primary cilium is a microtubule-based organelle serving as signaling hub for the cell. Structural or functional defects associated with centrosome/cilium complex lead to genetic diseases called ciliopathies. In order to understand molecular mechanism under ciliopathies, it is important to understand how centrosome/cilium complex is regulated in time and space. Centriolar satellites are the third component of the centrosome/cilium complex In the second chapter of my thesis, I characterized the cells without centriolar satellites in terms of ciliogenesis and ciliary function. I established the satellite-less cells by knocking-out PCM1 gene in inner medullary collecting duct (IMCD3) cells. PCM1 is the scaffolding protein of the centriolar satellites. Satellite-less IMCD3 cells do not ciliate efficiently as much as control cells. Content and function of cilia formed by satellite-less cells are also different. To understand how cellular process are affected in PCM1 knock-out cells, I applied tandem mass tag (TMT) labeling‐based quantitative analysis to compare the global proteome control and satellite-less cells. This analysis revealed that many processes such as actin cytoskeleton, cell migration and adhesion, endocytosis, neuronal processes are affected in the absence of PCM1 protein. With these observations, I showed that centriolar satellites are important to regulate ciliogenesis, ciliary content and function. However, the mechanism behind this regulation is poorly understood. In the third chapter of my thesis, I applied the miniTurbo labeling method to compare interaction partners of PCM1 in asynchronous and ciliated cells. The aim of this chapter to find interaction partners of PCM1 specific to ciliated cells to explain mechanism of regulation by centriolar satellites during ciliogenesis. After miniTurbo experiments and mass spectrometry analysis, we determined a set of protein which interact with PCM1 in ciliated cells. Among these proteins, we checked the localization and function of the TBC1D31 protein by loss of function experiments. In this chapter, we determined the set of protein might have a role in regulation of ciliogenesis with centriolar satellites. The interactions between these proteins with PCM1 and their relationship in the regulation of ciliogenesis are going to be explored in our future studies.
Deep learning approaches for vocal tract boundary segmentation in rtMRI
Recent advances in real-time Magnetic Resonance Imaging (rtMRI) provide an invaluable tool to study speech articulation. Development of automatic algorithms to detect the landmarks defining the boundaries of the vocal tract (VT) is crucial for a wide range of research, from speech modeling and synthesis to clinical research. In this thesis, we present two effective deep learning approaches for supervised detection and tracking of vocal tract contours in a sequence of rtMRI frames: (1) we propose a fully convolutional network to estimate the VT contour in heatmap regression fashion and (2) we introduce a deep temporal regression network which learns the non-linear mapping from a temporal overlapping fixed-length sequence of rtMRI frames to the corresponding articulatory movements. We as well introduce two post-processing algorithms succeeding the deep models, to further improve the quality of VT contour detection: (i) a novel appearance model based contour refinement to overcome the potential failures of data-driven approaches for highly deformable articulators and (ii) a spatiotemporal stabilization scheme to stabilize the estimated contours in space and time by removing the spatial outliers and temporal jitter. The proposed VT contour tracking models are trained and evaluated over the large audiovisual USC-TIMIT dataset. Performance evaluation is carried out using various objective assessment metrics for the spatial error and temporal stability of the contour landmarks in comparison with several baseline approaches from the recent literature. Results indicate significant improvements with the proposed methods over the state-of-the-art baselines. In addition, we develop a graphical user interface (GUI) for the analysis of the rtMRI data, integrated with various attributes including automatic segmentation of the VT boundaries using the proposed contour estimation methods and calculation of tract variables and cross-sectional distance.
