Search

With the dramatic progress of wearable devices, continuous collection of single lead ECG wave is able to be implemented in a comfortable fashion. Data mining on single lead ECG wave is therefore attracting increasing attention, where atrial fibrillation (AF) detection is a hot topic. In this paper, we propose a dual-channel neural network for AF detection from a single lead ECG wave. Two primary phases are included, the data preprocessing part followed by a dual-channel neural network. A two-stage denoising procedure is developed for data preprocessing, so as to tackle the high noise and disturbance which generally resides in the ECG wave collected by wearable devices. Then the time-frequency spectrum and Poincare plot of the denoised ECG signal are imported into the developed dual-channel neural network for feature extraction and AF detection. On the 2017 PhysioNet/CinC Challenge database, the F1 values were 0.83, 0.90, and 0.75 for AF rhythm and normal rhythm, and other rhythm, respectively. The results well validate the effectiveness of the proposed method for AF detection from a single lead ECG wave, and also indicate its performance advantages over some state-of-the-art counterparts.

Online users are typically active on multiple social media networks (SMNs), which constitute a multiplex social network. With improvements in cybersecurity awareness, users increasingly choose different usernames and provide different profiles on different SMNs. Thus, it is becoming increasingly challenging to determine whether given accounts on different SMNs belong to the same user; this can be expressed as an interlayer link prediction problem in a multiplex network. To address the challenge of predicting interlayer links, feature or structure information is leveraged. Existing methods that use network embedding techniques to address this problem focus on learning a mapping function to unify all nodes into a common latent representation space for prediction; positional relationships between unmatched nodes and their common matched neighbors (CMNs) are not utilized. Furthermore, the layers are often modeled as unweighted graphs, ignoring the strengths of the relationships between nodes. To address these limitations, we propose a framework based on multiple types of consistency between embedding vectors (MulCEVs). In MulCEV, the traditional embedding-based method is applied to obtain the degree of consistency between the vectors representing the unmatched nodes, and a proposed distance consistency index based on the positions of nodes in each latent space provides additional clues for prediction. By associating these two types of consistency, the effective information in the latent spaces is fully utilized. In addition, MulCEV models the layers as weighted graphs to obtain representation. In this way, the higher the strength of the relationship between nodes, the more similar their embedding vectors in the latent representation space will be. The results of our experiments on several real-world and synthetic datasets demonstrate that the proposed MulCEV framework markedly outperforms current embedding-based methods, especially when the number of training iterations is small.

High light-harvesting efficiency and low interfacial charge transfer loss are essential for the fabrication of high-efficiency quantum dot-based solar cells (QDSCs). Increasing the thickness of mesoporous TiO2 films can improve the loading of pre-synthesized QDs on the film and enhance the absorbance of photoanode, but commonly accompanied by the increase in the unfavorable charge recombination due to prolonged electron transmission paths. Herein, we systematically studied the influence of the balance between QD loading and TiO2 film thickness on the performance of QDSCs. It is found that the relative thin photoanode prepared by the cationic surfactant-assisted multiple deposition procedure has achieved a high QD loading which is comparable to that of the thick photoanode commonly used. Under AM 1.5G illumination, Zn–Cu–In–Se and Zn–Cu–In–S based QDSCs with optimized 11.8 μm photoanodes show the PCE of 10.03% and 8.53%, respectively, which are comparable to the corresponding highest PCE of Zn–Cu–In–Se and Zn–Cu–In–S QDSCs (9.74% and 8.75%) with over 25.0 μm photoanodes. Similarly, an impressive PCE of 6.14% was obtained for the CdSe based QDSCs with a 4.1 μm photoanode, which is slightly lower than the best PCE (7.05%) of reference CdSe QDSCs with 18.1 μm photoanode.

3D object detection from mobile phones in device to device (D2D) system provides a new smart payment tool for the next-generation of fintech, which is more flexible and efficient than the traditional barcode. In this paper, we propose a monoc-ular 3D object detection method based on depth guided local convolution. The method combines the information of RGB image mode and depth mode by using a convolution kernel through depth image, and works on a single RGB image locally. According to the multiscale input information, the convolution kernel is adaptively adjusted to capture the target objects of different scales, so as to improve the performance of 3D object detection. In addition, we use soft-NMS algorithm instead of traditional NMS to select the best prediction box. In order to further improve the accuracy of 3D object detection, the depth estimation network and 3D object detection network are jointly trained in this method to make the two networks constrain each other and achieve the best performance.

Edge computing provides computing capability at close-user proximity to reduce service latency for end users. To improve the efficiency of edge computing infrastructures, geographically-distributed edge datacenters can co-work with each other and with cloud datacenters, forming a new paradigm referred to as cooperative edge-cloud computing. In this context, applications typically run on a virtual machine (VM) that can be replicated at multiple sites, and thus user traffic can be served at all the sites where corresponding VMs reside. For the performance of many applications, latency is a critical parameter. In this work, taking applications latencies as the primary constraint, we model the problem of VM placement and workload assignment as a mixed integer linear program and develop heuristic algorithms accordingly. The goal is to minimize the consumption of information technology (IT) infrastructures for placing VMs in cooperative edge-cloud computing, while meeting the heterogeneous latency demands of different applications. Some preliminary results indicate that edge datacenters resource efficiency can be optimized by proper cross-site VM placement and workload re-direction.

Machine learning (ML) models are increasingly trained with distributed workers possessing heterogeneous resources. In such scenarios, model training efficiency may be negatively affected by \emph{stragglers}---workers that run much slower than others. Efficient model training requires eliminating such stragglers, yet for modern ML workloads, existing load balancing strategies are inefficient and even infeasible. In this paper, we propose a novel strategy, called \emph{semi-dynamic load balancing}, to eliminate stragglers of distributed ML workloads. The key insight is that ML workers shall be load-balanced at \emph{iteration boundaries}, being non-intrusive to intra-iteration execution. Based on it we further develop LB-BSP, an integrated worker coordination mechanism that adapts workers' load to their instantaneous processing capabilities---by right-sizing the sample batches at the synchronization barriers. We have designed distinct load tuning algorithms for ML in CPU clusters, in GPU clusters as well as in federated learning setups, based on their respective characteristics. LB-BSP has been implemented as a Python module for ML frameworks like TensorFlow and PyTorch. Our EC2 deployment confirms that LB-BSP is practical, effective and light-weight, and is able to accelerating distributed training by up to $54\%$ .

