7940 research outputs found
Sort by
Investigation of SOX9 Mediated Epidermal Growth Factor Receptor Tyrosine Kinase Inhibitor Resistance Reverting Potential of Piperlongumine in Lung Cancer
No full textSRY-related high mobility group box 9 (SOX9) has been identified as a promising transcription factor in the initiation and progression of various tumor types, including lung cancer. The present study investigates the role of SOX9 expression in EGFR-TKI (gefitinib) resistant lung cancer in vitro and ex vivo. It was found that upregulation of SOX9 was associated with gefitinib resistance with an increased rate of cell proliferation, migration and invasion, single-cell colony-forming ability, and reduced apoptosis in gefitinib resistant cells A549/GR and NCI-H522/GR. Moreover, upregulated SOX9 promoted epithelial to mesenchymal transition (EMT) via targeting β-catenin and its knockdown reversed the resistance and EMT phenotype in A549/GR and NCI-H522/GR cells. Similarly, it was found that multicellular spheroids of A549/GR and NCI-H522/GR cells showed a larger surface area with more dispersion, while SOX9 knockdown abolished these gained properties ex vivo. To date, many natural compounds have shown promising chemotherapeutic potential. Piperlongumine (PIP), a white to beige biologically active alkaloid has high pharmacological relevance as an anticancer agent. It has been found that PIP suppresses the cell proliferation, migration, and stem-cell-like phenotype by targeting SOX9, resulting in induction of apoptosis in A549 and NCI-H522 cells via attenuating EMT. PIP effectively showed cytotoxicity against A549 and NCI-H522 cells and induced reactive oxygen species (ROS) production and cell cycle arrest. Additionally, PIP induced dissipation in mitochondrial membrane potential and affected the expression of intrinsic apoptotic pathway proteins in lung cancer cells. PIP inhibited EMT and stemness of lung cancer cells. Further, ex vivo analysis revealed that PIP induced apoptosis and size reduction in multicellular spheroid of A549 and NCI-H522 cells. The emergence of resistance to gefitinib indicates that use of a single agent may not be sufficient for the effective treatment of lung cancer. Therefore, combinatorial effect of PIP with gefitinib in A549/GR and NCI-H522/GR cells was examined. The result suggested that co-treatment of PIP with gefitinib in lung cancer cells showed increased cytotoxicity as compared to gefitinib or PIP alone and reversed the resistance via inhibiting EMT. At last, the overall study delineated a deep insight into the mechanism of PIP mediated cancer inhibition and its combinatorial therapeutic effect in re-sensitization of SOX9 induced EGFR-TKI resistance in lung cancer
Synthesis and Functionalization of Sandwich Based Molecular Scaffold for their Study in Catalysis and Biological Behavior
No full textSandwich based molecular scaffolds are the most versatile molecular entities in organometallic chemistry because of their unique structural integrity, reversible redox behavior and huge potential in the development of material science, molecular wires, catalysis, sensing, electronic communication and bio-organometallic chemistry. Ferrocene quickly attracted the attention of the scientific community on account of its intriguing chemistry. Recent investigation supports ferrocenyl analogues obtained by diligent design and modification at different structural levels such as ferrocene conjugates with alkyl chain, organometallic moieties, heterocyclic moieties, phenol position, biomolecules and conjugated aromatic substituents causes enhancement of the cytotoxic activity. However, the functionalization of cyclopentadienyl rings of ferrocene by different fragments poses numerous challenges in regard to reaction conditions and methodologies. In view of the enormous opportunities of sandwich compounds and availability of limited reaction conditions we were motivated to employ a solid supported synthesis route to overcome such difficulties. In addition, we have been motivated to synthesize various ferrocenyl derivatives in monofunctionalized, unsymmetrically bi-functionalized form and explored their electrochemical, biological, photovoltaic and catalytic activities. The thesis with five chapters describes the synthesis and characterization of some novel ferrocene based organometallic systems containing unsymmetrical ferrocenyl hydrazone-thiosemicarbazone derivatives, diferrocenyl thiosemicarbazone enones, cyanovinyl donor-acceptor compounds and ferrocenyl cyclopalladated complexes. This report also describes interesting behavior of these compounds towards interaction with DNA and BSA. This sheds light on to their biological activities further supported by molecular docking study to determine relevant mechanism and modes of interaction. Investigation of cross coupling reactions of cyclopalladated complexes through different mechanistic path way by DFT calculation has also been incorporated. Investigation of their cytotoxicity and wide absorption studies have been described to understand their potential for biological and solar cell study
