Search

Background and purpose Delineating organs at risk (OARs) on computed tomography (CT) images is an essential step in radiation therapy; however, it is notoriously time-consuming and prone to inter-observer variation. Herein, we report a deep learning-based automatic segmentation (AS) algorithm (WBNet) that can accurately and efficiently delineate all major OARs in the entire body directly on CT scans. Materials and methods We collected 755 CT scans of the head and neck, thorax, abdomen, and pelvis and manually delineated 50 OARs on the CT images. The CT images with contours were split into training and test sets consisting of 505 and 250 cases, respectively, to develop and validate WBNet. The volumetric Dice similarity coefficient (DSC) and 95th-percentile Hausdorff distance (95% HD) were calculated to evaluate delineation quality for each OAR. We compared the performance of WBNet with three AS algorithms: one commercial multi-atlas-based automatic segmentation (ABAS) software, and two deep learning-based AS algorithms, namely, AnatomyNet and nnU-Net. We have also evaluated the time saving and dose accuracy of WBNet. Results WBNet achieved average DSCs of 0.84 and 0.81 on in-house and public datasets, respectively, which outperformed ABAS, AnatomyNet, and nnU-Net. WBNet could reduce the delineation time significantly and perform well in treatment planning, with clinically acceptable dose differences compared with those in manual delineation. Conclusion This study shows the feasibility and benefits of using WBNet in clinical practice.

Motivation Single-cell sequencing assay for transposase-accessible chromatin (scATAC-seq) provides new opportunities to dissect epigenomic heterogeneity and elucidate transcriptional regulatory mechanisms. However, computational modeling of scATAC-seq data is challenging due to its high dimension, extreme sparsity, complex dependencies and high sensitivity to confounding factors from various sources. Results Here, we propose a new deep generative model framework, named SAILER, for analyzing scATAC-seq data. SAILER aims to learn a low-dimensional nonlinear latent representation of each cell that defines its intrinsic chromatin state, invariant to extrinsic confounding factors like read depth and batch effects. SAILER adopts the conventional encoder-decoder framework to learn the latent representation but imposes additional constraints to ensure the independence of the learned representations from the confounding factors. Experimental results on both simulated and real scATAC-seq datasets demonstrate that SAILER learns better and biologically more meaningful representations of cells than other methods. Its noise-free cell embeddings bring in significant benefits in downstream analyses: clustering and imputation based on SAILER result in 6.9% and 18.5% improvements over existing methods, respectively. Moreover, because no matrix factorization is involved, SAILER can easily scale to process millions of cells. We implemented SAILER into a software package, freely available to all for large-scale scATAC-seq data analysis. Availability and implementation The software is publicly available at https://github.com/uci-cbcl/SAILER. Supplementary information Supplementary data are available at Bioinformatics online.

Aims There is paucity of clinical data comparing continuous infusion (CI) with bolus injection (BI) of intravenous loop diuretics in patients with acute decompensated heart failure (ADHF) and chronic renal dysfunction. This study aimed to compare the efficacy and safety of CI versus BI intravenous furosemide administration in patients with ADHF and moderate chronic renal insufficiency. Methods and results Acute decompensated heart failure and moderate chronic renal insufficiency [with estimated glomerular filtration rate (eGFR) 15.0–44.9 mL/min/1.73 m2] were randomized to start intravenous furosemide by BI or by a 6 h CI. End points included freedom from congestion at 72 h, the degree of dyspnoea assessed using the 0–10 Borg's category ratio scale, net daily urine output, weight loss during the study, length of hospital stay, total urinary sodium excretion, and development of acute kidney injury or electrolyte disturbance. After 72 h of treatment, the rate of the primary endpoint of freedom from congestion in the CI group was significantly higher than that in the BI group (69.05% vs. 43.59%, P = 0.02). The modified Borg scale indicated patients in the CI group had lower dyspnoea score than those in the BI group at 48 h (4.29 ± 1.23 vs. 5.97 ± 1.56; P = 0.02) and 72 h (1.15 ± 0.35 vs. 2.66 ± 0.83; P = 0.003). There were other significant differences favouring the CI group with regard to net urine output at 72 h (5145.98 ± 621.37 mL vs. 3755.95 ± 456.93 mL; P = 0.007), the mean body weight loss (4.72 ± 1.01 kg vs. 3.53 ± 0.73 kg; P = 0.02) and the total urinary sodium excretion (385.05 ± 38.15 vs. 320.33 ± 37.67; P = 0.02). The length of hospitalization in the CI group was significantly shorter than that in the BI group (10.36 ± 4.20 days vs. 15.68 ± 6.15 days; P = 0.02). No significant differences were observed between groups in the frequency of acute kidney injury, tinnitus, electrolyte disturbance or mortality. Conclusions Continuous intravenous infusion of furosemide resulted in significantly greater diuresis than bolus administration of an equal dose in patients with moderate chronic renal insufficiency and ADHF, while no differences emerged in terms of side effects or mortality.

