Search

Computer audition (CA) has experienced a fast development in the past decades by leveraging advanced signal processing and machine learning techniques. In particular, for its noninvasive and ubiquitous character by nature, CA-based applications in healthcare have increasingly attracted attention in recent years. During the tough time of the global crisis caused by the coronavirus disease 2019 (COVID-19), scientists and engineers in data science have collaborated to think of novel ways in prevention, diagnosis, treatment, tracking, and management of this global pandemic. On the one hand, we have witnessed the power of 5G, Internet of Things, big data, computer vision, and artificial intelligence in applications of epidemiology modeling, drug and/or vaccine finding and designing, fast CT screening, and quarantine management. On the other hand, relevant studies in exploring the capacity of CA are extremely lacking and underestimated. To this end, we propose a novel multitask speech corpus for COVID-19 research usage. We collected 51 confirmed COVID-19 patients' in-the-wild speech data in Wuhan city, China. We define three main tasks in this corpus, i.e., three-category classification tasks for evaluating the physical and/or mental status of patients, i.e., sleep quality, fatigue, and anxiety. The benchmarks are given by using both classic machine learning methods and state-of-the-art deep learning techniques. We believe this study and corpus cannot only facilitate the ongoing research on using data science to fight against COVID-19, but also the monitoring of contagious diseases for general purpose.

We present an EO Q-switched Tm:YAP laser constructed with a periodically poled Mg-doped congruent lithium niobate (PPMgCLN) Bragg modulator. A 2-mm-thick PPMgCLN modulator was adopted, and the internal stress generated in the electric poling process was released by annealing. The peak stress-induced refractive index variance for e-wave was deduced to be 2.29 × 10-5 at 1064 nm. When the absorbed pump power and repetition rate were 13.4 W and 1 kHz, respectively, the laser had a Q-switched pulse energy of 2 mJ, a pulse duration of 60 ns, and a peak power of 33.3 kW. By varying the reflection of the output coupler, output wavelengths of between approximately 1985 and approximately 1940 nm can be selected, making this laser a potential tool for pumping mid-IR sources and biomedical applications.

In a transcription-wide association study of nasal epithelium, we identify novel and previously reported susceptibility genes for atopic asthma in children and show that gene co-expression networks differ markedly between children with and without atopic asthma.

Recent advances in next-generation sequencing technology have yielded huge amounts of transcriptomic data. The discreteness and the high dimensions of RNA-seq data have posed great challenges in biological network analysis. Although estimation theories for high-dimensional modified Poisson-type graphical models have been proposed for the network analysis of count-valued data, the statistical inference of these models is still largely unknown. We herein propose a two-step procedure in both edgewise and global statistical inference of these modified Poisson-type graphical models using a cutting-edge generalized low-dimensional projection approach for bias correction. Extensive simulations and a real example with ground truth illustrate asymptotic normality of edgewise inference and more accurate inferential results in multiple testing compared to the sole estimation and the inferential method under normal assumption. Furthermore, the application of our method to novel RNA-seq data of childhood atopic asthma in Puerto Ricans demonstrates more biologically meaningful results compared to the sole estimation and the inferential methods based on Gaussian and nonparanormal graphical models.

Lung fibrosis, or the scarring of the lung, is a devastating disease with huge unmet medical need. There are limited treatment options and its prognosis is worse than most types of cancer. We previously discovered that MK-0429 is an equipotent pan-inhibitor of all αv integrins that reduces proteinuria and kidney fibrosis in a preclinical model. In the present study, we further demonstrated that MK-0429 significantly inhibits fibrosis progression in a bleomycin-induced lung injury model. In search of newer integrin inhibitors for fibrosis, we characterized monoclonal antibodies discovered using Adimab’s yeast display platform. We identified several potent neutralizing integrin antibodies with unique human and mouse cross-reactivity. Among these, Ab-31 blocked the binding of multiple αv integrins to their ligands with IC50s comparable to those of MK-0429. Furthermore, both MK-0429 and Ab-31 suppressed integrin-mediated cell adhesion and latent TGFβ activation. In IPF patient lung fibroblasts, TGFβ treatment induced profound αSMA expression in phenotypic imaging assays and Ab-31 demonstrated superior in vitro activity at inhibiting αSMA expression, suggesting that the integrin antibody is able to modulate TGFβ action though mechanisms beyond the inhibition of latent TGFβ activation. Together, our results highlight the potential to develop newer integrin therapeutics for the treatment of fibrotic lung diseases.One Sentence Summarytargeting integrin in lung fibrosis