Learning Centric Feature Extraction and Classification Models for OCR
No full textCharacter recognition is the process of enabling computers to classify the characters from their image presentation. The practical application involves bankcheck reading, book reading for the blind, postal address reading, and many more. With emerging trends in technology, mobile devices have been equipped with high definition cameras and powerful processors. This pattern enlarges the scope of mobile applications to read business cards, street signboards, medical prescriptions, and local language translation for travellers, etc. The essential step in all these applications is the recognition of characters. Traditionally the application uses handcrafted features but in recent years, the nonhandcrafted feature extraction methods have gained increasing popularity for solving pattern classification tasks due to their inherent ability to extract robust features and handle outliers. This dissertation focuses on the design of nonhandcrafted learning models for character recognition application. The research primarily involves the proposition of hyperparameterless learning models that can be used for the image classification task. The dissertation proposed three newly developed learning methodology for character classification from raw images. The first contribution overcomes limitations of classification over Single Layer Feed Forward Network. The network is a widely validated model for the classification task. The limitations of such model are due to the need for hyperparameter tuning. Extreme Learning Machine was developed as a hyperparameterless to overcome this limitation. However, the random input weight in Extreme Learning Machine makes it suffer from the illposed problem. Hence, a hyperparameterless algorithm namely BackwardForward Extreme Learning Machine (BFELM) is developed that learns the input and output weights in one backward and one forward pass respectively. The second contribution extended the BFELM framework to learn the weights of the convolutional neural network. The newly developed model is named Convolutional Network with Backward Forward Extreme Learning Machine (CNBFELM). Indepth analysis of the proposed model over the various publicly available dataset, prove its efficiency in hyperparameterless nonhandcrafted learning. The third contribution involves the development of feature learning by optimization. First, feature learning is modeled as an optimization problem that does not depend on classification error or accuracy. It is worth mentioning that the optimization model can generalize even with a smaller number of the training sample. The final contribution is about designing the convolutional neural network model for handwritten word recognition in Indic language. The CMATERdb2.1.2 dataset is used to study the proposed model in depth
Weakly-supervised Anomaly Detection and Classification in Untrimmed Surveillance Videos
No full textAnomaly detection and classification is a crucial task for ensuring public safety in real world surveillance videos. Due to the increasing number of crimes, act of terrorism, and accidents, timely detection and classification of such anomaly cases can minimize the loss. However, the task of anomaly detection and classification is challenging due to the unavailability of large annotated video data, untrimmed nature of the surveillance videos, and difficulties in obtaining discriminative feature representations for anomaly categories. This thesis aims at mitigating the mentioned challenges by adopting weakly supervised learning methods, efficient temporal dependency modeling schemes in untrimmed videos and extracting discriminative feature representations from space time convolutional neural network (3DCNN). For this, three contributions have been made in this thesis and each of the contribution mitigate one or more challenges. The first contribution focus on extracting discriminative spatiotemporal feature