25 research outputs found
Generation of mid-infrared laser using solid state laser and nonlinear optical frequency conversion
No full textCoherent MIR radiation has several applications such as remote sensing, laser surgery, nonlinear spectroscopy, countermeasures, etc. Generation of coherent MIR radiation is achieved by solid state laser and quantum cascade laser with direct MIR lasing, or nonlinear frequency conversion from laser with shorter wavelength. In this thesis, we developed several different laser configurations to access the MIR wavelengths.
Firstly, a high repetition rate high power Ho:YAG laser was developed to generate high average power MIR ns pulse in an intra-cavity ZGP OPO configuration. A high energy Ho:YAG MOPA laser pumped ZGP crystal in an extra-cavity OPO configuration was demonstrated. A novel uncoated wedge ZGP OPO configuration was studied, addressing the AR-coating limitation encountered in conventional ZGP OPO.
Secondly, Coherent Polarization Locking configuration was implemented in Ho:YAG oscillator. Studies show the Coherent Polarization Locking configuration improves the beam quality in CW operation and reduces the risk of intra-cavity optical damage in pulse operation. Improvement in beam quality is due to the reduction of thermal related effects in the oscillator. Splitting the intra-cavity intensity in the Coherent Polarization Locking cavity reduces the possibility of intra-cavity optical damage.
Lastly, a high energy Yb:CaF2 regenerative amplifier was developed as the pump source for a multi-stage ZGP OPA setup to generate fs MIR laser pulses. To increase the efficiency of fs MIR generation, a high energy Ho:YAG regenerative amplifier was explored as the direct pump source for ZGP OPA system.DOCTOR OF PHILOSOPHY (SPMS
Nonspherical particle suspensions in wall turbulence
No full textIn the present PhD thesis, we investigate the dynamics of small non-spherical particles suspended in a fully developed turbulent channel flow. Non-spherical particles are approximated as axisymmetric spheroidal and triaxial ellipsoidal particles, which are characterized by their inertia and shape. As a starting point, the rotational dynamics of an inertial oblate spheroidal particle suspended in a creeping linear shear flow is investigated, which is later used in interpreting the results obtained in complex turbulent flows. The three-dimensional turbulent flow field is obtained from the Navier-Stokes equations by means of direct numerical simulation in an Eulerian reference frame. The particles are tracked using Lagrangian point particle approach. Existing methodology to simulate the prolate spheroidal particles in wall-bounded turbulent channel flow is extended to investigate the dynamics of oblate spheroidal particles and triaxial ellipsoidal particles. The effects of particle inertia, particle shape, and fluid shear on particle rotation and orientation are reported with respect to both fluid and particle reference frame. Particles at the channel center and near the wall show different rotation patterns and surprisingly different effects of particle inertia. Finally, we report the systematic investigation of the gravity force effects on the dynamics of prolate spheroidal particles
Measuring spin polarization of Cu-doped bulk indium tin oxide using point contact Andreev reflection method.
No full textSpin polarization of a novel material is of interest to researcher in the field of spintronic devices. Direct measurement of the spin polarization of a novel material using Point Contact Andréev Reflection (PCAR) method is presented. PCAR method is realized using the home-build point contact device.Bachelor of Science in Physic
Supervised and unsupervised learning for 3D brain image segmentation
No full textM.Phil.Supervised and unsupervised automatic brain structure segmentation play pivotal roles in diagnosing brain diseases, including brain tumors, neuropsychiatric disorders and so on, thus are important in the clinical practice. However, developing an automatic approach to segment the brain structures remains very challenging in both supervised and unsupervised learning area. Supervised methods suffer from capturing the detail of complex brain structures, i.e., the ambiguous boundary, complex anatomical structures, and large variance in shapes of brain tissues. While for unsupervised segmentation methods, current registration-based segmenting architectures often fail to capture the hidden mapping relationship from unaligned image to reference image, thereby reducing the discriminativeness of learned features. In this thesis, to alleviate the challenging problems, we propose two deep-learning-based methods for supervised and unsupervised learning, respectively.First of all, we focus on the