A paper in INTERSPEECH 2020

Set classification aims to classify a set of observations as a whole, as opposed to classifying individual observations separately. To formally understand the unfamiliar concept of binary set classification, we first investigate the optimal decision rule under the normal distribution, which utilizes the empirical covariance of the set to be classified. We show that the number of observations in the set plays a critical role in bounding the Bayes risk. Under this framework, we further propose new methods of set classification. For the case where only a few parameters of the model drive the difference between two classes, we propose a computationally-efficient approach to parameter estimation using linear programming, leading to the Covariance-engaged LInear Programming Set (CLIPS) classifier. Its theoretical properties are investigated for both independent case and various (short-range and long-range dependent) time series structures among observations within each set. The convergence rates of estimation errors and risk of the CLIPS classifier are established to show that having multiple observations in a set leads to faster convergence rates, compared to the standard classification situation in which there is only one observation in the set. The applicable domains in which the CLIPS performs better than competitors are highlighted in a comprehensive simulation study. Finally, we illustrate the usefulness of the proposed methods in classification of real image data in histopathology., 86 pages, 5 figures

Reasonably optimizing the operation structure of the power grid, ensuring power quality and safety to the greatest extent, and meeting user-side power demand are the primary tasks of power grid operation.In recent years, thanks to the deepening of the construction and transformation of distribution networks in various places and the promotion of typical wiring of distribution networks, the power point and grid structure of distribution networks tend to be reasonable, which makes it possible to transfer loads on a large scale through a medium voltage distribution network.In addition, the implementation of distribution automation and the continuous improvement of smart grid construction have enabled a large number of switches to complete operations in a short period of time.Intelligent and real-time operation of the distribution network has been strengthened.To this end, this paper constructs a dynamic optimization operation scheme for distribution networks that takes into account demand response(DR) and distribution network reconstruction.The article uses IEEE33 node for power flow simulation. The results show that the method described in this paper can effectively guarantee the safety and reliability of power distribution lines, reduce the cost of power network operation and maintenance, and improve the efficiency of the power grid.

Micro simulation asynchronous motor unit is used to simulate the real asynchronous motor load, which is an important component of the micro dynamic simulation laboratory of power system. The real-time and accuracy of monitoring and control of micro simulation asynchronous motor will affect the results of micro dynamic simulation experiments directly. This paper designs and implements an integrated control instrument for micro simulation asynchronous motor unit (ICIMSAM). ICIMSAM consists of data communication unit, power amplifier unit, data processing unit, power supply unit, operation execution unit and liquid crystal display unit, realizing the functions of real-time running state monitoring, automatic starting/stopping and automatic power regulation for the motor unit. Light synchronized phasor measurement unit (PMU Light) is employed to get the real-time operation information of the micro simulation asynchronous motor unit. Real-time operation information includes three-phase frequency, voltage amplitude, voltage phasor, current amplitude, current phasor, which would be sent from the PMU Light to ICIMSG through the ethernet network. Appropriate software was developed on server to communicate with ICIMSG in a two-way manner through the network. Four group of micro dynamic simulation experiments have been completed to test the functions of ICIMSAM.

As a result of a growing and aging population, as well as an increase in associated costs, there is a continual stretching of health care services worldwide. This issue is motivating researchers all over the world to optimize medical resources by utilizing digital tools explicitly addressed to health care and well-being. One of the main fields of research in this regard is artificial intelligence (AI), the endowment of machines with human-like learning, reasoning, and decision-making abilities. Combined with high penetration of sensor-based technologies—such as smartphones and wearables—in modern society, advancements in AI mean we are entering a new age of health care. Soon, we will be able to monitor vital signs and lifestyle habits, in real-time, in such a way that will help clinicians to monitor patients’ evolution and progress in a nonintrusive and remote manner. This chapter intended to be an introductory, higher-level overview, of the core concepts relating to the branch of AI known as machine learning (ML). Readers are introduced to the ML train-test pipeline and given an overview of commonly used ML algorithms. The chapter finishes by discussing challenges that need to be overcome to help fully realize the potential of ML in everyday digital health settings.

Cardiovascular diseases have a high morbidity, and remain the leading cause of mortality. In the past two decades, developing an intelligent auscultation system has attracted tremendous efforts from the field of signal processing and machine learning. We propose a novel framework based on wavelet representations and deep recurrent neural networks for recognising three heart sounds, i. e., normal, mild, and severe. The Heart Sounds Shenzhen corpus (n = 170) is used to validate the proposed method. The experimental results demonstrate the efficacy of the proposed method in a rigorous subject independent scenario, which can reach an unweighted average recall at 43.0 % (chance level: 33.3%).