Full-color holographic near-eye displays
Near-eye displays (NEDs) for augmented reality (AR) applications are expected to be the next computing paradigm. NEDs offer to combine computer-generated visuals with our physical world in a seamless fashion. Depth cues and natural blurring of images are all critical for a comfortable 3D experience with near-eye displays. Holography for NEDs is a commonly accepted strong candidate in meeting the human visual system's demands by offering natural depth cues. Holographic near-eye displays (HNEDs) deliver virtual images using computer-generated holograms (CGHs) displayed on spatial light modulators (SLMs). Holographic displays allow for a vast range of optical architectures that are not possible with conventional microdisplay-based designs. However, due to the current technological limitations of SLMs, most existing HNEDs have a limited field of view (FOV) and viewing region around the eye pupil (i.e., eyebox size). However, improving SLMs have a positive impact on improving FOV and eyebox size in HNED design. In this thesis, we propose various solutions for overcoming the current technological limitations of SLMs and HNEDs. We start by offering a paraxial matrix optics-based analysis of a conventional HNED design to formulate the relation between eyebox, FOV, and SLM characteristics. We developed a CGH computation procedure that applies to arbitrary paraxial optical architectures, where the SLM illumination beam can be collimated, converging, or diverging. The virtual or real SLM image as seen by the eyebox plane may form at an arbitrary location. Using this approach, we designed full-color HNEDs with varying FOV and resolution characteristics and proper depth control. We demonstrated a lensless HNED architecture with diverging beam illumination, which provides 3D images within a wide FOV (70°) at retinal resolution (30 cycles-per-degree), exceeding 4,000 resolvable pixels on a line. The experiments using binary holograms imprinted on masks prove that the proposed CGH computation procedure eliminates chromatic aberrations and speckle noise observed in all other laser-based displays. We also designed two systems with 10° and 20° FOV using dynamic SLM. While the first design has a uniform resolution, the second design demonstrates a foveated display, which has gradually degrading resolution across the FOV. To further explore HNED architectures, we analyzed the light source coherence requirements and investigated HNED designs utilizing light-emitting diodes (LEDs). While laser light sources have the highest degree of spatial and temporal coherence, lasers' usage in direct contact with a human may cause health hazards. We show that under certain design restrictions, it is possible to utilize LEDs instead of lasers and get better quality holographic images. We analyzed the effect of LED emission areas on image resolution, quality, and depth perception. Lastly, we designed micro-mirror array (MMA) based thin components and demonstrated that those could be used as off-axis thin lenses in AR displays to reduce size and volume.
mmWave channel model for intra-vehicular wireless sensor networks
Intra-vehicular wireless sensor networks (IVWSNs) have significant potential to reduce part, manufacturing and repair costs, facilitate the integration of new nodes and provide more freedom for the sensor placement in previously impossible locations by obviating the need for wiring harness. mmWave stands up as a promising candidate to fulfill the high reliability, security and low latency requirements of IVWSNs, exploiting the availability of large bandwidth at high frequencies and high nominal gains with directional antennas. This work focuses on building the channel model for the engine compartment, passenger compartment and beneath the chassis of a vehicle by conducting vast number of measurements for 14x14, 13x13and 15x15 transmitter and receiver links in a Fiat Linea, respectively. The path loss exponent is approximately 3, showing almost no variation within different compartments. The power variation around the path loss model has a Generalized Extreme Value (GEV) distribution with zero mean for all compartments and 5 dB standard deviation for the engine compartment and approximately 7.6 dB standard deviation for the other two compartments. A modified Saleh - Valenzuela (SV) model is used to represent the clustering of power delay profiles (PDPs). Log-normal distribution is used to model the inter-arrival times of clusters, while the dependencies of cluster amplitude and ray decay rate on the cluster arrival times are represented by a dual slope linear fit model with breakpoints 1.2 ns and 5.6 ns for engine compartment, respectively, and 1.6 ns and 2.6 ns for the other two compartments, respectively. The experimental PDPs vary around the SV model and these variations are represented by a normal distribution with zero mean and 5.8 dB standard deviation, which is independent of both the delay bins and the compartment of the vehicle. All these findings are used to build a simulation model for each compartment of the vehicle. The simulation model is validated by comparing the distributions of the received powers and Root Mean Square (RMS) delay spreads of the experimental and simulated PDPs.