This study aims to find out the possible optimal therapy and assess the prognosis properly for patient with spontaneous rupture of hepatocellular carcinoma (HCC).Propensity score matching (PSM) analysis was used to study the data from 325 patients with ruptured HCC (RHCC) and 2,291 patients with non-RHCC.The incidence and hospital mortality of RHCC were 5.1% and 0.8% respectively, with a median overall survival (OS) time of 17 months. There was no difference between ruptured and non-RHCC patients undergoing conservation treatment in terms of OS. Trans-arterial embolization (TAE) was carried out in 69 (21.2%) cases with RHCC, with a median OS of 7 months, which was no difference from that of non-RHCC (pre- and post-PSM). One hundred and sixty-nine (52.0%) RHCC cases underwent one-stage hepatectomy, with a median OS and disease-free survival (DFS) of 30 and 6 months respectively, which were shorter than that of non-RHCC (post-PSM). TAE plus two-stage hepatectomy was performed in 30 RHCC cases, with a median OS and DFS of 28 and 10 months respectively; these outcomes were better than that from RHCC patients undergoing TAE alone or one-stage hepatectomy (post-PSM), which were no difference from that of non-RHCC patients undergoing hepatectomy. The risk of death for RHCC patient undergoing one-stage hepatectomy is 1.545 times higher than that of one undergoing TAE + two-stage hepatectomy.TAE plus two-stage hepatectomy might be the optimal treatment for RHCC patient. Under the premise of the same pathological properties, there is no difference in prognosis between ruptured and non-RHCC patients if the therapy is appropriate.

Background Surgical treatment of severe thoracolumbar kyphosis (TLK) secondary to late osteoporotic vertebral compression fracture (OVCF) presents several challenges to spine surgeons. Proper selection of distal fusion level is an important issue in surgical planning to decrease the likelihood of distal mechanical complications. This study was designed to compare the clinical and radiographic outcomes in elderly patients suffering from severe TLK related with late OVCF between different distal fusion strategies and to recommend a superior distal fusion level for these patients. Methods A total of 57 consecutive subjects with a minimum follow-up of two years were retrospectively reviewed. TLK was defined as the hyperkyphosis with an apex below T10. Severe TLK was defined as the Cobb angle of kyphosis ≥60°. Patients fused to sagittal stable vertebra (SSV) were assigned to Group SSV, while those fused to the vertebra above and below SSV were assigned to Groups SSV- and SSV+, respectively. Bone cement was used to enhance pedicle screw fixation. Clinical and radiographic results were collected and compared between different groups. Results Deformity corrections and living quality improvements at the latest follow-up were superior in Group SSV than Group SSV- with shorter fusion levels, while to the equal extent with Group SSV+ with longer fusion levels. 7 cases of distal complications were observed in Group SSV-. Negatively balanced lowest instrumented vertebra was revealed to be the independent factor predicting distal complications. Patients’ satisfaction of their surgical management was greater in Group SSV than Group SSV- (83.2 ± 4.4% vs. 70.5 ± 10.9%, P = 0.024), while comparable with Groups SSV + SSV+ (84.8 ± 5.7%). Conclusions Fusion to SSV with cement-augmented pedicle screws could limit the development of distal junctional mechanical complications after surgical treatment for severe TLK secondary to late OVCF in elderly patients, while achieves satisfactory deformity correction with the preservation of necessary lumbar motility.

Currently, depression has become a common mental disorder, especially among postgraduates. It is reported that postgraduates have a higher risk of depression than the general public, and they are more sensitive to contact with others. Thus, a non-contact and effective method for detecting people at risk of depression becomes an urgent demand. In order to make the recognition of depression more reliable and convenient, we propose a multi-modal gait analysis-based depression detection method that combines skeleton modality and silhouette modality. Firstly, we propose a skeleton feature set to describe depression and train a Long Short-Term Memory (LSTM) model to conduct sequence strategy. Secondly, we generate Gait Energy Image (GEI) as silhouette features from RGB videos, and design two Convolutional Neural Network (CNN) models with a new loss function to extract silhouette features from front and side perspectives. Then, we construct a multi-modal fusion model consisting of fusing silhouettes from the front and side views at the feature level and the classification results of different modalities at the decision level. The proposed multi-modal model achieved accuracy at 85.45% in the dataset consisting of 200 postgraduate students (including 86 depressive ones), 5.17% higher than the best single-mode model. The multi-modal method also shows improved generalization by reducing the gender differences. Furthermore, we design a vivid 3D visualization of the gait skeletons, and our results imply that gait is a potent biometric for depression detection.

Citation recommendation is important in the environment of scholarly big data, where finding relevant papers has become more difficult because of information overload. Applying traditional collaborative filtering (CF) to citation recommendation is rather challenging due to the cold start problem and the lack of paper ratings. To address these two challenges, in this paper, we propose a Collaborative filtering with Network representation learning framework for Citation Recommendation dubbed as CNCRec, which is a hybrid user-based CF considering both paper content and network topology. It aims at recommending citations in heterogeneous academic information networks. CNCRec creates the paper rating matrix based on attributed citation network representation learning, where the attributes are topics extracted from the paper text information. Meanwhile, the learned representations of attributed collaboration network is utilized to improve the selection of nearest neighbors. By harnessing the power of network representation learning, CNCRec is able to make full use of the whole citation network topology compared with previous context-aware network-based models. Extensive experiments on both DBLP and APS datasets show that the proposed method outperforms state-of-the-art methods in terms of precision, recall, and Mean Reciprocal Rank. Moreover, CNCRec can better solve the data sparsity problem compared with other CF-based baselines.

Eigendecomposition of symmetric matrices is at the heart of many computer vision algorithms. However, the derivatives of the eigenvectors tend to be numerically unstable, whether using the SVD to compute them analytically or using the Power Iteration (PI) method to approximate them. This instability arises in the presence of eigenvalues that are close to each other. This makes integrating eigendecomposition into deep networks difficult and often results in poor convergence, particularly when dealing with large matrices. While this can be mitigated by partitioning the data into small arbitrary groups, doing so has no theoretical basis and makes it impossible to exploit the full power of eigendecomposition. In previous work, we mitigated this using SVD during the forward pass and PI to compute the gradients during the backward pass. However, the iterative deflation procedure required to compute multiple eigenvectors using PI tends to accumulate errors and yield inaccurate gradients. Here, we show that the Taylor expansion of the SVD gradient is theoretically equivalent to the gradient obtained using PI without relying in practice on an iterative process and thus yields more accurate gradients. We demonstrate the benefits of this increased accuracy for image classification and style transfer., Comment: IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) PREPRINT 2021

BACKGROUND/PURPOSE Clarithromycin-based standard triple therapy is still commonly adopted by 81.4% of physicians in real-world practice but yields low eradication rates. Therefore, we conducted this study to compare the efficacy of gastric juice-guided therapy for first-line eradication with the standard triple therapy, in order to provide an alternative to real-world practice. METHODS A total of 182 treatment-naive Hp-infected patients were included and randomly allocated to either susceptibility-guided therapy (SGT) with gastric juice PCR or Clarithromycin-based standard triple therapy (STT) for 7 days. RESULTS The intention-to-treat eradication rates were 89% (81/91) in SGT and 75.8% in STT (p < 0.031). The per-protocol eradication rates were 91.0% (81/89) in SGT and 79.3% (69/87) in STT (p < 0.034). Among the subgroups of different antibiotic resistance, patients with SGT demonstrated superior eradication rates (91.7% vs 45.5%, p < 0.027) in the subgroup of both clarithromycin resistance and levofloxacin resistance. CONCLUSION This prospective randomized controlled trial demonstrated the reliable efficacy of susceptibility-guided therapy via gastric juice PCR for the first-line Hp eradication. In Asia-Pacific area, where standard triple therapy is still adopted by the majority of the physicians, it is a recommended alternative to overcome the increasing antibiotic resistance.