Metabolic reprogramming is one of the hallmarks of cancer. Nrf2 pathway is one of the critical signaling cascades involved in cell defense and survival against oxidative stress. The significance of Nrf2 in cancer metabolism begins to be recognized. In this minireview, we focus on the Nrf2-mediated cancer metabolic reprogramming and intend to highlight the role of Nrf2 in the regulation of malignant transformation, cancer proliferation, and the development of treatment resistance via metabolic adaptations. We hope for the development of noninvasive biomarkers and novel therapeutic approaches for cancer based on Nrf2-directed cancer metabolic reprogramming in the near future.

While deep learning is undoubtedly the predominant learning technique across speech processing, it is still not widely used in health-based applications. The corpora available for health-style recognition problems are often small, both concerning the total amount of data available and the number of individuals present. The Bipolar Disorder corpus, used in the 2018 Audio/Visual Emotion Challenge, contains only 218 audio samples from 46 individuals. Herein, we present a multi-instance learning framework aimed at constructing more reliable deep learning-based models in such conditions. First, we segment the speech files into multiple chunks. However, the problem is that each of the individual chunks is weakly labelled, as they are annotated with the label of the corresponding speech file, but may not be indicative of that label. We then train the deep learning-based (ensemble) multi-instance learning model, aiming at solving such a weakly labelled problem. The presented results demonstrate that this approach can improve the accuracy of feedforward, recurrent, and convolutional neural nets on the 3-class mania classification tasks undertaken on the Bipolar Disorder corpus.

The goal of Acoustic Scene Classification (ASC) is to recognise the environment in which an audio waveform has been recorded. Recently, deep neural networks have been applied to ASC and have achieved state-of-the-art performance. However, few works have investigated how to visualise and understand what a neural network has learnt from acoustic scenes. Previous work applied local pooling after each convolutional layer, therefore reduced the size of the feature maps. In this paper, we suggest that local pooling is not necessary, but the size of the receptive field is important. We apply atrous Convolutional Neural Networks (CNNs) with global attention pooling as the classification model. The internal feature maps of the attention model can be visualised and explained. On the Detection and Classification of Acoustic Scenes and Events (DCASE) 2018 dataset, our proposed method achieves an accuracy of 72.7 %, significantly outperforming the CNNs without dilation at 60.4 %. Furthermore, our results demonstrate that the learnt feature maps contain rich information on acoustic scenes in the time-frequency domain.

This study investigates the performance of wavelet as well as conventional temporal and spectral features for acoustic scene classification, testing the effectiveness of both feature sets when combined with neural networks on acoustic scene classification. The TUT Acoustic Scenes 2017 Database is used in the evaluation of the system. The model with wavelet energy feature achieved 74.8 % and 60.2 % on development and evaluation set respectively, which is better than the model using temporal and spectral feature set (72.9 % and 59.4 %). Additionally, to optimise the generalisation and robustness of the models, a decision fusion method based on the posterior probability of each audio scene is used. Comparing with the baseline system of the Detection and Classification Acoustic Scenes and Events 2017 (DCASE 2017) challenge, the best decision fusion model achieves 79.2 % and 63.8 % on the development and evaluation sets, respectively, where both results significantly exceed the baseline system result of 74.8 % and 61.0 % (confirmed by one tailed z-test \(p

Machine learning based heart sound classification represents an efficient technology that can help reduce the burden of manual auscultation through the automatic detection of abnormal heart sounds. In this regard, we investigate the efficacy of using the pre-trained Convolutional Neural Networks (CNNs) from large-scale image data for the classification of Phonocardiogram (PCG) signals by learning deep PCG representations. First, the PCG files are segmented into chunks of equal length. Then, we extract a scalogram image from each chunk using a wavelet transformation. Next, the scalogram images are fed into either a pre-trained CNN, or the same network fine-tuned on heart sound data. Deep representations are then extracted from a fully connected layer of each network and classification is achieved by a static classifier. Alternatively, the scalogram images are fed into an end-to-end CNN formed by adapting a pre-trained network via transfer learning. Key results indicate that our deep PCG representations extracted from a fine-tuned CNN perform the strongest, 56.2% mean accuracy, on our heart sound classification task. When compared to a baseline accuracy of 46.9%, gained using conventional audio processing features and a support vector machine, this is a significant relative improvement of 19.8% (p∠.001 by one-tailed z-test).