Chemically induced assay for centriolar satellite mispositioning reveals their functions at the primary cilium
Centriolar satellites are membrane-less, electron dense granular structures that localize and move around centrosomes and cilia. The satellite proteome is composed of over 200 protein components, which were implicated in a wide range of functions such as centriole duplication, cell division, cellular signaling, primary cilium biogenesis and microtubule dynamics. Importantly, various proteins mutated in ciliopathies and primary microcephaly were also identified as part of the satellite proteome, suggesting an intimate link between centriolar satellite function and development. Although centriolar satellites have remained as understudied structures since their discovery more than 60 years ago, recent work showed that satellites store, modify and traffic centrosome/cilium proteins and play important roles in the regulation of centrosome/cilium biogenesis and function. To mediate their trafficking function, satellite cluster and move around centrosomes in most cell types. However, satellite distribution varies in response to different stimuli such as cell cycle cues and across different cell types, suggesting context-dependent functions for satellites. Dissecting these spatial and temporal functions have been challenging using traditional approaches such as loss-of-function studies, in particular, in ciliated cells. To overcome these challenges, I developed a chemical based inducible trafficking assay that allows efficient redistribution of centriolar satellites to cell periphery or cell center. Using this assay, I showed that satellite mispositioning disrupts centrosomal targeting of key regulators of cilium biogenesis, identifying a direct role for satellites in centrosomal protein targeting and sequestration. To identify the functional consequences of satellite mispositioning, I used functional assays to probe cilium biogenesis and microtubule dynamics in cells where satellites were mispositioned at the membrane. The results of these assays showed that satellites regulate primary cilium assembly, maintenance and disassembly. Taken together, our results showed a direct link between satellite functions and their pericentrosomal clustering in ciliated cells and also provided a new tool for studying acute functions of satellites in a context-dependent way. Finally, given the crucial roles of the primary cilium as the signaling center for developmentally important signaling pathways such as Hedgehog signaling, our results sheds light into why satellite proteins are mutated in developmental disorders.
Optimal scheduling for full duplex wireless powered communication networks
According to recent Ericsson mobility report, 24.6 billion sensor nodes are expected to be installed by 2025. Increasing the lifetime of this massive battery-powered installation, efficient spectrum utilization and strict delay requirements are the major challenges. Low power transceivers with energy harvesting capability, intelligent medium access protocol and full-duplex (FD) communication can overcome these challenges. Therefore, we investigate a FD wireless powered communication network (WPCN), in which a hybrid access point transmits wireless energy by using radio frequency signals and users harvest this energy to transmit information. We consider minimum length scheduling problem (MLSP) and sum throughput maximization problem (STMP) subject to traffic demand, energy causality and maximum transmit power of the users for a continuous rate (CR), discrete rate (DR) and constant transmission rate models. The novel formulated optimization problems are non-convex and combinatorial in nature, thus, difficult to solve for the global optimum. As a solution strategy, we demonstrate that the power control problems (PCPs) and scheduling problems can be solved separately in the optimal solution. For CR-MLSP, we optimally solve the PCP by evaluating Karush-Kuhn-Tucker conditions. For the scheduling, we introduce a penalty function allowing reformulation of problem as a sum penalty minimization problem. Based on the characteristics of the penalty function and optimality analysis, we propose two polynomial-time heuristic algorithms and a reduced-complexity exact algorithm employing smart pruning techniques. Next, many WPCNs are expected to use low-power transceivers with finite discrete configurations, we consider a novel DR-MLSP, where users select a rate from a finite set of discrete-rate levels. We optimally solve the PCP by using the optimality conditions of minimum length scheduling (MLS) slot, which is defined as a slot of minimum transmission completion time while starting transmission at any time after the decision time. Then, for scheduling, we classify the problem based on whether the MLS slots of the users overlap over time. We present the optimal algorithm for non-overlapping slot scenario based on the allocation of MLS slots, and a polynomial-time heuristic algorithm for overlapping scenario by allocating the transmission slot to the user with earliest MLS slot. Besides, we consider a multi-cell WPCN with concurrent transmission of users for constant and continuous rate models. We solve the PCPs by proposing