The remarkable advances of Machine Learning (ML) have spurred an increasing demand for ML-as-a-Service on publiccloud: developers train and publish ML models as online services to provide low-latency inference for dynamic queries. The primarychallenge of ML model serving is to meet the response-time Service-Level Objectives (SLOs) of inference workloads while minimizingserving cost. In this paper, we proposes MArk (Model Ark), a general-purpose inference serving system, to tackle the dual challenge ofSLO compliance and cost effectiveness. MArk employs three design choices tailored to inference workload. First, MArk dynamicallybatches requests and opportunistically serves them using expensive hardware accelerators (e.g., GPU) for improved performance-costratio. Second, instead of relying on feedback control scaling or over-provisioning to serve dynamic workload, which can be too slow ortoo expensive, MArk employs predictive autoscaling to hide the provisioning latency at low cost. Third, given the stateless nature ofinference serving, MArk exploits the flexible, yet costly serverless instances to cover occasional load spikes that are hard to predict. Weevaluated the performance of MArk using several state-of-the-art ML models trained in TensorFlow, MXNet, and Keras. Compared withthe premier industrial ML serving platform SageMaker, MArk reduces the serving cost up to 7.8while achieving even better latencyperformance.

A solid-liquid triboelectric nanogenerator (TENG) has attracted increasing research interest in relation to the development of regeneration energy based on water resources. The output of solid-liquid TENG remains unsolved, however, because of the low voltage output that impedes wide applications. To this end, in this work we developed a miniaturized microfluidic channel-based TENG device for highly efficient conversion of energy from the transport of a water droplet to an electrical output. We investigated an optimized design in a triboelectric material, the droplet transport and the electrostatic induction layer to provide a high voltage output and stable energy harvesting. The optimized device demonstrated maximum voltage amplitude 102 mV with an ultralow liquid consumption, 0.36 [Formula: see text], resulting in sample-energy conversion 283.33 mV/ [Formula: see text]. This novel device is expected potentially to address the limitations imposed by sample consumption in energy harvesting in the future.

Blau syndrome (BS) is a monogenic autoinflammatory disease caused by mutations in nucleotidebinding oligomerization domain containing 2 (NOD2). BS is characterized by the clinical triad of granulomatous dermatitis, arthritis and recurrent uveitis. Due to the low incidence of BS and the lack of treatment studies with large samples, a specific treatment scheme has not been established. We report the case of a patient with BS that was uncontrollable with various immunosuppressive therapies but had a good response to thalidomide. She had the typical triad of rash, arthritis and uveitis. Gene sequencing indicated a NOD2 heterozygous missense variant (c.1759C > T, p.R587C), which has been reported as a pathogenic mutation. The BS diagnosis was confirmed. After treatment with methotrexate, an anti-tumour necrosis factor (TNF)-α inhibitor and corticosteroids, the patient's clinical symptoms and inflammatory indicators remained uncontrolled, and she experienced multiple side effects, such as hypertension and growth retardation attributed to prolonged corticosteroid use. After treatment with thalidomide, her condition was controlled without recurrence or side effects, and corticosteroids were stopped as soon as possible. This report suggests that thalidomide may be effective for BS treatment, but more research is needed to evaluate its long-term efficacy and side effects.

The idea of ‘citizen sensing’ and ‘human as sensors’ is crucial for social Internet of Things, an integral part of cyber-physical-social systems (CPSS). Social media data, which can be easily collected from the social world, has become a valuable resource for research in many different disciplines, e.g. crisis/disaster assessment, social event detection, or the recent COVID-19 analysis. Useful information, or knowledge derived from social data, could better serve the public if it could be processed and analyzed in more efficient and reliable ways. Advances in deep neural networks have significantly improved the performance of many social media analysis tasks. However, deep learning models typically require a large amount of labeled data for model training, while most CPSS data is not labeled, making it impractical to build effective learning models using traditional approaches. In addition, the current state-of-the-art, pre-trained Natural Language Processing (NLP) models do not make use of existing knowledge graphs, thus often leading to unsatisfactory performance in real-world applications. To address the issues, we propose a new zero-shot learning method which makes effective use of existing knowledge graphs for the classification of very large amounts of social text data. Experiments were performed on a large, real-world tweet dataset related to COVID-19, the evaluation results show that the proposed method significantly outperforms six baseline models implemented with state-of-the-art deep learning models for NLP.

A long-standing vision of indoor localization is to eliminate infrastructure and deployment costs. Recent innovations make it possible to enable device-to-device (D2D) localization while requiring multiple antennas for the systems. We ask the following question: can we localize the more generally-used single-antenna devices (e.g., IoT) using another single-antenna device (e.g., smartphone or smartwatch) in a smart environment where low-cost backscatter tags are widely deployed on walls or smart objects. In this paper, we present TagLoc, a light-weight system that enables device-to-device localization without relying on large antenna arrays. Our observation is that the reflected signals from the ambient smart environment can be exploited to eliminate the requirement of bulky antenna arrays that are unachievable for the simple-designed IoT devices. Specifically, TagLoc creates multiple direction signatures using backscatter arrays in smart environments. Then, the receiver can accurately estimate the direction signatures from the transmitter to the arrays and then localize the target by cooperating all tag arrays. We prototype TagLoc using two single-antenna Intel NUCs with off-the-shelf Intel 5300 WiFi cards and customized backscatter tags. The results show TagLoc can achieve robust performance in a real indoor environment with a median localization error of 0.82m.

Face manipulation techniques improve fast with the development of powerful image generation models. Two particular face manipulation methods, namely face swap and expression reenactment attract much attention for their flexibility and ease to generate high quality synthesis results. Recently, these two subjects are actively studied. However, most existing methods treat the two tasks separately, ignoring their underlying similarity. In this paper, we propose to tackle the two problems within a unified framework that achieves high quality synthesis results. The enabling component for our unified framework is the clean disentanglement of 3D pose, shape, and expression factors and then recombining them for different tasks accordingly. We then use the same set of 2D representations for face swap and expression reenactment tasks that are input to a common image translation model to directly generate the final synthetic images. Once trained, the proposed model can accomplish both face swap and expression reenactment tasks for previously unseen subjects. Comprehensive experiments and comparisons show that the proposed method achieves high fidelity results in multiple aspects, and it is especially good at faithfully preserving source facial shape in the face swap task, and accurately transferring facial movements in the expression reenactment task.