supervised learning method for 3D brain structure segmentation. We present a novel deep convolutional neural network ( -Net), which aims at boosting the information flow and enhancing the feature propagation. Specifically, we formulate a densely convolutional LSTM module (DCLSTM) to selectively aggregate the convolutional features (with the same spatial resolution) at the same stage of a CNN, which promotes the discriminativeness of features at each CNN stage. By stacking multiple DC-LSTMs among different stages, we can progressively and selectively aggregate the highly semantic features in deep stages to refine the fine detail features in shallow stages for better brain segmentation. Extensive experiments on two public benchmark datasets on sub-cortical brain structure segmentation validates the efficacy and effectiveness of our proposed network.Second, towards unsupervised segmentation, we study segmentation based on registration architecture, which transfers one brain image into the coordinate system of another brain image with the matched imaging contents. By employing this transformation relation between unlabeled images (unaligned images) and one brain volume with annotation (reference image), we can easily transfer the segmentation mask of the reference image to all unlabeled brain images. Specifically, to capture the internal transformation relation between a pair of 3D brain images, we develop a novel probabilistic multilayer regularization network for unsupervised brain image registration and segmentation. We first introduce direct regularizations into the hidden layers of two deep convolutional neural networks (CNNs) by adopting probabilistic models between hidden layers in these two CNNs. Furthermore, we embed the regularization terms into multiple layers of the CNNs and produce the feature-level latent variables in different layers. Combining the predicted feature-level latent variables of all layers, we can obtain a transformation metrics with rich information for a robust 3D brain image registration and segmentation. We employ two common benchmark datasets for unsupervised 3D brain image segmentation. Experimental results show that our method clearly outperforms state-of-the-art methods on both benchmark datasets by a large margin.監督式和無監督式的自動腦結構分割在診斷腦疾病的過程中起著關鍵作用(例如腦腫瘤, 神經精神疾病等), 因此該類方法在臨床實踐中具有重要意義。然而, 在監督式和無監督式學習領域中, 開發一種自動分割大腦結構的方法仍然非常具有挑戰性。監督式學習的方法難以捕獲复雜的大腦結構的細節, 例如腦溝模糊的邊界, 復雜的解剖結構以及大腦組織形狀的巨大差异。而當前基於配準架構的無監督分割模型通常無法捕獲從未對齊圖像到參考圖像的隱藏映射關系, 從而降低了用於圖像配準和圖像分割的特徵的判别能力。为了解决以上問題, 本文提出了兩種基於深度學習的監督式和無監督式的學習方法。首先, 針對於3D 腦結構分割的監督學習方法, 我們提出了一種新穎的深度卷積神經網絡( -Net) 來促進神經網絡的信息流和特徵傳播。具體來说, 我們設計了密集卷積長短期記憶模塊(DC-LSTM), 該模塊在CNN 的同一階段選擇性地聚合具有相同的空間分辨率的卷積特徵從, 而達到促進了每個CNN階段聚合特徵的判别能力。基於此, 通過在不同CNN 階段之間堆砌多個DC-LSTM, 我們可以逐步有選擇地聚合深層CNN階段的高度語義特徵, 用聚合的特徵來精煉更具精細細節的淺層CNN 階段的特徵, 以此實現更好的腦部分割。我們在兩個關於大腦皮層下結構的公共基準數據集上進行的大量實驗, 實驗結果驗證了我們提出的網絡的有效性和高效性。其次, 對於無監督式的分割方法, 我們研究方向主要針對基於配準架構的分割方法。該類分割方法主要涉及到將一個大腦圖像轉移到另一個大腦圖像的坐標系的配準的過程。通過應用配準過程中的没有分割結果的圖像(未對齊圖像) 和帶有分割結果的圖像(參考圖像) 間的轉化關系, 我們可以輕鬆地將參考圖像的分割結果轉化到所有没有分割結果的大腦圖像上。具體來说, 为了捕獲一對3D 腦圖像之間的内部轉化關系, 我們提出了一種新穎的多層概率正則化模型來對腦圖像進行無監督配準和分割。。我們首先在兩個深度卷積神經網絡的隱藏層之間建立概率模型, 將概率正則化模型直接加入這兩個深度卷積神經網絡的隱藏層。此外, 我們將概率正則化模型項嵌入到CNN的多個階段中, 然後在不同CNN 階段中生成一個特徵級的隱藏變量。結合所有階層預測出來的特徵級的隱藏變量, 我們可以獲得一個具有更豐富轉化信息的轉化指標, 基於此指標來達到更爲魯棒的3D 腦圖像的無監督式配準和分割。我們在兩個常見的基準數據集進行無監督的3D 腦圖像配準和分割。實驗結果表明, 在兩個數據集上, 我們的方法明顯優於最新方法。Liu, Lihao."December 2019."Thesis M.Phil. Chinese University of Hong Kong 2020.Includes bibliographical references (leaves 74-85).Abstracts also in Chinese.Title from PDF title page (viewed on February 17, 2022)
Membrane mechanisms during exo- and endocytosis in excitable cells
No full textExcitable cells, like endocrine cells and neurons release hormones and transmitters through exocytosis to mediate many important functions, such as stress responses, immune responses, control of blood glucose, and synaptic transmission that mediates communication between neurons in the brain. This process relies on delicate endocytic mechanisms, to retrieve vesicular membranes and proteins previously fused at the plasma membrane and thus maintains the ability of exocytosis for excitable cells.Upon exocytosis, vesicle fusion with the plasma membrane generates an Ω-shape membrane profile with a narrow fusion pore (Clathrin-mediated endocytosis (CME), the most common form of endocytosis, is considered to retrieve vesicles from the flat plasma membrane. With conventional and platinum replica electron microscopy (EM), we found that clathrin-coated pits prefer to be localized at the bulk invagination of the plasma membrane instead of the flat plasma membrane. Using super-resolution STED microscopy, we observed for the first-time direct budding of vesicles from bulk membrane invagination. These results suggest that bulk membrane invaginations