optimal algorithms based on the evaluation of Perron-Frobenius conditions and usage of bisection method for constant and continuous rate models, respectively. Then, the solutions of PCPs are used to solve the scheduling problems. For the constant rate scheduling problem, we propose a heuristic algorithm which aims at maximizing the allowable interference on each user within a concurrently transmitting set. For the CR scheduling problem, we define a penalty function representing the advantage of concurrent transmission over individual transmission of those users. Then, following the optimality analysis and demonstration of the equivalence between MLSP and minimization of the sum of penalties, we propose a heuristic algorithm which allocates the users concurrently to minimize the sum penalties over the schedule. Furthermore, we consider an on-off transmission scheme for a single hop and relay-based WPCN, in which users either transmit at constant power or remain silent. For single hop problem, we propose a polynomial-time optimal scheduling algorithm. For relay-based system, following an optimality analysis, we propose a heuristic algorithm that performs very close to the optimal solution. Finally, for the CR-STMP, the PCP is proven to be convex and solved optimally. For scheduling, based on the derived optimality conditions, we propose a fast heuristic algorithm, which performs very close-to-optimal solution. Then, we characterize a novel optimization framework for DR-STMP to determine the rate adaptation and transmission schedule. We investigate the characteristics of the solution and propose a polynomial time heuristic algorithm for rate adaptation and scheduling problem.
A novel control mechanism of mitotic exit in Saccharomyces cerevisiae
Mitotic exit is the cell cycle stage in which the cell transits from M phase to a new G1 phase. While mitosis is under control of cyclin-dependent kinases (CDK), mitotic exit depends on inactivation of CDKs. Timely coordination of mitotic CDK inactivation with respect to chromosome segregation is important to avoid aneuploidy and maintain ploidy. A comprehensive understanding of regulation of mitotic exit is missing. In budding yeast, mitotic exit is achieved by the use of a special network, called Mitotic Exit Network. An inhibitor to Mitotic Exit Network, Kin4 kinase, leads to lethality for the cells when is expressed in high doses. Our goal is to find out novel mechanisms of mitotic exit control. In this thesis, we identified a temperature sensitive mutant that rescues the lethality of Kin4 overexpression. Through a "dosage suppressive genetic screening", we found two genes, SAN1 and PHO2, that retarded growth this temperature sensitive mutant when Kin4 was overexpressed. We further characterized the effect of these genes on mitotic exit, with a focus on SAN1. Our results indicate a novel regulatory mechanism for mitotic exit. Data presented in this thesis will pave the way to illuminate a new section in the process of exiting from mitosis. Owing to the fact that the basic cellular tasks are preserved from yeast to human, we envisage that characterization of this newly emerged role in the mitotic exit of S. cerevisiae would contribute to the understanding of analogous control mechanisms in more complex organisms.
Novel optoelectronic and plasmonic bulk heterojunction neurointerfaces for controlled capacitive charge transfer
Artificial control of neural activity allows for understanding complex neural networks and improving therapy of neurological disorders. Light is a non-invasive communication trigger with biological systems. Proper transduction of light to bioelectrical stimuli via artificial photoactive devices requires simultaneous satisfaction of safety, efficiency, and current direction control. For safety, we demonstrated novel photovoltaic neurointerfaces that incorporate biocompatible materials and induce capacitive charge-transfer based on charging and discharging of double layer at the electrode-electrolyte interface without irreversible Faradaic reactions. For that, we developed a single-junction, wireless and capacitive-charge-injecting biointerface by using a high open-circuit voltage (0.75 V) bulk heterojunction of PTB7-Th:PC71BM. For efficiency, we integrated plasmonic interactions to optoelectronic biointerfaces. So far, plasmonics has been primarily used for heat-induced cell stimulation due to membrane capacitance change (i.e., optocapacitance). For the first time, we demonstrated that plasmonic coupling to photocapacitor biointerfaces improves safe and efficacious neuromodulating displacement charges for an average of 185% in the entire visible spectrum while maintaining the Faradaic currents below 1%. For current direction control, we show that utilization of photovoltaic biointerfaces combined with light waveform shaping can generate safe capacitive currents for bidirectional modulation of neurons. The differential photovoltage response of the double-layer capacitor facilitates the direction control of capacitive currents depending on the slope of light intensity. Hence, the findings of this thesis show the great promise of optoelectronic neurointerfaces for non-genetic, all-optical and safe modulation of neurons.