Miniature radar has demonstrated its great potential in smart homes to understand the wellness of the residents and provide ubiquitous interactions. While it has many promising applications, it also results in congested RF (radio frequency) environments as there is an unprecedented amount of traffic in a smart home. To ease the strain on the limited spectrum, we ask the question that, can we reuse the sensing signals for data communication With such a capability, we can improve the spectrum utilization by sharing the spectrum between sensing and communication systems. However, radar signals are customized for the sensing purpose and are incompatible with legacy communication standards. To address this challenge, we have an observation that, non-linearity effect in RF circuits can convert wideband radar signals into a LoRa signal. Based on this observation, in this paper, we present LoRadar, which enables an FMCW (Frequency-Modulated Continuous Wave) radar to carry LoRa signals in sensing waves. We present both the downlink and uplink design, enabling a LoRadar device to communicate with LoRa nodes in a bi-directional way. We implement LoRadar and evaluation results show that LoRadar can achieve home-level coverage with 3:4kbps data rate while it preserves the sensing resolution of the radar.

The stator-PM machine concept has been practiced on the recently proposed stator-PM variable reluctance (VR) resolver. When inheriting the simple and sturdy rotor structure of the conventional VR resolver, the novel stator-PM VR resolver provides a chance to avoid the high-frequency (HF) excitation by the unique PM back-EMF decoding solution. This paper extends this stator-PM VR resolver category with the axial magnetic field technology. The axial slotless structure makes the VR resolver compact and much smaller than the radial field prototypes. The thin and light leaf-style rotor is suitable for the super-high-speed applications, as the additional inertia due to a sensor is not welcome in those scenarios. Additionally, the printed circuit board (PCB) replaces the winding coils in the regular VR resolvers, significantly simplifying the manufacturing. The design with even and odd poles cannot share the same principle, so they are discussed separately. The prototypes with 4-X and 5-X configurations are tested in the experiment setup respectively to validate the analysis of the proposed structure. Moreover, the 4-X resolver is integrated with a permanent magnet synchronous machine (PMSM), and a motor driver with the space vector modulation (SVM) on the basis of the specific resolver decoder is built.

Lipid nanoparticle (LNP) is commonly used to deliver mRNA vaccines. Currently, LNP optimization primarily relies on screening ionizable lipids by traditional experiments which consumes intensive cost and time. Current study attempts to apply computational methods to accelerate the LNP development for mRNA vaccines. Firstly, 325 data samples of mRNA vaccine LNP formulations with IgG titer were collected. The machine learning algorithm, lightGBM, was used to build a prediction model with good performance (R2 > 0.87). More importantly, the critical substructures of ionizable lipids in LNPs were identified by the algorithm, which well agreed with published results. The animal experimental results showed that LNP using DLin-MC3-DMA (MC3) as ionizable lipid with an N/P ratio at 6:1 induced higher efficiency in mice than LNP with SM-102, which was consistent with the model prediction. Molecular dynamic modeling further investigated the molecular mechanism of LNPs used in the experiment. The result showed that the lipid molecules aggregated to form LNPs, and mRNA molecules twined around the LNPs. In summary, the machine learning predictive model for LNP-based mRNA vaccines was first developed, validated by experiments, and further integrated with molecular modeling. The prediction model can be used for virtual screening of LNP formulations in the future.

Background This study is designed to evaluate the predictive value of derived monocyte to lymphocyte ratio (dMLR), neutrophil to lymphocyte ratio (NLR) and eosinophil to lymphocyte ratio (ELR) for lymph node metastasis (LNM) in gastric signet ring cell carcinoma (SRCC) patients. Methods We carried out a retrospective analysis for 203 eligible patients who underwent radical gastrectomy between January 2011 and September 2020. Results The diagnostic sensitivity and specificity for dMLR were 60.3% and 72.2%, respectively; while for NLR and ELR were 64.1% and 66.6%, and 91.6% and 23.6%, respectively. The dMLR and NLR were identified as independent predictors of LNM, but ELR was not. A nomogram was developed incorporating dMLR and clinical characteristics with AUC of 0.868, and the outcome of DCA supported good clinical benefit. Conclusions dMLR is a promising novel preoperative biomarker to predict LNM of gastric SRCC.

Medium access control (MAC) protocol identification in the context of cognitive radio is a challenging issue, which is usually solved by the support vector machine (SVM) method. To avoid manual feature extraction in SVM and realize accurate identification, a graphical scheme is adopted in this work. Firstly, the identification problem is modeled to analyze the distinctions of four MAC protocols by exploiting time and power features of the received signal. Then, we propose a convolutional neural network (CNN)-based MAC protocol identification (C-MPI) method, which combines CNN with the time-frequency image that contains time, frequency and energy three dimensional information. Moreover, in order to further improve the identification performance, a modified CNN with a squeeze-and-excitation mechanism-based MAC protocol identification (CSE-MPI) approach is developed, which plugs the SE block, a channel-wise attention mechanism, into classical CNN. Numerical simulations are provided to demonstrate the effectiveness of the proposed approaches.

Understanding and predicting human motion behavior with social interactions have become an increasingly crucial problem for a vast number of applications, ranging from visual navigation of autonomous vehicles to activity prediction of intelligent video surveillance. Accurately forecasting crowd motion behavior is challenging due to the multimodal nature of trajectories and complex social interactions between humans. Recent algorithms model and predict the trajectory with a single resolution, making them difficult to exploit the long-range information and the short-range information of the motion behavior simultaneously. In this paper, we propose a temporal pyramid network for pedestrian trajectory prediction through a squeeze modulation and a dilation modulation. The hierarchical design of our framework allows to model the trajectory with multi-resolution, then can better capture the motion behavior at various tempos. By progressively combining the global context with the local one, we finally construct a coarse-to-fine hierarchical pedestrian trajectory prediction framework with multi-supervision. Further, we introduce a unified spatial-temporal attention mechanism to adaptively select important information of persons around in both spatial and temporal domains. We show that our attention strategy is intuitive and effective to encode the influence of social interactions. Experimental results on several benchmarks demonstrate the superiority of our proposed scheme.