are a major platform for mediating clathrin-dependent endocytosis, calling for modification of the current view that clathrin-mediated endocytosis mostly originated from the flat plasma membrane. In summary, the studies described here improve our understanding of the regulation of hormone and transmitter release as well as endocytosis.List of scientific papersI. GIANVITO ARPINO, Lihao Ge, Shin Wonchul, Seth Villareal, Uri Ashery, Agila Somasundaram, Justin Taraska, Oleg Shupliakov, Ling-Gang Wu. Bulk invagination at the plasma membrane is the primary site for clathrin-mediated endocytosis. [Manuscript]II. Peter J. Wen*, Staffan Grenklo*, GIANVITO ARPINO, Xinyu Tan, Hsien- Shun Liao, Johanna Heureaux, Shi-Yong Peng, Hsueh-Cheng Chiang, Edaeni Hamid, Wei-Dong Zhao, Wonchul Shin, Tuomas Näreoja, Emma Evergreen, Yinghui Jin, Roger Karlsson, Steven N. Ebert, Albert Jin, Allen P. Liu, Oleg Shupliakov and Ling-Gang Wu. Actin dynamics provides membrane tension to merge fusing vesicles into the plasma membrane. Nature Commun. 2016, 7, p. 12604. *Contributed equally to this work. https://doi.org/10.1038/ncomms12604 III. Wonchul Shin, Lihao Ge, GIANVITO ARPINO, Seth Villareal, Edaeni Hamid, Huisheng Liu, Wei-Dong Zhao, Peter J. Wen P, Hsueh-Chiang and Ling-Gang Wu. Visualization of Membrane Pore in Live Cells Reveals a Dynamic-Pore Theory Governing Fusion and Endocytosis. Cell. 2018, 4, p. 934. https://doi.org/10.1016/j.cell.2018.02.062 IV. Wonchul Shin*, GIANVITO ARPINO*, Sathish Thiyagarajan*, Rui Su*, Liaho Ge*, McDargh Zachary, Xiaoli Guo, Lisi Wei, Oleg Shupliakov, Albert Jin A, Ben O'Shaughnessy B, Ling-Gang Wu. Vesicle Shrinking and Enlargement Play Opposing Roles in the Release of Exocytotic Contents. Cell Reports. 2020, 30, p. 421. *Contributed equally to this work. https://doi.org/10.1016/j.celrep.2019.12.044 </p
Study on the inhibitory mechanism of epigallocatechin gallate against foodborne pathogens and its application
No full textObjectiveThe bacteriostatic mechanism of epigallocatechin gallate (EGCG) against Staphylococcus aureus (S. aureus) and Salmonella typhimurium (S. typhimurium) and the elimination effect of S. aureus and S. typhimurium in purple yam chip were studied.MethodsThe bacteriostatic effect of EGCG in vitro was determined by microbroth gradient dilution method, and the bacteriostatic mechanism of EGCG was further determined by potassium ion determination, PI staining, OD260 nm value, OD280 nm value and cell morphology observation. Finally, the clearance rate of S. aureus and S. typhimurium in purple yam chip was studied by EGCG combined with pasteurization.ResultsThe minimum inhibitory concentration (MIC) of EGCG for S. aureus and S. typhimurium was 0.512 mg/mL and 0.256 mg/mL, and the minimum bactericidal concentration was 1.024 mg/mL and 0.512 mg/mL, respectively. After MIC and 2MIC treatment, the K+ concentration of S. aureus increased to 2.26 mg/L and 2.39 mg/L, respectively, while after MIC and 2MIC treatment, the K+ concentration of S. typhimurium increased to 1.79 mg/L and 1.84 mg/L, respectively. All strains were higher than those without EGCG treatment. EGCG could destroy the cell membranes of S. aureus and S. typhimurium, promote the increase of membrane permeability, and then appear the phenomenon of cell contents leakage and cell shrinkage, resulting in bacterial death. The clearance rate of S. aureus and S. typhimurium in purple yam chip was 100% after treatment with EGCG at MIC combined with pasteurization.ConclusionEGCG has antibacterial effect on foodborne pathogens in vitro, and can be used as a natural antibacterial preservative in food
Newton-ADE FDTD Method for Time-Varying Plasma
No full textA novel finite-difference time-domain (FDTD) method is proposed in this article for electromagnetic (EM) wave propagation in time-varying plasma. It is formulated based on the fundamental Newton’s equation of motion, which governs the relationships among the time-varying electron density, current density and electric field. Utilizing the auxiliary differential equation (ADE) FDTD scheme, the method is aptly referred to as the Newton-ADE FDTD method for time-varying plasma. Traditionally, the previous ADE FDTD methods directly apply the time-varying electron density for the plasma frequency in the update equations of current density and electric field. This is inadequate and incorrect in general time-varying plasma conditions. The formulation of the Newton-ADE FDTD method is provided and compared to that of the traditional ADE FDTD method. It is found that the key difference lies in the new compact term being the time-derivative of logarithm of electron density. Two discretization schemes for this term are provided. The Newton-ADE FDTD method is validated based on the matrix