Shape memory alloy design by machine learning for biomedical and high-temperature applications
Shape memory alloys (SMAs) are of great importance due to their extensive usage in biomedical applications, aerospace engineering, or robotics. In recent years, although there has been a considerable amount of research to achieve optimum compositions of SMAs for these applications, due to the high experimental costs, the demand for alloys with optimum properties has not been met for many applications yet. In this research, using the predictive power of artificial intelligence and machine learning, a systematic approach to predict the optimum composition of the SMAs was proposed to address two problems related to the binary NiTi and NiTi-based SMAs. In particular, in chapter two, the optimum chemical composition was proposed to minimize the Ni ion release in the binary NiTi SMA. The method to do so was to gather a database from the existing literature and using it to train a special algorithm that provides the information for predicting the desired compositions. In chapter three, using the same approach, two models were developed to predict the phase transformation temperatures and thermal hysteresis of multi-component NiTi-based SMAs. These models were used to predict the optimum alloy with the highest Phase transformation temperatures with the least possible thermal hysteresis.
Navigation based on inertial sensor data using deep learning techniques
Estimating the location of pedestrians continuously and in real-time indoors and outdoors is an important problem in medical rehabilitation, occupational health and safety as well as retail applications. When a clear view of satellites is available outdoors, the Global Positioning System (GPS) can provide accurate positions to smart phones and smart watches. GPS signals, however, are not always available especially in indoor or dense urban environments due to multi-path reflection or signal blockage by buildings. Utilizing the existing wireless infrastructures like cell-tower or wireless local area networks (WLANs) is another possibility for indoor pedestrian navigation by triangulation methods. Nevertheless, these solutions also suffer from multipath loss and similar signal problems. Strapdown inertial navigation with zero-velocity update (ZUPT) or pedestrian dead reckoning (PDR) methods provide a navigation solution based on on- board inertial measurement units (IMU) without depending on any external infrastructure. These dead reckoning methods, however, have unbounded consumer-grade IMU positioning errors due to sensor error accumulation by integration. These errors originate from finite and bounded random drift, axis misalignment, scale factor, and thermomechanical white noise. An appropriate way for continuous pedestrian positioning is to use an accurate IMU based method as a sub-system of an integrated navigation unit. The IMU-based positioning is a functional solution even in the case of a temporary signal blockage of the position-fixing subsystem such as the GPS or WLAN. Recently, it has been shown that deep learning (DL) based dead reckoning methods outperform the classical ZUPT and PDR methods in terms of the positioning accuracy. In this thesis, the state-of-the-art deep inertial odometry methods have been refined, made more accurate, smaller in memory size and latency. Recent DL-based dead reckoning methods show that deep recurrent neural networks can yield highly accurate trajectories compared with other shallow techniques without resorting to visual odometry. By refining the DL model architecture, we present a compact and robust deep inertial odometry methodology. While the root-mean-squared error (RMSE) for the estimated position decreases by 26% in our model, the number of trainable parameters and the latency of the artificial neural network (ANN) are decreased by 64% and 50%, respectively. Thus, the ANN method has become more feasible to implement on mobile devices and embedded systems. Furthermore, the proposed DL architecture is extended on drone positioning problem using IMUs. Using three different drone positioning datasets, DL architectures have been trained and tested. While drone-positioning using only IMU data is feasible, this problem presents additional challenges due to complex motion dynamics. Thus, this model can help improve positioning accuracy in applications involving mobile devices with indoor uses such as search and rescue, sports performance measurements, drone localization.