Response Evaluation Criteria in Solid Tumors (RECIST 1.1) is classical and popular for public. However, there are some problems recently. For example, partial response ranges from 30% to 99%, objective response is dichotomous, so there may be some heterogeneity. New metrics for evaluating the efficacy have been investigating, such as early tumor shrinkage, time to response and depth of response (DpR). DpR has been used in hematologic malignancies and is considered as a predictor of efficiency. In solid tumors, DpR was firstly proposed by Mansmann et al. in metastatic colorectal cancer (mCRC) and defined as the percentage of tumor shrinkage, which is a continuous metric, and could avoid the information loss due to dichotomization of responses that has been widely applied to several kinds of solid tumors. Some authors have found associations between DpR and OS, DpR is a valuable surrogate endpoint for mCRC, metastatic breast cancer, metastatic melanoma and advanced pancreatic cancer. However, the predictive value of DpR is still uncertain in the research of lung cancer and gastric cancer. Which indicating that a mature and unified application standard has not yet been formed for DpR. This article summarizes researches on the DpR as a predictor of the long-term outcomes for solid tumors, it also discusses the challenges and limitations in the applications of DpR.

What is known and objective Smoking has a notable influence on the efficacy of medications for lung cancer. Previous studies illustrated the correlation between smoking and the efficacy of first-line Epidermal Growth Factor Receptors-Tyrosine Kinase Inhibitors (EGFR-TKIs). The benefit of smokers in immunotherapy was still controversial. Here, we investigated the impact of smoking on clinical outcomes of molecularly targeted therapies or immunotherapy in Non-Small Cell Lung Cancer (NSCLC). Methods We performed meta-analysis including trials comparing EGFR-TKIs, Anaplastic Lymphoma Kinase (ALK) inhibitors or Immune Checkpoint Inhibitors (ICIs) against chemotherapy in NSCLC. The Progression-Free Survival (PFS)-Hazard Ratios (HRs) of two groups served as the index and we used random effects to pool outcomes. Results and discussion Twenty randomized trials were selected. Compared with chemotherapy, treatment with EGFR-TKIs had similar benefit in never-smokers (PFS: HR = 0.46, 95% CI 0.30 to 0.69) and smokers (PFS: HR = 0.68, 95% CI 0.50 to 0.91; p = 0.135) while non-smokers (PFS: HR = 0.32, 95% CI 0.23 to 0.44) had better benefit from first-line EGFR-TKIs than smokers (PFS: HR = 0.54, 95% CI 0.41 to 0.71; p = 0.02). Treatment with ALK inhibitors had similar benefits in never-smokers (PFS: HR = 0.43, 95% CI 0.35 to 0.53) and smokers (PFS: HR = 0.56, 95% CI 0.44 to 0.71; p = 0.406). The benefit of ICIs in smokers (PFS: HR = 0.79, 95% CI 0.64 to 0.98) was significantly greater than never-smokers (PFS: HR = 1.81, 95% CI 1.27 to 2.57; p = 0.004). What is new and conclusion Smoking status is an important clinical predictor of therapy in NSCLC. Never-smokers and smokers have similar benefit with EGFR-TKIs therapy compared with chemotherapy, while never-smokers have greater benefit after first-line EGFR-TKIs therapy. There was similar benefit in never-smokers and smokers when using ALK inhibitors over chemotherapy. Additionally, ICIs treatment over chemotherapy leads to more favourable PFS in smokers both in first-line and second-line settings.