representation from a proposed multilevel 3DCNN for anomaly video sequences and subsequently utilizes long short term memory (LSTM) module for obtaining effective temporal dependency encoding in anomaly detection task. From result analysis, it is found that the feature representation obtained from the multilevel 3DCNN is not only superior over existing competent methods but also it is robust to challenges like illumination changes, partial occlusion. In the second contribution, the knowledge of image and video classification is used through inflated 3DCNN for obtaining enhanced feature representation. In addition, a temporal interintra clip pooling strategy is proposed for efficient temporal encoding by LSTM module in anomaly detection task. The effectiveness of this contribution has been demonstrated over existing methods and the previous contribution through exhaustive experimentation. Since the first two contributions only focus on anomaly detection task in untrimmed surveillance videos, the task of anomaly classification still remains challenging. Thus, the third contribution of this thesis proposes a framework that jointly handles the anomaly detection and classification task. In addition, the proposed framework is driven by twolevels of temporal attention mechanism which highlights the temporal saliency in the feature map as well as develops a dependency between the detection and classification task. From exhaustive experimental analysis and visualization, it is observed that the proposed method not only boosts the anomaly detection performance but also substantially improves the classification accuracy
Metal Organic Frameworks as Potential Candidates in Photocatalysis and Chemical Sensing Applications
No full textThis work demonstrates a systematic investigation on aqueous phase applications of contemporary Metal Organic Frameworks (or MOFs). The proposed areas of research in this study were photocatalysis and chemical sensing. The investigated MOFs were primarily MIL-125(Ti), NH2-MIL-125(Ti), and MIL-53(Fe). Different variants of MIL-53 (Fe) were synthesized via cation (Li+, Na+, K+), anion (S2-), and metal (Au, Ag) doping. Similarly, Polyaniline (PA) functionalization of NH2-MIL-125(Ti) was successfully carried out. All parent MOFs and their variants were systematically synthesized via facile microwave routes in a high-pressure microwave reactor. Syntheses recipes were thoroughly optimized for reaction temperature and time for consistency in product formation. Optical band gap energy of hybrid MOFs from UV-vis diffused reflectance analysis was estimated using both Kubelka-Munk (K-M) and dielectric models and compared. A host of conventional inorganic photocatalysts (semiconductors), conductors, and insulators were successively analyzed for comparison. The dielectric model was found to be equally effective when compared with Kubelka-Munk (K-M)) method and the variations were within 0.16 - 7.07% for different sets of conventional and hybrid materials. Additionally, the dielectric model was found to be more definitive when multiple absorption peaks/unresolved peaks were present in the diffused reflectance spectra (a common feature with hybrid MOF or poly-functional materials). The dispersion-dissipation plot based on the dielectric model could rightly predict the material class (semi-conductor/conductor/insulator) accordingly and was shown in the study. MIL-125(Ti) showed remarkable aqueous phase stability with measured BET surface area of ca. 112 m2g-1. The product was found to be photo-catalytically active with a band gap energy of 3.14 eV and successfully degraded methylene blue dye (ca. 96.77% in 360 minutes). Water stable NH2-MIL-125(Ti) was successfully functionalized with Polyaniline (PA) with an increase in electron conductivity (as reflected in various intrinsic parameters such as fluorescence intensities, quantum yield, and dielectric functions) leading to successful detection of dopamine (DA) with excellent sensitivity. The corresponding LOD was estimated to be 10 nM and 5 nM in the case of NH2-MIL-125 (Ti) and PA-NH2-MIL-125 (Ti) respectively via Stern Volmer model at Signal-to-Noise (S/N) = 3.0. Band gap tuning of synthesized MIL-53(Fe) through cation (Li+, Na+, K+), anion (S2-), and metal (Au, Ag) doping (structural/interstitial) and their influences in photocatalytic applications viz. dye degradation and peroxidase activity were investigated. The influence of cation doping showed a decrease in band gap energy: Li+ (1.81 eV), Na+ (1.875 eV) and K+ (1.89 eV) when compared with