International audience; Motivated by the encouraging results recently obtained by genera-tive adversarial networks in various image processing tasks, we propose a conditional adversarial training framework to predict dimensional representations of emotion, i. e., arousal and valence, from speech signals. The framework consists of two networks, trained in an adversarial manner: The first network tries to predict emotion from acoustic features, while the second network aims at distinguishing between the predictions provided by the first network and the emotion labels from the database using the acoustic features as conditional information. We evaluate the performance of the proposed conditional adversarial training framework on the widely used emotion database RECOLA. Experimental results show that the proposed training strategy outperforms the conventional training method, and is comparable with, or even superior to other recently reported approaches, including deep and end-to-end learning.

Spectrogram representations of acoustic scenes have achieved competitive performance for acoustic scene classification. Yet, the spectrogram alone does not take into account a substantial amount of time-frequency information. In this study, we present an approach for exploring the benefits of deep scalogram representations, extracted in segments from an audio stream. The approach presented firstly transforms the segmented acoustic scenes into bump and morse scalograms, as well as spectrograms; secondly, the spectrograms or scalograms are sent into pre-trained convolutional neural networks; thirdly, the features extracted from a subsequent fully connected layer are fed into U+0028 bidirectional U+0029 gated recurrent neural networks, which are followed by a single highway layer and a softmax layer; finally, predictions from these three systems are fused by a margin sampling value strategy. We then evaluate the proposed approach using the acoustic scene classification data set of 2017 IEEE AASP Challenge on Detection and Classification of Acoustic Scenes and Events U+0028 DCASE U+0029. On the evaluation set, an accuracy of 64.0 U+0025 from bidirectional gated recurrent neural networks is obtained when fusing the spectrogram and the bump scalogram, which is an improvement on the 61.0 U+0025 baseline result provided by the DCASE 2017 organisers. This result shows that extracted bump scalograms are capable of improving the classification accuracy, when fusing with a spectrogram-based system.

ObjectiveTo investigate the effect of salvianolic acid A and C component molecules, which are involved in drug compatibility, on inflammatory cytokine expression that affects human chemokine ligand 5 (CCL5) and chemokine ligand 10 (CXCL10) levels in rats with unilateral ureteral obstruction (UUO).MethodsFifty Sprague Dawley rats were randomly divided into five groups: normal, model, salvianolic acid A, salvianolic acid C and salvianolic acid A and C groups. The normal group was used as the control, and the other groups of rats had a UUO model established. The control group had free access to food and water, and the other groups received the corresponding drugs for 2 weeks. After the last administration, urine β2-microglobulin (β2-MG) and N-acetyl-β-D-glucosaminidase (NAG) levels were analyzed. After 24 h, all rats were sacrificed and the serum was analyzed for creatinine (Cr) and blood urea nitrogen (BUN) levels. Rat kidneys were removed, and CCL5 and CXCL10 inflammatory cytokine mRNA expression was measured using real-time fluorescent quantitative reverse transcription-polymerase chain reaction (RT-PCR). Kidney fibrosis was observed using hematoxylin-eosin (HE) staining and Masson trichrome staining.ResultsIn the salvianolic acid A and salvianolic acid C treatment groups, serum Cr and urine NAG levels were significantly lower than in the model group (both P < 0.05). In all treatment groups, urine β2-MG levels were significantly lower than in the model group (all P < 0.05). Compared with model group, the pathological changes and collagen deposition improved to varying degrees (both P < 0.05). CCL5 and CXCL10 mRNA expression decreased to different degrees compared with the model group (both P < 0.05).ConclusionSalvianolic acid A and C are component molecules of drug compatibility, and they may protect renal function and improve tubular function and renal pathology to a certain degree in UUO. This improvement may be related to a reduction in inflammatory cytokines CCL5 and CXCL10 secretion in the UUO rat kidney.

Light field, which is captured by a plenoptic camera, is always limited in its narrow field of view (FOV) by the physical size of the aperture. To break through the restriction, we propose to extend the FOV using multiview light fields. A series of light fields are acquired by translating the camera at isometric spatial positions. In contrast to previous methods, our algorithm is the first that achieves light field registration and rendering based on epipolar plane image (EPI) properties, including disparity and color consistencies. Furthermore, the aliasing caused by the under-sampling in the angular space is eliminated by synthesizing novel views in the EPI space. Experimental results on the real scene data have demonstrated the effectiveness of our algorithm.