Temporal action detection is a challenging task in video understanding, due to the complexity of the background and rich action content impacting high-quality temporal proposals generation in untrimmed videos. Capsule networks can avoid some limitations of the invariance caused by pooling and inability from convolutional neural networks, which can better understand the temporal relations for temporal action detection. However, because of the extremely computationally expensive procedure, capsule network is difficult to be applied to the task of temporal action detection. To address this issue, this paper proposes a novel U-shaped capsule network framework with a k-Nearest Neighbor (k-NN) mechanism of 3D convolutional dynamic routing, which we named U-BlockConvCaps. Furthermore, we build a Capsules Boundary Network (CapsBoundNet) based on U-BlockConvCaps for dense temporal action proposal generation. Specifically, the first module is one 1D convolutional layer for fusing the two-stream with RGB and optical flow video features. The sampling module further processes the fused features to generate the 2D start-end action proposal feature maps. Then, the multi-scale U-Block convolutional capsule module with 3D convolutional dynamic routing is used to process the proposal feature map. Finally, the feature maps generated from the CapsBoundNet are used to predict starting, ending, action classification, and action regression score maps, which help to capture the boundary and intersection over union features. Our work innovatively improves the dynamic routing algorithm of capsule networks and extends the use of capsule networks to the temporal action detection task for the first time in the literature. The experimental results on benchmarks THUMOS14 show that the performance of CapsBoundNet is obviously beyond the state-of-the-art methods, e.g., the mAP@tIoU=0.3, 0.4, 0.5 on THUMOS14 are improved from 63.6% to 70.0%, 57.8% to 63.1%, 51.3% to 52.9%, respectively. We also got competitive results on the action detection dataset of ActivityNet1.3.

Acute myocardial infarction (AMI) remains a dominant origin of morbidity, mortality and disability worldwide. Increases in reactive oxygen species (ROS) are key contributor to excessive cardiac injury after AMI. Here we developed an immobilized enzyme with Superoxide Dismutase (SOD) activity cross-link with Zr-based metal-organic framework (ZrMOF) (SOD-ZrMOF) for mitigate ROS-caused injury. In vitro and in vivo evidence indicates that SOD-ZrMOF exhibits excellent biocompatibility. By efficiently scavenging ROS and suppressing oxidative stress, SOD-ZrMOF can protect the function of mitochondria, reduce cell death and alleviate inflammation. More excitingly, long-term study using an animal model of AMI demonstrated that SOD-ZrMOF can reduce the infarct area, protect cardiac function, promote angiogenesis and inhibit pathological myocardial remodeling. Therefore, SOD-ZrMOF holds great potential as an efficacious and safe nanomaterial treatment for AMI., Graphical abstract Image 1, Highlights • We synthesized a superoxide dismutase immobilized nanocomposite with enhanced stability. • We tested the biocompatibility of SOD-ZrMOF and found it could promote the cell viability and proliferation of H9c2 cells. • SOD-ZrMOF shows excellent performance in protecting mitochondrial function of CMs by eliminating excessive ROS under hypoxia. • SOD-ZrMOF exhibited therapeutic effects on MI heart through inhibiting oxidative stress injury by NF-κB/HIF-1α pathway.

In modern data centers, enterprise cloud instances run not only foreground applications like web and databases, but also different background services (e.g., backup, virus/compliance scan, batch) to manage the cloud instances securely and improve the overall resource utilization. These background services often incur resource storms that suddenly consume a lot of shared resources on cloud instances. The resource storms significantly degrade the performance of foreground applications by interfering in the preemption of the shared resources, resulting in frequent SLA violations. However, stock OS schedulers are not designed to handle these situations, and prior works are insufficient to address such resource storms under highly dynamic cloud workloads. This work presents Orchestra, a cloud-specific framework for controlling multiple applications in the user space, aiming at meeting corresponding SLAs. Orchestra takes an online approach with lightweight monitoring and performance models for both applications on the fly. It optimizes the resource allocations to meet corresponding SLAs. We evaluate the performance of Orchestra on a production cloud with a diverse range of SLAs. Orchestra guarantees the foreground application's performance SLAs at all times. At the same time, Orchestra maintains the background's performance by minimizing its performance penalty with proper allocation of the shared resources.

Primary central nervous system anaplastic lymphoma kinase (ALK)-negative anaplastic large cell lymphoma (ALCL) is an extremely rare type of primary central nervous system lymphoma (PCNSL). There are only nine cases reported in the literature to date, most of which have an overall survival time of no more than 8 months. Herein, we report such a rare case who has a good outcome with the longest survival time and perform a review of the literature. A 19-year-old male patient was admitted to the hospital complaining of dizziness. CT and MRI imaging showed a heterogeneous enhanced lesion in the left parieto-occipital lobe and the leptomeninges of the occipital lobe and the cerebellum. The lesion was resected and confirmed to be ALK-negative ALCL by pathological examination. Then, the patient received 10 cycles of chemotherapy with high-dose methotrexate (HD-MTX) and whole-brain radiotherapy. The patient recovered well and was regularly followed up. He was free of symptoms without recurrence on imaging examination 3 years later. ALCL is a rare type of PCNSL. HD-MTX combined with radiation is an effective therapeutic approach. However, further prospective studies with a large number of patients are needed to identify the biological characteristics of this rare type of PCNSL.

Background and aims Observational studies showed that coronavirus disease 2019 (COVID-19) attacks universally and its most menacing progression uniquely endangers the elderly with cardiovascular disease (CVD). The causal association between COVID-19 infection or its severity and susceptibility of atrial fibrillation (AF) remains unknown. Methods and results The bidirectional causal relations of COVID-19 (including COVID-19, hospitalized COVID-19 compared with not hospitalized COVID-19, hospitalized COVID-19 compared with population, and severe COVID-19) and AF are determined by using two-sample Mendelian randomization (MR) analysis. Genetically predicted severe COVID-19 was not significantly associated with risk of AF [odds ratio (OR), 1.037; 95% confidence interval (CI), 1.005-1.071; P = 0. 023, q = 0.115]. In addition, genetically predicted AF was also not causally associated with severe COVID-19 (OR, 0.993; 95% CI, 0.888-1.111; P = 0.905, q = 0.905). There was no evidence to support association between of genetically determined COVID-19 and risk of AF (OR, 1.111; 95% CI, 0.971-1.272; P = 0.127, q = 0.318), and vice versa (OR, 1.016; 95% CI, 0.976-1.058; P = 0.430, q = 0.851). Besides, no significant association was observed for hospitalized COVID-19 with AF. MR-Egger indicated no evidence of directional pleiotropy. Conclusion Overall, this MR study provides no clear support that COVID-19 is causally associated with the risk of AF.

Childhood brucellosis present various non-specific clinical symptoms, and limited laboratory data exist for clinical diagnosis. A better understanding of these clinical and laboratory characteristics can avoid clinical misdiagnosis and mistreatment. In this case-series study, a total of 78 children with confirmed diagnosis of brucellosis were evaluated retrospectively. We observed that the incidence rate was higher in the first two quarters every year. The most common symptom was fever. Osteoarticular involvement was found in 44.87% of the patients. Laboratory tests showed that the values of erythrocyte sedimentation rate(ESR), C-reactive protein(CRP), Hemoglobin(Hb), Neutrophils(NEU), Alanine aminotransfease(ALT) and Ferritin in childhood brucellosis with osteoarticular involvement had significant differences than those without osteoarticular involvement or control group (P