the parent MIL-53(Fe) (1.95 eV) whereas the trend was reversed for anion (S2-) doping: 1% S (2.224 eV); 4% S (2.376 eV), 7% S (2.448 eV) and 10% S (2.579 eV). Cation doping, leading to the distortion of the octahedral geometry of Fe3+-Fe2+ in MIL-53(Fe) played a significant part in influencing the performance of doped variants of MIL-53 (Fe). The IVCT band intensity was highest for 1% S when compared with MIL-53(Fe) and its cation doped counterparts which were reflected on the photoactivity. The intrinsic material properties viz. band gap energy, electron conductivity, and surface area had a combinatorial influence on the photocatalytic performances. The incorporation of Au/Ag nanoparticles (NPs) into MIL-53(Fe) matrix occurred interstitially and it could be observed in HR-TEM analysis. Similarly, the plasmonic activity of gold was observed at 540 nm with no change in band gap for MIL-53(Fe). The study showed marked improvement in H2O2 detection with LOD of 0.0623 μM within a linear range (1- 10 μM) (S/N = 3). The presence of H2O2 in spiked milk samples were analyzed and modeled through linear regression and the resulting model was validated with an accuracy of 0.465% and 0.45% with RSD % lying within 1.93% and 3.81% for two different models falling within linear ranges (1-10 μM) and (10-100 μM) respectively. The interference/sensitivity studies showed inherent pitfalls in the assay due to the presence of certain common milk constituents. Overall, Au@MIL-53(Fe) demonstrated as an excellent colorimetric sensor for the detection of H2O2 both in model solution as well as in processed milk and could trace possible adulteration of milk. The material holds promise in the fabrication of various diagnostic prototypes for specific applications in food processing, medical and environmental remediation
Biological Synthesis of Metal / Metal Oxide Nanoparticles and Their Applications in Food Safety
No full textInfringement of bacterial contamination leads to devastating infectious diseases worldwide and can affect human health in two possible ways. Firstly, due to the lack of quality control in food processing sectors, food is contaminated by bacterial pathogens. Secondly, inadequate access to safe and portable clean water along with poor hygiene and sanitation facilities can lead to contamination with pathogenic bacteria—both of the cases resulting in numerous foodborne disease outbreaks. Food industries is one of the potential candidate concerned with the outbreaks of the infectious diseases as it has the susceptibilities for the pathogen contamination. The current practices for controlling the outspread of microbial infections include careful monitoring of the contamination pathway of different pathogenic bacteria by various food safety measures, such as the development of a biosensing platform for pathogen detection and application of antimicrobial agents. Conventional strategies for pathogen biosensing have several drawbacks, like being less sensitive for real sample analysis, time-consuming and laborious. Similarly, several inorganic and organic materials have been implemented as antibacterial agents, which possess some shortcomings due to the high toxicity and cost-effectiveness. The present study focused on the biological synthesis of metal and metal oxide nanoparticles using plant sources as reducing agents and finding their suitable applications in two different aspects of food safety measures, (1) antimicrobial agent, and (2) biosensing of foodborne pathogens. Two different plant extracts were utilized for the synthesis of metal and metal oxide nanoparticles- Prunus nepalensis [fruit extract for gold nanoparticles (Au NPs)] and Bauhinia purpurea [leaf extract for magnesium oxide (MgO) nanoflakes]. Then the detailed characterizations of the biogenic Au NPs and MgO nanoflakes were performed. The average size of the synthesized Au NPs and MgO nanoflakes was found to be 6 ± 1 nm and 11 nm (width), respectively, with a crystalline structure. Electron microscopy was used to investigate the morphology of the biogenic nanoparticles. Additionally, the presence of antioxidants, phenolics, and flavonoids in B. purpurea leaf extract and P. nepalensis fruit extract was studied by using different assays, which suggested the efficacy of leaf and fruit extracts as a potential reducing agent for nanoparticle synthesis. Antimicrobial activity of positively charged (+2.70 mV) MgO nanoflakes was investigated against Staphylococcus aureus, a gram-positive bacteria known to cause various infections in humans. Results