Background and Purpose— Moyamoya disease (MMD) is a rare, genetically heterogeneous cerebrovascular disease resulting from occlusion of the distal internal carotid arteries. A variant in the Ring Finger 213 gene ( RNF213 ), altering arginine at position 4810 (p.R4810K), is associated with MMD in Asian populations. However, there are a lack of data on the role of RNF213 in patients with MMD of additional ethnicities and diasporic Asian populations. We investigate the contribution of RNF213 alterations to MMD in an ethnically diverse population based in the United States. Methods— We initially sequenced RNF213 exons 43, 44, and 45 (encoding the eponymous RING finger domain) and exon 60 (encoding p.R4810K) in 86 ethnically diverse patients with MMD. Comprehensive exome sequencing data from 24 additional patients with MMD was then analyzed to identify RNF213 variants globally. Segregation of variants with MMD and other vascular diseases was assessed in families. Results— RNF213 p.R4810K was identified in 56% (9/16) of patients with MMD of Asian descent and not in 94 patients of non-Asian descent. 3.6% (4/110) of patients had variants in the exons encoding the RING finger domain. Seven additional variants were identified in 29% (7/24) of patients with MMD who underwent exome sequencing. Segregation analysis supported an association with MMD for 2 variants and a lack of association with disease for 1 variant. Conclusions— These results confirm that alterations in RNF213 predispose patients of diverse ethnicities to MMD, and that the p.R4810K variant predisposes individuals of Asian descent in the United States to MMD.

Recycling of End of Life Vehicle (ELV) is a coming social issue in China. For the purpose of establishing recycling enterprise rationally, all the aspect should be taken into account, especially the cost of recycling. In this paper a cost model of recycling process is presented and analyzed to the disassembly and recycling process. The configuration of equipment with different mode of disassembly is also discussed.

We present a near-field waveguide probe with an embedded array of split-ring resonators (SRRs) that operate at 16.65 GHz. When the probe scans a metal surface, the magnetic flux generated from the currents on the metallic surface strongly affects the equivalent lumped parameters of the SRRs. This, in turn, affects the reflection coefficient of the feeding port leading to high sensitivity in detecting anomalies such as cracks in the surface. Experimental results indicated the feasibility of detecting cracks, as small as 25 μm in width, in metallic surfaces. Additionally, the probe achieves good resolution in detecting closely spaced cracks.

We present a ∞at lens design that provides focusing with no aberration. By proflling the refractive index of the lens to generate a spherical wavefront at the exit side of the lens, the transmitted flelds converge at a specifled focal point. The focusing is achieved using primarily the dispersion phenomenon. We show through numerical examples that focusing without aberration can be achieved at a speciflc frequency and that focusing is possible over a narrow range of frequencies providing that the dispersion is minimal. Additionally, we show that the same principle used to design the lens can be used to design ∞at re∞ectors with a focal point focusing.

In order to solve the problem of the strong ore pressure in the mining process of a shallow buried coal seam working face, the ore pressure law of a shallow buried coal seam working face was observed. According to the observation results of ore pressure, the law and characteristics of ore pressure on the working face of shallow buried coal seam are analyzed, the basic law and parameters of roof movement and the roof plate falling situation on the end face on the working face are obtained, and the rationality of the support is studied.

A design of electronic laser harp based on SCM (the abbreviation for single chip microcomputer) is proposed. It adopts the semiconductor laser transmitter and the receiving tube as a sensor, and uses the laser transmitter emits beams to imitate the 8 strings of harp. The main control chip STC89C52RC collects and processes the signal from receiving tube circuit, then STC89C52RC outputs special signal to produce sound of harp by audio. Compared with the traditional method, this design greatly enhances the anti-interference ability of the system, and it owns a certain degree of innovation and creation in not only the form of performance but also design proposal of system.

This paper presents the University of Passau’s approaches for the Multimodal Emotion Recognition Challenge 2016. For audio signals, we exploit Bag-of-Audio-Words techniques combining Extreme Learning Machines and Hierarchical Extreme Learning Machines. For video signals, we use not only the information from the cropped face of a video frame, but also the broader contextual information from the entire frame. This information is extracted via two Convolutional Neural Networks pre-trained for face detection and object classification. Moreover, we extract facial action units, which reflect facial muscle movements and are known to be important for emotion recognition. Long Short-Term Memory Recurrent Neural Networks are deployed to exploit temporal information in the video representation. Average late fusion of audio and video systems is applied to make prediction for multimodal emotion recognition. Experimental results on the challenge database demonstrate the effectiveness of our proposed systems when compared to the baseline.