exponential method. Novel stability and convergence analyses are provided for the proposed method in time-varying dispersive media. These analyses show that our proposed method is stable and achieves second-order temporal accuracy. Numerical results are presented for various time-varying plasma conditions. It is demonstrated that when the absolute value for time-derivative of logarithm of electron density is large, e.g., greater than the collision frequency, the Newton-ADE FDTD method can provide correct results, while the traditional ADE FDTD method yields significant differences.Ministry of Education (MOE)Submitted/Accepted versionThis work was supported in part by the National Natural Science Foundation of China under Grant 62201430, Grant 61627901, and Grant 62171349; in part by the Singapore Ministry of Education under Grant RG49/21; in part by the Innovation Capability Support Program of Shaanxi under Grant S2022-ZC-TD-0060; and in part by the Fundamental Research Funds for the Central Universities under Grant 20103227273
Efficient few-cycle mid-IR pulse generation in the 5-11 µm window driven by an Yb amplifier
No full textHigher anthocyanin accumulation was associated with higher transcription levels of anthocyanin biosynthesis genes in spinach
No full textSpinach (Spinacia oleracea L.) is widely cultivated as an economically important green leafy vegetable crop for fresh and processing consumption. The red purple spinach shows abundant anthocyanin accumulation in the leaf and leaf petiole. However, the molecular mechanisms of anthocyanin synthesis in this species are still undetermined. In the present study, we investigated the pigments formation and identified anthocyanin biosynthetic genes in spinach, and performed the expression analysis of anthocyanin related genes in the purple and green cultivar by quantitative PCR. Results showed that accumulation of anthocyanin was the dominant pigment resulting in the red coloration in spinach, and 22 biosynthesis genes and 25 regulatory genes were identified in spinach, based on the spinach genomic and transcriptomic database. Furthermore, the expression patterns of genes encoding enzymes indicated that SoPAL, SoUFGT3 and SoUFGT4 were possible candidate genes for anthocyanin biosynthesis in red purple spinach. The expression patterns of transcription factors indicated that two SoMYBs, three SobHLHs and one SoWD40 were drastically up-regulated and co-expression in red purple spinach, suggesting an essential role of regulatory genes in the anthocyanin biosynthesis of spinach. The results above enhanced our understanding about the molecular mechanisms of anthocyanin biosynthesis in purple spinach.The accepted manuscript in pdf format is listed with the files at the bottom of this page. The presentation of the authors' names and (or) special characters in the title of the manuscript may differ slightly between what is listed on this page and what is listed in the pdf file of the accepted manuscript; that in the pdf file of the accepted manuscript is what was submitted by the author
Multimodal critical-scenarios search method for test of autonomous vehicles
No full textPurpose – The purpose of this paper is to search for the critical-scenarios of autonomous vehicles (AVs) quickly and comprehensively, which is essential for verification and validation (V&V). Design/methodology/approach – The author adopted the index F1 to quantitative critical-scenarios' coverage of the search space and proposed the improved particle swarm optimization (IPSO) to enhance exploration ability for higher coverage. Compared with the particle swarm optimization (PSO), there were three improvements. In the initial phase, the Latin hypercube sampling method was introduced for a uniform distribution of particles. In the iteration phase, the neighborhood operator was adapted to explore more modals with the particles divided into groups. In the convergence phase, the convergence judgment and restart strategy were used to explore the search space by avoiding local convergence. Compared with the Monte Carlo method (MC) and PSO, experiments on the artificial function and critical-scenarios search were carried out to verify the efficiency and the application effect of the method. Findings – Results show that IPSO can search for multimodal critical-scenarios comprehensively, with a stricter threshold and fewer samples in the experiment on critical-scenario search, the coverage of IPSO is 14% higher than PSO and 40% higher than MC. Originality/value – The critical-scenarios' coverage of the search space is firstly quantified by the index F1, and the proposed method has higher search efficiency and coverage for the critical-scenarios search of AVs, which shows application potential for V&V