suggested the high efficacy of MgO nanoflakes as a potential antibacterial agent against S. aureus at a meagre dose size of 250 μg/mL. Possible mode of action was investigated through surface morphology analysis of bacterial cells by field emission scanning electron microscopy (FESEM), and intracellular reactive oxygen species (ROS) generation by fluorescence microscopy. Another main aim of the study was to utilize biologically synthesized nanoparticles in biosensing applications for the detection of foodborne pathogens. Characterization of the peroxidase mimicking activity of the negatively charged (-20.6 mV) biogenic Au NPs in x comparison with Horseradish peroxidase (HRP, a natural enzyme), was performed. Biogenic Au NPs showed 9.64 times higher maximum reaction velocity following Michaelis–Menten kinetics at 6% H2O2 with a higher affinity towards its substrate, 3,3′,5,5 -Tetramethylbenzidine (TMB) than that of HRP. The reaction velocity was also dependent on the environmental pH, temperature, substrate, and H2O2 concentration. The Michaelis–Menten constant (Km) for biogenic Au NPs and HRP was found to be 6.9×10-2 mM and 7.9×10-2 mM, respectively, at 6% H2O2. Based on this phenomenon, two different biosensing assays were developed for the detection of Mycobacterium bovis [through indirect enzyme linked immunosorbent assay (iELISA)] and Listeria monocytogenes [through recombinase polymerase amplification (RPA)-based paper disc biosensing assay]. As a proof-of-concept for suitable biosensing applications with biogenic Au NPs, an immunoassay was developed to detect M. bovis cells using biogenic Au@1F11conjugates where Au@1F11acted as peroxidase mimic and signalling tag. It was found that the biogenic Au@1F11conjugates were capable of sensitive detection of M. bovis within a range up to 105 CFU/mL with high specificity. Then RPA based-paper disc biosensing assay was performed for the detection of L. monocytogenes where Listeriolysin O (LLO), a secreted protein encoded by hlyA gene was targeted for the detection assay. Oligonucleotide conjugated biogenic Au NPs (Au@ LM122-DP-Amino) showed intrinsic peroxidase-like activity, which acted as a detection probe for the biosensing assay. The developed RPA-paper disc biosensing assay system for L. monocytogenes detection demonstrated high sensitivity for the target pathogen. It was able to detect as low as 2 CFU/mL, or 10 fg of the genomic DNA/ 50 μL reaction. Peroxidase mimicking Au NPs played a significant role in amplifying the detection signal, which could be visualized by naked eye and could be utilized in further developments of point-of-care (POC) detection assays for various analytes, such as pathogen, disease biomarkers, and toxins for further applications
Exploring Antimony-Based Nanostructured Hybrids for Alkali Metal-Ion Batteries Anode and Visible-Light-Driven Photocatalytic Dye Degradation
No full textOver the last few decades, energy and environmental problems have increased manifolds owing to various socio-economic reasons. However, an uninterrupted supply of electricity is essential for modern lifestyles due to a surge in demand for electronic gadgets and the emergence of hybrid electric vehicles in the 21st century. In this regard, alkali metal-ion batteries viz. rechargeable Li-ion batteries (LIBs) are still considered the ‘holy grail’ for energy storage applications, starting from portable electronics to EVs. However, the exorbitant cost of lithium compounds and the dilapidated global reserve is a severe threat to the future of LIBs. In the recent past, conceptually similar sodium-ion batteries (SIBs) have been proposed as a low-cost alternative to the existing LIBs in view of the abundance and wide distribution of the sodium-bearing compounds on earth. However, it is crucial to develop high-capacity anodes for practical deployment and realization of SIBs to reach their full potential. In addition, the globalization of textile industries has witnessed adverse effects on the environment due to water pollution from effluent waste-water containing hazardous organic toxic dye. The treatment of toxic organic effluents is highly desirable for the preservation of clean water. Therefore, it is paramount to develop low-cost, visible-light-driven photocatalysts to degrade these toxic pollutants. The present work tries to make an attempt to identify and explore Sb-based materials as a potential solution to the global energy storage and water pollution problem mentioned above by designing and developing Sb/C hybrid LIBs/SIBs anodes and Sb2S3 