Most of the previous microwave near-field probes and imaging techniques focused on surface imaging, providing ultrahigh lateral resolution. Few microwave near-field probes were developed for subsurface detection, offering both high lateral and depth resolution with varying degrees of effectiveness. In this work, a novel microwave near-field probe using a single split-ring resonator is introduced with the primary focus of subsurface detection. The design is simple, compact, inexpensive, and easy to fabricate using printed circuit board technology. Fourier spatial analysis of the field of the new probe reveals a substantial enhancement of the evanescent field, thus making a significant difference in subsurface detection. Experimental results illustrate that a small 3.24-mm aluminum block immersed in 1% sodium chloride (NaCl) solution and positioned 4 mm away from the surface was successfully detected using a probe operating at 1.218 GHz. For this particular experiment, where the size of the object was λ/74 , the detection ability of the new probe was tested using 2% and 3% NaCl solution as well. The phase changes due to the depth of the object demonstrate that the new probe is able to sense the presence of the same object in very lossy medium (3% NaCl whose loss tangent is approximately unity) with depth of 1-2 mm in spite of a standoff distance of 1-mm air and a container thickness of 6.35 mm.

Recently, it was shown that single negative media can significantly enhance the sensitivity of near-field probes. Inspired by this recent finding, a new near-field probe is proposed for noninvasive subsurface detection. The new probe uses a single split ring resonator (SRR) instead of a periodic arrangement of SRRs for negative material realization. Experimental tests were conducted to detect the presence of cracks on aluminum plates, the presence of small aluminum blocks located behind a layer of ground chicken (lossy medium), and the presence of a small aluminum block submerged in sodium chloride solution (lossy medium) with a salinity of 1%. Preliminary results show that the proposed near-field probe enables detection of electrically small targets buried in lossy media.

(2007). Renewal Application Research in Design of Suspension Bridge Stiffened by Steel Slab Girder. Structural Engineering International: Vol. 17, No. 3, pp. 214-216.

We propose a novel probe for detecting magnetostatic surface resonant spheres. The probe is composed of a two-loop resonator in which the inner loop is loaded with a capacitor whose value can be adjusted to achieve probe resonance at desired frequency. This probe produces substantial variation in its reflection coefficient in presence of a magnetostatic resonant sphere with negative real permeability, whereas non-resonant spheres do not yield any significant effect. The mechanism behind this is explained and can be generalized to detect magnetostatic resonant spheres in low microwave frequencies. Probe design is presented along with full-wave numerical simulations.

This paper formulated and extended Mie theory to spherical particles with negative permeabilities or negative coatings. The analytic solution was validated by comparison with the numerical solution obtained using the commercial full-wave simulation tool HFSS. The results presented here are expected to help understand the interesting and recently studied phenomenon of magneto-static resonance.

This paper presents a BackPropagation Neural Network (BPNN) approach to predict the location of a buried object. A small circuit board is buried in sand at three different location levels and an electrically-small probe is used as a detection sensor. The standard deviation of the phase of reflection coefficient is used as input for the Neural Network (NN) while the location levels of the small circuit are designated as the output of the network. The network shows an accuracy of more than 90% in predicting the location of the buried circuit.

The purpose of this study was to determine whether compound zhi zhu xiang (CZZX) exerts anxiolytic-like effects in rats. The animals were orally administered CZZX (0.75, 1.5, and 3 g/kg daily) for 10 days and tested in the elevated plus maze (EPM), Vogel conflict test (VCT), and open field. Repeated treatment with CZZX (3 g/kg/day, p.o.) significantly increased the percentage of both entries into and time spent on the open arms of the EPM compared with saline controls. In the VCT, repeated treatment with CZZX (1.5 and 3 g/kg/day, p.o.) significantly increased the number of punished licks. The drug did not change the total entries into the open arms of the EPM or interfere with water consumption or nociceptive threshold, discarding potential confounding factors in the two tests. In the open field, locomotion was not reduced, discarding the possible sedative effect of CZZX. In the binding assay, the binding of [3H] Ro 15-1788 (flumazenil) to the benzodiazepine binding site in washed crude synaptosomal membranes from rat cerebral cortex was affected by CZZX. These data indicate an anxiolytic-like profile of action for CZZX without sedative side effects, and this activity may be mediated by benzodiazepine binding site modulation at γ-aminobutyric acid-A receptors.