To evaluate the ocular and systemic determinants of the choriocapillaris flow deficits percentage (CC FD%) in normal eyes.Observational cross-sectional study.Healthy Chinese participants without ocular or systemic diseases underwent detailed ophthalmic evaluations, including swept-source optical coherence tomography angiography (SS-OCTA) with 6- × 6-mm macular choriocapillaris images. The CC FD% was assessed in circular regions with diameters of 1.0 and 5.0 mm, rings with 1.0- to 2.5-mm diameters, and rings with 2.5- to 5.0-mm diameters.The study included 830 individuals (mean ± SD age 58.66 ± 8.75 years). CC FD% (mean ± SD 22.05% ± 1.13%) was the lowest in the 2.5- to 5.0-mm ring, followed by the 1.0- to 2.5-mm ring, and highest in the 1.0-mm circle. In multivariable analysis, a higher CC FD% was associated with older age (β = 0.16 [95% confidence interval {CI}] 0.08-0.24, P.001), higher intraocular pressure (β = 0.34 [95% CI 0.25-0.42], P.001), higher serum concentrations of high-density lipoprotein cholesterol (β = 0.24 [95% CI 0.17-0.32], P.001), and lower image quality score (β = -0.22 [95% CI -0.30 to -0.14], P.001). CC FD% was also independent of axial length.In healthy adult Chinese individuals, a higher CC FD% was associated with older age, higher intraocular pressure, and higher high-density lipoprotein cholesterol serum concentration. These factors may influence clinical assessments of the choriocapillaris. The lack of an association between CC FD% and axial length is consistent with similar findings for Bruch membrane thickness, macular retinal thickness, and macular retinal pigment epithelium cell density.

Rationale and Objectives To detect the diffusion characteristics of lumbosacral nerve roots in patients with chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) and further to explore their correlations with electrophysiological parameters of lower extremity nerves. Materials and Methods Eighteen CIDP patients and 18 age and sex-matched healthy volunteers were enrolled in this study from August 2019 to August 2020. Axial diffusion tensor imaging (DTI) of lumbosacral plexus was performed in all subjects and fractional anisotropy (FA), axial diffusivity (AD), radial diffusivity (RD), and mean diffusivity (MD) of lumbosacral nerve roots were measured. Two-sample t test or Mann–Whitney U test was used to compare the difference of DTI parameters between two groups. Receiver operating characteristic curves were plotted to determine the diagnostic accuracy. All patients also underwent nerve conduction studies. Correlations between DTI parameters of lumbosacral nerve roots and electrophysiological parameters were analyzed with Pearson or Spearman coefficients. Results CIDP patients showed significantly lower FA as well as higher AD, RD, and MD values of lumbosacral nerve roots (FA:0.24±0.054, 0.32±0.044; AD:2.31±0.256, 2.11±0.230 (×10−3mm2/s); RD:1.28±0.189, 1.13±0.106 (×10−3mm2/s); MD:1.68±0.268, 1.45±0.186 (×10−3mm2/s) in CIDP and control group, respectively, all p Conclusion DTI can be used to assess the microstructure alterations of lumbosacral nerve roots in CIDP patients. FA and RD may serve as potential markers reflecting the conduction function of tibial and common peroneal nerves.

In this paper, the authors analyze the algorithm optimization and anomaly detection simulation based on extended jarvis-patrick clustering and outlier detection. We perform detection by using the jarvis-patrick graph-based clustering method. After that, to further improve the false alarm rate (FAR) of the algorithm, we use an extra outlier detection step combined with our proposed EJP to create a new anomaly detection method called LD-EJP. Using LD-EJP, the false alarm rate improved much (experiments show that the false alarm rate can reach 4.1% while the best JP clustering can reach is 7.4%). Then, we tested LD-EJP against two other anomaly detection methods using k-means and LGCCB, showing that our algorithm has a better detection rate and false alarm rate than these two clustering-based anomaly detection methods. In addition, the detection rate and false positives of the algorithm also have some room for improvement. In the labeling process, the proportion of anomaly clusters to normal clusters needs to be manually adjusted to find a better detection rate. In addition, the detection rate we chose can consume some of the detection rate gained in extended JP clustering to have the LD-EJP obtain a better FAR. Therefore, our future work contains finding or proposing another outlier detection algorithm with better performance than our LD-EJP method.

For continuous monitoring of cardiovascular diseases, this paper presents a novel framework for heart sound acquisition. The proposed approach uses compressed sensing for signal sampling, and a two-stage reconstruction is developed for reconstruction. The first stage aims to give a tentative recovered signal, on which a peak detection technique is developed to identify whether there is a peak in current segment and, if so, its location. With such information, an adaptive dictionary is selected for the second round reconstruction. Because the selected dictionary is adaptive to the morphology of current frame, the signal reconstruction performance is consequently promoted. Experiment results indicate that a satisfactory performance can be obtained when the frame length is 256 and the signal morphology is divided into 16 categories. Furthermore, the proposed algorithm is compared with a series of counterparts and the results well demonstrate the advantages of our proposal, especially at high compression ratios.

Purpose : The association between visual impairment (VI) and the risk of dementia has been poorly understood. We sought to investigate the VI-dementia relationship in the UK Biobank Study. Design : Prospective cohort study Methods : A total of 117,187 volunteers (aged 40-69 years) deemed free of dementia at baseline were included. Habitual distance visual acuity worse than 0.3 logMAR units in the better-seeing eye was used to define VI. The incident dementia was based on electronically linked hospital inpatient and death records. Results : During a median follow up of 5.96 years, the presence of VI was significantly associated with incident dementia (HR=1.78, 95% CI: 1.18-2.68, P=0.006). There was a clear trend between the severity of VI and the risk of dementia (P for trend=0.002). Conclusions : We found VI was associated with increased risk of dementia, with a progressively greater risk among those with worse visual acuity. Our findings suggested that visual impairment might be a modifiable risk factor for dementia and highlighted the potential value of VI elimination to delay the manifestation of dementia.

The development of technologies that can generate Deepfake videos is expanding rapidly. These videos are easily synthesized without leaving obvious traces of manipulation. Though forensically detection in high-definition video datasets has achieved remarkable results, the forensics of compressed videos is worth further exploring. In fact, compressed videos are common in social networks, such as videos from Instagram, Wechat, and Tiktok. Therefore, how to identify compressed Deepfake videos becomes a fundamental issue. In this paper, we propose a two-stream method by analyzing the frame-level and temporality-level of compressed Deepfake videos. Since the video compression brings lots of redundant information to frames, the proposed frame-level stream gradually prunes the network to prevent the model from fitting the compression noise. Aiming at the problem that the temporal consistency in Deepfake videos might be ignored, we apply a temporality-level stream to extract temporal correlation features. When combined with scores from the two streams, our proposed method performs better than the state-of-the-art methods in compressed Deepfake videos detection.