based compounds for visible-light-driven photocatalyst. Among various materials explored to date, nanostructured antimony-based (Sb) hybrid compounds are considered a promising anode for LIBs/SIBs and solar-light-driven photocatalyst towards dye degradation owing to its attractive electrical and electronic properties. However, Sb-based materials suffer from poor electrochemical performance and photocatalytic activity due to large volumetric variance during (de)alloying with Li/Na and fast recombination rate of photogenerated electron-hole pairs, respectively. Sb-based anodes have two major disadvantages includes significant volume swelling (~135%-293%) and detachment of active phase from the current collector upon (de)alloying with Li and Na. One way to improve the electrochemical performance in LIB/SIBs and photocatalytic activity of Sb/Sb2S3 based materials is hybridizing with carbon and analog and developing myriads of nanostructured compounds. Carbonaceous materials viz. reduced graphene oxide (rGO) and resorcinol-formaldehyde (RF) derived porous carbon are particularly attractive for Sb-based carbonaceous composite formulation for LIBs/SIBs anodes and photocatalytic degradation of industrial dyes. In the present work, Sb-based nanostructured (SbNPs, Sb/rGO, core-shell Sb@C, yolk-shell Sb@void@C, Sb2S3 nanorods, Sb2S3/rGO, Sb2S3-NPs, core-shell Sb2S3@C, Sb2S3-NPs/rGO, and Bi1.09Sb0.91S3/rGO) anodes have been successfully synthesized via facile hydrothermal and sol-gel routes. In addition, Sb2S3 nanorods, Sb2S3 nanorods/rGO, Sb2S3-NPs, Sb2S3-NPs/rGO, Sb2S3-NPs@C, and BixSb2-xS3 [x = 0.536, 1.09, 1.68] narrow bandgap photocatalysts have been explored for photocatalytic degradation of Rhodamine-B (RhB) dye. Various analytical tools viz. XRD, FTIR, TG, Raman spectroscopy, FESEM, TEM, XPS, and BET surface area analysis were utilized to determine phase, structure, morphology, valence, and specific surface area of as-synthesized Sb-based nanostructured hybrid. The effect of varying rGO content, the role of porous carbon coating on the stable electrochemical performance of Sb-based anodes were investigated in detail using LIBs and SIBs half-cell and full-cell. The electrochemical results show that carefully chosen optimal amount of rGO and ultrathin porous carbon improves electrochemical performance due to fast charge-transfer, minimum solid electrolyte interphase (SEI) resistance across electrode/electrolyte interphase, validated by EIS, Li+/Na+ diffusion coefficient, ex-situ FESEM, XRD, and XANES analysis of the cycled electrodes. In addition, binder-free electrodes fabricated by electrophoretic deposition (EPD) in quick time (~3 min) showed stable electrochemical performance compare to the conventional doctor blade technique offering a useful strategy for preventing Sb delamination. The dissertation work successfully demonstrates that Sb-based SIBs anodes would be suitable candidates for future energy storage needs, and narrow band gap Sb2S3 and BixSb2-xS3-based photocatalysts exhibiting excellent photocatalytic activity would be potentially suitable for waste-water treatments
VLSI Circuits for Orthogonal Matching Pursuit Algorithm with Performance Trade-offs
No full textThe conventional Shannon-Nyquist sampling theory sets the goal for a signal to be sampled at a rate twice the highest frequency component present in the signal. This requires high sampling Analog to Digital Converter (ADC) for measuring high frequency signals like Radio Frequency (RF) signals. Compressed Sensing (CS) is a recently developed sampling paradigm which proves to be an alternative to the Nyquist sampling requirements. CS exploits the sparsity property of signals and enables sparse signals to be acquired and measured at sub-Nyquist rate. CS is achieved by means of two processes, compressive sampling and signal recovery. In the compressive sampling step, random linear measurements are generated from the input signal by some hardware like Analog to Information Converter (AIC) or Random Modulation Pre-Integrator (RMPI). The original signal is recovered from the measurement samples by use of recovery algorithms in the second step. Out of these algorithms Orthogonal Matching Pursuit (OMP) algorithm is very popular and widely used for its moderate complexity and better performance. The major steps of OMP algorithm are Atom Searching (AS) and Least Squares (LS) evaluation. In the AS step, a column is identified from the measurement matrix which is most correlated to the current residual and a least squares problem is solved