We briefly describe the polarization state of the spike groups and fast fine structures superimposed on the 47GB microwave burst of 1991 May 16. At two frequencies 100 MHz apart, namely, 2645 and 2545 MHz, the polarization showed rapid reversals on short time scales of the order of 50 ∼ 100ms. We think this is probably related to fluctuations on both short time and small spatial scales, in the plasma density and magnetic field of the emission region. These oscillations cause the plasma and cyclotron frequency ratio to vary around the value of 2 , thus making the electron cyclotron maser instability growth rate to be dominated alternately by the X mode and the O mode and the polarization state to switch from one sign to the other.

Near-field probes are attractive because they provide sub-wavelength resolution. Typical detection using near-field probes involves defects or irregularities present on the surface of the material. Sub-surface defects or irregularities are much more difficult to detect due to the fast decay of evanescent waves emanating from the probe. This work present novel probes employing metamaterial particles which work to substantially enhance the evanescent field energy in the vicinity of the probe, thereby increasing the sensitivity of the probe and enabling subsurface detection. Two experiments are presented where conducting metallic objects immersed in saline water and sand are detected. The results and findings presented here are expected to have strong implications in a wide range of subsurface detection challenges including detection of biological anomalies in human tissue, land mine detection, through-wall detection amongst other applications.

Evanescent fields have been used to resolve material features in subwavelength resolution. The technique provides a wide range of applications in noninvasive surface and subsurface detection, such as detection of holes covered with solder in circuit boards [1], imaging of micro-structures in semiconductors, localization of buried objects [2], detection of defects covered by paint [3] and imaging of abnormalities in biological samples [4, 5].

Eye diagram is a very successful way of showing the quality and parametric information about the serial data signal. But traditional test instrument has some drawbacks for implementation of eye diagram construction and analysis. The third generation oscilloscope--Digital Phosphor Oscilloscope (DPO) can overcome the shortcomings of traditional test instrument. In this thesis, the design and implementation of eye diagram construction and analysis in the DPO developed by us is discussed. And the testing results of our design are given at the end of the thesis. The design has been used in the prototype of DPO and works well.

Recently, we showed that single negative media can significantly enhance the sensitivity of near-field probes. Inspired by this recent finding, we propose a new near-field probe that uses only one single split ring resonator instead of a periodic arrangement of these resonators. We show through numerical simulations that the new proposed probe lead to significant enhancement in the sensitivity as measured by the change in phase in the reflection coefficient. Experimental tests were conducted to detect the presence of a small aluminum block located behind a layer of ground chicken (lossy medium). Preliminary results show that with our new probe, a phase change in the reflection coefficient of 14 degrees can be achieved for 2 mm thick ground chicken and 6.24 mm aluminum cube at an operation frequency of 1.16 GHz.

A high speed (47,000 A-scan/s), high resoluiton FD-OCT system, operating in the 1060nm wavelength range was used to acquire in-vivo 3D image of healthy and pathological rat retinas. The images were acquired with ~4.3µm axial and ~5µm lateral resolution in the rat eye and 102dB sensitivity at 1.3mW optical power of the imaging beam. The images of the healthy rat retinas show increased penetration into the choroid, clear visualization of all intra-retinal layers and the choroidal blood network, as well as part of the underlying sclera. The high imaging resolution of the OCT system is also sufficient for resolving tiny capillaries imbedded in the inner - and outer plexiform layers of the retina. The high data acquisition rate of the FD-OCT system combined with the high axial resolution is also suitable for probing light induced physiological processes in the retina simultaneously with the morphological imaging.

Based on identification and prediction ability of neural networks for nonlinear systems, an improved BP network was adopted to predict the seismic responses of the bridge structures. First, a multi-player BP networks based on Levenberg-Marquardt algorithm was formed. Then, the improved neural network was trained by the imitated seismic responses of the first 4 seconds which were obtained from artificial earthquake waves by finite element method. Thirdly, the seismic responses of 1st to 8th seconds for the same bridge structure were predicted use the neural network which has been trained, and the predict responses were compared with the calculation data. The error curves between the prediction and the calculation results show that the BP network combined with is Levenberg-Marquardt algorithm has very good convergence rate, and the artificial neural network can predict the dynamic response of bridge structures well enough.