RATIONALE: In predominantly White populations, early chronic obstructive pulmonary disease (COPD) (i.e., COPD in people aged

Photodiodes (PDs) which convert optical signals to electrical signals have been widely used in many fields. In contrast to pin PDs, in uni-traveling carrier (UTC) PDs, the absorption layer is p type doped and only photo-generated electrons are transported, which results in several superior characteristics including super high speed operation, high saturation current and low voltage operation. In this work, high speed InP based waveguide UTC-PDs have been fabricated by Zn diffusion doping for the first time to our knowledge, in contrast to the in situ doping of previous UTC-PDs. Diffusion doping will make it much easier for the UTC-PDs to be integrated with high quality passive waveguide and multi-quantum well (MQW) based active devices for photonic integrated circuit (PIC) applications. UTC-PDs with 3 × 15, 3 × 25 and 3 × 50 µm2 absorption areas are fabricated, all having over 50 GHz bandwidth at -1V bias voltage. At -3 V bias, the RF power at 40 GHz shows no saturation at up to 17 mA photocurrent for the 3 × 25 µm2 PD.

Accident analysis models play a significant role in accident investigation and analysis. Most traditional models, usually based on a chain of events, have serious limitations when used for complex systems, especially for coal mine accident. HFACS is a widely accepted accident cause model with complex structure and content. Based on HFACS and accident investigation need in China, a new systemic accident analysis method “Six-hierarchy model” was proposed. And the index database of accident causal factors at each hierarchy of coal mine accidents was established. This model provides clear guidance for the prevention of accidents from six levels, which helps to the targeted rectification after accidents and the improvement of safety management level. The source, application scope, advantages and disadvantages of HFACS and six-hierarchy model are compared. The “2.14” gas explosion accident in Sunjiawan coal mine is analyzed by using HFACS and six-hierarchy model and the safety management suggestions are put forward. The study shows that the Six-hierarchy model provides a clear guide for the prevention of accidents from six levels. And it benefits the targeted rectification after accidents and the continuous improvement of safety management.

Domain adaptation aims to exploit domain-invariant features by aligning the cross-domain distributions in the manifold subspace for applying the classifier trained on the source domain to the target domain. However, two limitations may still deteriorate their performances: (1) the influences of noisy or irrelevant features in the original feature space are ignored, which may unexpectedly hurt the classification of target samples; (2) the graph constructed directly in the original data space cannot accurately capture the inherent local manifold structures of high-dimensional data due to the curse of dimensionality, which may seriously mislead the transferable features learning. In this paper, we propose a novel approach to address these problems, referred to as joint Adaptive Dual Graph and Feature Selection for domain adaptation (ADGFS). Specifically, feature selection can characterize the relative importance of different features through a scaling factor, which enables ADGFS to not only reduce the impacts of noisy or irrelevant features on knowledge transfer but also learn informative domain-invariant features. Meanwhile, ADGFS adaptively optimizes the dual graph by learning the similarity matrices of both instance-level and feature-level graphs in the projected low-dimensional manifold subspace rather than the original high-dimensional space, such that the intrinsic local manifold structures of data can be captured precisely. Moreover, ADGFS simultaneously aligns the marginal and conditional probability distributions in the nonnegative matrix factorization framework to narrow the distribution discrepancies between the two different domains, which can adequately transfer knowledge from the source domain to the target domain. Comprehensive experiments on four benchmark datasets can demonstrate that the effectiveness of the proposed approach in cross-domain image classification.

Background: Prurigo nodularis (PN) as an extremely pruritic and hyperplastic chronic dermatosis induces psychologically and physiologically stressful responses. PN-induced responses in the hypothalamic-pituitary-adrenal (HPA), hypothalamic-pituitary-gonadal (HPG) axes and endocannabinoid system (ECS) are abnormal. Extant studies on the PN’s pathogenesis mostly focused on the PN’s psychological responses. To date, the PN’s physiological responses remain not been fully uncovered yet. Objectives: To investigate the PN-induced physiological responses via the levels of five steroids and two endocannabinoids combined with their ratios in plasma and examine the association between the psychological and physiological responses. Materials and methods: Thirty-six patients with PN, 36 age- and gender-matched healthy controls were recruited. The PN’s psychological symptoms including pruritus severity, pain and life quality were measured with the visual analogue scale, the prurigo score index, numerical rating scale, verbal rating scale and dermatology life quality index. Their concentrations of steroids and endocannabinoids were determined with liquid chromatography-tandem mass spectrometry. Results: Compared to controls, the PN patients showed lower plasma levels in cortisol, cortisone, N-arachidonoyl-ethanolamine (AEA), and the ratio of DHEA to 1-arachydonoyl glycerol (1-AG), which negatively moderately and over correlated with PN’s symptoms, especially with the pruritus severity. Additionally, the PN patients exhibited higher levels in the ratios of testosterone and 1-AG to cortisol, which positively moderately and over correlated with pruritus severity. Thus, the seven biomarkers would be sensitive and reliable biomarkers for assessing the pruritus severity of PN because they met the screening criteria that the biomarkers show intergroup differences and showed moderate or over correlation with the pruritus severity of PN. Conclusions: To the best of our knowledge, this is the first study exploring PN-induced physiological responses. The findigs suggest that alterations in these three endocrine systems may lead to new insights to psychological mechanisms and responses to prurigo nodularis.

The 2016 WHO classification introduced the category of high-grade B-cell lymphoma (HGBL), which includes one poorly understood subset, blastoid-HGBL. Establishing the diagnosis and distinguishing blastoid-HGBL from B-acute lymphoblastic leukemia (B-ALL) in bone marrow can be challenging. We assessed 31 cases of blastoid-HGBL diagnosed initially in bone marrow and compared this group to 36 cases of B-ALL using immunophenotyping, fluorescence in situ hybridization, and targeted next generation sequencing analysis. The 31 blastoid-HGBL cases included 14 HGBL with MYC and BCL2 and/or BCL6 rearrangements (double hit lymphoma, DHL), 13 HGBL, not otherwise specified (NOS), and four cases with TdT expression that were difficult to classify. Compared with B-ALL, blastoid-HGBL cases more often showed increased intensity/bright expression of CD20, CD38, CD45, BCL-6, and MYC, and less frequent bright expression of CD10 and TdT. Cases of blastoid-HGBL also more frequently had MYC rearrangement, a complex karyotype and TP53 mutation (p