to find the signal estimate in the LS step. In this work, different Very Large Scale Integration (VLSI) circuits for OMP algorithm are proposed and the performance indices like hardware complexity, recovery speed, power consumption and Recovery Signal to Noise Ratio (RSNR) are estimated to demonstrate the circuit trade-offs. The architectures of OMP algorithm differ in the technique of solving the LS problem. In this work, three different architectures are proposed to implement the OMP algorithm on Field Programmable Gate Array (FPGA) which are termed as architectures 1, 2 and 3. Architecture 1 solves the LS problem using pseudoinverse computation based on QR Decomposition (QRD). Partial QRD is evaluated by modified Gram-Schimdt algorithm. Architecture 2 is an improved version of the previous one and uses same QRD but avoids pseudoinverse computation. Architecture 2 implements a novel Low Power Fast OMP (LPF-OMP) algorithm which is proposed in this work. LPF-OMP algorithm, an improved version of classic OMP algorithm, has high recovery speed compared to other versions of OMP algorithm. QRD based architectures are suitable for low sparse signals where the resource utilization loosely depends on signal sparsity but suffers from low RSNR performance. Architecture 3 solves the LS problem by Gaussian Elimination (GE) to achieve high RSNR performance by keeping other performances within acceptable limit. A novel Incremental Gaussian Elimination (IGE) technique is proposed which efficiently implements GE. In this work, Gaussian test pulses for Radio Detection And Ranging (RADAR) application and multi-tone signals are taken as the input signals to validate the designs. The random measurements from the input signals are taken using the MATLAB-Simulink model of the AIC or RMPI. The proposed designs are targeted to Virtex6 FPGA device to compare with other reported works for K = 256, N = 1024 and Im = 36 where K is the measurement vector length, N is the length of the signal and Im is the maximum number of iterations. Sparsity of RADAR pulses is unknown a priori and thus OMP algorithm can be stopped after maximum number of iterations or when a stopping criteria is fulfilled. A novel stopping criteria is also proposed in this work. Classic OMP can recover an m-sparse signal in Im = m iterations but LPF-OMP may run for extra few iterations for correct recovery. RSNR performance is measured after Im = m = 36 number of iterations for comparison. Original signal reconstruction is also an objective of this research and thus a novel Look Up Table (LUT) based approach is proposed to generate the elements of Gabor dictionary. Reconstruction of the practical signals is demonstrated by implementing the architectures on Artix7 FPGA device. Application Specific Integrated Circuit (ASIC) implementation of OMP algorithm is also attempted in this work as a part of Linear Frequency Modulated Continuous Wave (LFM-CW) RADAR receiver for detection of stationary targets. Two new ASIC realizations are proposed and they are termed as architectures 4 and 5. These architectures implement the LPF-OMP algorithm to achieve fast reconstruction speed and use IGE technique to achieve better RSNR performance. These architectures are modified version of architecture 3 and are implemented for parameters K = 64, N = 256 and Im = 16 using SCL180 nm standard cell library provided by Semi-Conductor Laboratory (SCL) India. OMP implemented on ASIC provides performance advantage over FPGA implementations and the proposed ASIC realizations are hardware efficient compared to other reported ASIC implementations. Improvement of the range solution of the CS based LFM-CW RADAR is also attempted here by using partial Fourier dictionar
Polymer and Glass-Ceramics Composite Scaffolds for Bone Tissue Engineering
No full textThe use of conventional bone repair techniques involving bone grafts has limited potential because of the likelihood of disease transmission and because of their limited supply. Bone tissue engineering is a potential alternative to overcome these limitations, which enables implant scaffolds to stimulate tissue formation directly in-vivo. It is a technique aimed at induced bone regeneration by combining different biomaterials and other functional materials. Arrays of biomaterials have been used in bone tissue engineering due to their attractive properties, such as osteo-conductivity and biocompatibility. The widely used materials include biodegradable polymers and ceramics such as hydroxyapatite, bioactive glasses and glass ceramics.