In recent years, computational intelligence methods are widely used to solve problems in engineering structural field by more and more researchers. In this paper, a method based on combining artificial neural networks (ANN) and genetic algorithm (GA) in identifying the prestress loss of the long span bridge was proposed. A model with some parameters was constructed to access the prestress loss. Based on the differences of the bridge deck elevation of the long span bridge in service between damaged structure and intact structure, the model parameters were identified by the genetic algorithm, and the prestress loss values of different time were obtained. At the same time, the differences of responses were constructed within BP neural networks. In fact, the trained BP neural networks were used as a subroutine in the process of identification using GA. It is demonstrated that the method combining BP neural networks with GA of identification for prestress loss is efficient and the identified results fit well with the actual value.

In the present paper we report the results of a correlation analysis for 57 microwave impulsive bursts observed at six frequencies in which we have obtained a regression line between the peak frequency and the corresponding rise time of microwave impulsive bursts: {ie361-01} (with a correlation coefficient of - 0.43). This can be explained in the frame of a thermal model. The magnetic field decrease with height has to be much slower than in a dipole field in order to explain the weak dependence of f p on t r . This decrease of magnetic field with height in burst sources is based on the relationship between f p and t r found by assuming a thermal flare model with a collisionless conduction front.

Laser direct imprint technology of silicon nanostructure has only 56cm/s change in surface recombination velocity, with excimer laser peak intensity of 1~2J/cm2, imprint pressure of 10~100g/cm2, and high vacuum at 10-6torr.

Excimer laser assisted imprint technology is implemented to create nanostructure on silicon substrate. Patterns with 25-nm feature sizes are successfully fabricated on quartz mold using electron beam lithography for nano-imprint. Transfer of nano-meter features to silicon is experimented. Parameters for optimal imprint are investigated. With a peak light intensity of 1/spl sim/2 J/cm/sup 2/, imprint pressure of 10/spl sim/100 g/cm/sup 2/, and high vacuum at 10/sup 6/ torr, nano-imprint has very good results. The carrier lifetimes before and after laser-assisted imprint are studied and discovered to vary from 1818 /spl mu/s to 640 /spl mu/s. This small damage of Si substrate is attributed to recrystallization of silicon.

Fourteen patients (20 renal units) with cystine stones were treated by extracorporeal shock wave lithotripsy (SWL, Dornier HM3). The cases consisted of four ureteral stones, two renal + ureteral st...

A high-speed (47,000 A-scans/s), ultrahigh axial resolution Fourier domain optical coherence tomography (OCT) system for retinal imaging at approximately 1060 nm, based on a 1024 pixel linear array, 47 kHz readout rate InGaAs camera is presented. When interfaced with a custom superluminescent diode (lambda(c) = 1020 nm, Deltalambda = 108 nm, Pout = 9 mW), the system provides 3.3 microm axial OCT resolution at the surface of biological tissue, approximately 4.5 microm in vivo in rat retina, approximately 5.7 microm in vivo in human retina, and 110 dB sensitivity for 870 microW incident power and 21 mus integration time. Retinal tomograms acquired in vivo from a human volunteer and a rat animal model show clear visualization of all intraretinal layer and increased penetration into the choroid.

In this paper we investigate the contribution of gyro-resonance to the SVC of the solar radio emission. We calculated the spatial distributions of the gyroresonance radiation over uinpolar and bipolar active regions at a number of wavelengths and we found some interesting features. We also calculated the spectra of flux and polarization and our results were in general agreement with observations. This suggests that the gyro-resonance radiation is indeed an Important mechanism for the solar SVC.

Acoustic Scene Classification (ASC) aims to classify the environment in which the audio signals are recorded. Recently, Convolutional Neural Networks (CNNs) have been successfully applied to ASC. However, the data distributions of the audio signals recorded with multiple devices are different. There has been little research on the training of robust neural networks on acoustic scene datasets recorded with multiple devices, and on explaining the operation of the internal layers of the neural networks. In this article, we focus on training and explaining device-robust CNNs on multi-device acoustic scene data. We propose conditional atrous CNNs with attention for multi-device ASC. Our proposed system contains an ASC branch and a device classification branch, both modelled by CNNs. We visualise and analyse the intermediate layers of the atrous CNNs. A time-frequency attention mechanism is employed to analyse the contribution of each time-frequency bin of the feature maps in the CNNs. On the Detection and Classification of Acoustic Scenes and Events (DCASE) 2018 ASC dataset, recorded with three devices, our proposed model performs significantly better than CNNs trained on single-device data., Comment: IEEE Transactions on Multimedia