OBJECTIVE Previous studies have suggested the use of 1.0 g/kg of 20% mannitol at the time of skin incision during neurosurgery in order to improve brain relaxation. However, the incidence of brain swelling upon dural opening is still high with this dose. In the present study, the authors sought to determine a better timing for mannitol infusion. METHODS One hundred patients with midline shift who were undergoing elective supratentorial tumor resection were randomly assigned to receive early (immediately after anesthesia induction) or routine (at the time of skin incision) administration of 1.0 g/kg body weight of 20% mannitol. The primary outcome was the 4-point brain relaxation score (BRS) immediately after dural opening (1, perfectly relaxed; 2, satisfactorily relaxed; 3, firm brain; and 4, bulging brain). The secondary outcomes included subdural intracranial pressure (ICP) measured immediately before dural opening; serum osmolality and osmole gap (OG) measured immediately before mannitol infusion (T0) and at the time of dural opening (TD); changes in serum electrolytes, lactate, and hemodynamic parameters at T0 and 30, 60, 90, and 120 minutes thereafter; and fluid balance at TD. RESULTS The time from the start of mannitol administration to dural opening was significantly longer in the early administration group than in the routine administration group (median 66 [IQR 55–75] vs 40 [IQR 38–45] minutes, p < 0.001). The BRS (score 1/2/3/4, n = 14/26/9/1 vs 3/25/18/4, p = 0.001) was better and the subdural ICP (median 5 [IQR 3–6] vs 7 [IQR 5–10] mm Hg, p < 0.001) was significantly lower in the early administration group than in the routine administration group. Serum osmolality and OG increased significantly at TD compared to levels at T0 in both groups (all p < 0.001). Intergroup comparison showed that serum osmolality and OG at TD were significantly higher in the routine administration group (p < 0.001 and = 0.002, respectively). Patients who had received early administration of mannitol had more urine output (p = 0.001) and less positive fluid balance (p < 0.001) at TD. Hemodynamic parameters, serum lactate concentrations, and incidences of electrolyte disturbances were comparable between the two groups. CONCLUSIONS Prolonging the time interval between the start of mannitol infusion and dural incision from approximately 40 to 66 minutes can improve brain relaxation and decrease subdural ICP in elective supratentorial tumor resection.

Background Alpha-1 antitrypsin deficiency (AATD) is an autosomal co-dominant condition that predisposes to emphysema, cirrhosis, panniculitis, and vasculitis. Under-recognition has prompted efforts to enhance early detection and testing of at-risk individuals. Direct-to-consumer (DTC) genetic testing represents an additional method of detection. Research Question The study addressed three questions: 1) Does a DTC testing service identify previously undetected individuals with AATD? 2) What was the time interval between initial AATD-related symptoms and initial diagnosis of AATD in such individuals? and 3) What was the behavioral impact of learning about a new diagnosis of AATD through a DTC test? Study Design and Methods In this cross-sectional study, 195,014 individuals responded to a survey within the 23andMe, Inc. research platform. Results Among 195,014 study participants, the allele frequency for either the PI*S and PI*Z AATD variants was 21.6% (6.5% for PI*Z and 15.1% for PI*S); 0.63% were PI*ZZ, half of whom reported having a physician confirm the diagnosis. Approximately 27% of those with physician-diagnosed AATD reported first becoming aware of AATD through the DTC test. Among those newly-aware participants, the diagnostic delay interval was 22.3 years. Participants frequently shared their DTC test results with healthcare providers (HCPs) and the reported impact of learning a diagnosis of AATD was high. For example, 51.1% of PI*ZZ individuals shared their DTC result with an HCP. The odds ratio for PI*ZZ smokers to report smoking reduction as a result of receiving the DTC result was 1.7 [CI 1.4, 2.2] compared to those without a Z allele and for reduced alcohol consumption was 4.0 [CI 2.6, 5.9]. Interpretation In this largest available report on DTC testing for AATD, this test, in combination with clinical follow-up, can help to identify previously undiagnosed AATD patients. Moreover, receipt of the DTC AATD report was associated with positive behavior change, especially among those with risk variants.

Nonalcoholic fatty liver disease (NAFLD) encompasses a spectrum of pathologies, ranging from steatosis to nonalcoholic steatohepatitis (NASH). The factors promoting the progression of steatosis to NASH are still unclear. Recent studies suggest that mitochondrial lipid composition is critical in NASH development. Here, we showed that CDP-DAG synthase 2 (Cds2) was downregulated in genetic or diet-induced NAFLD mouse models. Liver-specific deficiency of Cds2 provoked hepatic steatosis, inflammation and fibrosis in five-week-old mice. CDS2 is enriched in mitochondria-associated membranes (MAMs), and hepatic Cds2 deficiency impaired mitochondrial function and decreased mitochondrial PE levels. Overexpression of phosphatidylserine decarboxylase (PISD) alleviated the NASH-like phenotype in Cds2f/f;AlbCre mice and abnormal mitochondrial morphology and function caused by CDS2 deficiency in hepatocytes. Additionally, dietary supplementation with an agonist of peroxisome proliferator-activated receptor alpha (PPARα) attenuated mitochondrial defects and ameliorated the NASH-like phenotype in Cds2f/f;AlbCre mice. Finally, Cds2 overexpression protected against high-fat diet-induced hepatic steatosis and obesity. Thus, Cds2 modulates mitochondrial function and NASH development.

We analyze the coupling between an adjacent field ring and coupling ring of the two-dimensional optical resonator array to obtain topological edge state. Using finite-difference time-domain method, the energy band structure, light field distribution and transmission spectrum are simulated to investigate the optimised coupling parameters under the topological boundary state. When the coupling coefficient is approximately 0.775, the system can form a topological state in the wavelength of 1.55 μm-1.56 μm while output can maintain a flat spectral response with normalized power above 0.9. The research in this paper will provide a basis for the potential application of topological edge states of tow-dimensional coupled resonant optical waveguide in the field of photonic communication.

In this paper, we demonstrate the transmission spectral and surrounding refractive index (SRI) sensing performance of long-period fiber gratings (LPFGs) at 2 m waveband. The cladding modes operating at 2 m waveband show higher SRI sensitivity than 1.55 m waveband. The influence of a little off resonance region of grating period, refractive index modulation and duty cycle of gratings are analyzed theoretically. Additionally, the LPFGs of LP08 mode near dispersion turning point operating at 2 m waveband are investigated theoretically and experimentally. Mode splitting brings out with the increasing SRI. After the resonant dip splitting into two dips (dip A and dip A), the SRI sensitivity of dip A are measured to be as high as -8495.2 nm/RIU in the SRI region of 1.445-1.457. The proposed 2-m-waveband LPFG makes it a potential in biochemical and environmental monitoring.

A five-wavelength-switchable thulium-doped fiber laser with a fiber Bragg grating (FBG)-based five-channel transmitting Fabry-Pérot (F-P) filter and a multi-coupler based triple-ring cavity (CTRC) filter, is experimentally proposed. For the first time, a multi-channel F-P filter and specially-designed coupler-based compound cavity filter were combined to realize multi-wavelength-switchable operation, with single-longitudinal-mode (SLM) lasing for each wavelength channel in a 2 µm band fiber laser. The theoretical analysis of the CTRC filter and the principle for achieving SLM operation by corporately using the F-P filter and CTRC filter were presented in detail. The experimental results demonstrated the good performance of the proposed filters. Five lasing wavelengths (1941.13 nm, 1941.22 nm, 1941.30 nm, 1941.39 nm, and 1941.47 nm) were obtained with good switchability, and each operated in a stable SLM state. The OSNR of each lasing wavelength was higher than 50 dB, and the linewidths of all obtained wavelengths were less than 8 kHz.