Natural polymers like alginate, gelatin and chitosan are attractive biomaterials because of their close resemblance with extra-cellular matrix components, thus making them highly biocompatible. However, these polymers suffer from poor bioactivity and low strength. Making a composite scaffold using the bioactive glass-ceramics and polymer can improve the toughness and mechanical stability of the scaffold while maintaining good bioactivity and biocompatibility.
The glass-ceramics (GC) in the system P2O5-Na2O-CaO-SiO2 was synthesized using sol gel technique. The GC showed to contain combeite crystalline phase which is important for bioactivity and bio-dissolution of the GC. The GC showed good bioactivity and biocompatibility. The GC powder was used in the fabrication of composite scaffold. Two types of scaffolds were fabricated where GC was used; (a) as filler in a polymer (chitosan, alginate and gelatin) matrix scaffolds prepared by freeze drying method, and (b) as a ceramic skeleton in the sponge replicated GC scaffolds those coated with three different polymers.
The composite scaffolds prepared by freeze drying showed highly macroporous (~80%) structures with good pore connectivity which is ideal for bone tissue engineering application. In general, composite scaffolds showed the improvement of different properties such as mechanical strength, bioactivity, cell adhesion, alkaline phosphatase activity and osteogenicity with increase in GC powder loading in polymer matrix of the scaffold. Among all, gelatin based scaffolds showed highest compressive strength exhibiting 5.8 MPa in the scaffold containing 5 wt% GC. Gelatin based scaffolds also showed excellent cell adhesion and proliferation compared to the other. Alginate-chitosan-GC based scaffolds showed improved strength compared to chitosan-GC based scaffolds and better cell proliferation compared to alginate-GC scaffolds.
Sponge replication was used to prepare GC scaffolds doped with 6 wt% titania through sintering of GC. The strength of the porous GC scaffold was improved by doping with titania which acts as a sintering additive. The strength of the scaffolds was further improved by coating with alginate, gelatin and chitosan. Among all, gelatin coated GC scaffolds showed highest compressive strength and relatively better cell adhesion properties. Microtomography was used for in-detail examination of pore characteristics in scaffolds. The scaffolds exhibited highly porous architecture with open pores. A clear understanding of pore size distribution was obtained. The scaffolds showed macoporous structure ideal for bone tissue engineering. The pore sphericity, orientation and interconnectivity were analyzed in details. The results demonstrated that the composite scaffolds are potential candidates for bone tissue engineering
A Study on Singular and Non-singular Elliptic and Parabolic PDEs
No full textThe thesis is concerned about singular elliptic, parabolic partial differential equations (PDEs) and non-singular elliptic PDEs. A problem having nonlinearities that blow up to infinity near zero is considered as a ‘singular problem’. Most of the problems in the thesis are involved with nonlocal elliptic operators such as the fractional order Laplacian or p-Laplacian with constant or variable exponents. The objective of the thesis is to investigate the existence and multiplicity/unicity of weak solutions to PDEs involving a singular nonlinearity or/and critical nonlinearities or/and an L1/measure datum, assigned with Dirichlet boundary conditions. The approach varies from problem to problem. Variational techniques comprising of the variants of mountain pass theorem, the Ekeland’s variational principle are the commonly used techniques in the thesis. The PDEs involving a singular nonlinearity, a Radon measure or an L1 function are approached by the weak convergence method. The concentration compactness principles and the method of layer potentials are used to tackle the problems with critical exponents and the problems with mixed boundary value type respectively. A concentration compactness type principle in two critical exponent set up is also derived. The solutions are found to be in fractional or integer order Sobolev spaces with constant or variable exponents. Furthermore, various notions of solutions like entropy solution, very weak solution, etc. are also introduced for a few singular problems. The presence of irregular data like L1/measure data affects the regularity of solutions, i.e. the solutions are obtained as limits of approximations (SOLA) with a lesser degree of differentiability or/and integrability