What is known and objective The efficacy of epidermal growth factor receptor tyrosine kinase inhibitors (EGFR-TKIs) in patients diagnosed with non-small-cell lung cancer (NSCLC) has been confirmed by a large number of studies. However, hepatotoxicity caused by EGFR-TKIs has not been widely investigated. This review compares the hepatotoxicity of different EGFR-TKIs through a network meta-analysis. Methods PubMed, EMBASE, Cochrane Library and ClinicalTrials.gov were systematically searched from their individual inceptions to 20 May 2020 with the goal of identifying randomized controlled trials (RCTs) reporting hepatotoxicity in NSCLC patients receiving EGFR-TKIs. A random-effects pairwise meta-analysis and network meta-analysis were performed within a frequentist framework. Risk ratios (RRs) and 95% confidence intervals (CIs) were calculated. Results Twelve eligible RCTs, including data from 6,280 patients diagnosed with NSCLC, were analysed. In our network meta-analysis, gefitinib was associated with a higher risk for hepatotoxicity compared to placebo (RR, 2.55; 95% CI, 1.32-4.89) and dacomitinib (RR, 2.60; 95% CI, 1.30-5.20) in terms of all-grades alanine transaminase (ALT) elevation. As for all-grades aspartate transaminase (AST) elevation, gefitinib and erlotinib showed a significantly increased risk for hepatotoxicity compared to afatinib, dacomitinib and placebo (erlotinib vs. afatinib: RR, 1.85; 95% CI, 1.05-3.24; erlotinib vs. dacomitinib: RR, 1.68; 95% CI, 1.19-2.36; erlotinib vs. placebo: RR, 3.38; 95% CI, 1.69-6.73; gefitinib vs. afatinib: RR, 2.23; 95% CI, 1.32-3.79; gefitinib vs. dacomitinib: RR, 2.03; 95% CI, 1.51-2.73; gefitinib vs. placebo: RR, 4.08; 95% CI, 2.11-7.91). There was a high risk of high-grade ALT elevation in patients treated with gefitinib compared to patients treated with erlotinib (RR, 4.31; 95% CI, 2.15-8.66), dacomitinib (RR, 6.95; 95% CI, 1.85-26.05) or placebo (RR, 8.38; 95% CI, 1.56-45.01). No statistically significant differences were identified among the five agents analysed in terms of all-grades TB elevation and high-grade AST elevation. The surface under the cumulative ranking curve (SUCRA) revealed that gefitinib showed a potentially higher risk for ALT and AST elevation compared to other EGFR-TKIs regardless of grade. What is new and conclusion Current evidence indicates that the association between afatinib or dacomitinib and risk of liver enzyme elevation remains uncertain in patients diagnosed with NSCLC. Some evidence suggests that gefitinib and erlotinib may be associated with a significantly increased risk for hepatotoxicity in patients with NSCLC. However, given that the elevation of liver enzymes was not definitely associated with EGFR-TKIs and publication bias, further studies are required to confirm these results.

Green-computing technology and energy-saving design have become the focus of research in various fields in recent years. As a ubiquitously deployed infrastructure, WiFi can be considered as a platform for green sensing, and a plethora of efforts have been made in WiFi-based passive detection recently. However, little work has been done on the exercise activity recognition. In this paper, we propose SEARE, a novel energy-efficient solution using WiFi for exercise activity recognition. It is prototyped by fine-grained CSI extracted from existing commercial WiFi devices. Different from traditional features like mean or max value exploited in previous activity recognition works, involving either time or frequency information, we select CSI-waveform shape as activity feature, which contains the information from both of these two domains. A series of de-noise methods are designed, including low-pass, PCA, and median filtering, where PCA can remove the in-band noise that traditional low-pass filters fail to do. Finally the evaluation of activities quality can be made. Extensive experimental result validates the great performance of SEARE in both LOS and NLOS scenarios, with average recognition accuracies of 97.8 and 91.2 percent respectively.

BACKGROUND: The recent development of metagenomic sequencing makes it possible to massively sequence microbial genomes including viral genomes without the need for laboratory culture. Existing reference-based and gene homology-based methods are not efficient in identifying unknown viruses or short viral sequences from metagenomic data. METHODS: Here we developed a reference-free and alignment-free machine learning method, DeepVirFinder, for identifying viral sequences in metagenomic data using deep learning. RESULTS: Trained based on sequences from viral RefSeq discovered before May 2015, and evaluated on those discovered after that date, DeepVirFinder outperformed the state-of-the-art method VirFinder at all contig lengths, achieving AUROC 0.93, 0.95, 0.97, and 0.98 for 300, 500, 1000, and 3000 bp sequences respectively. Enlarging the training data with additional millions of purified viral sequences from metavirome samples further improved the accuracy for identifying virus groups that are under-represented. Applying DeepVirFinder to real human gut metagenomic samples, we identified 51,138 viral sequences belonging to 175 bins in patients with colorectal carcinoma (CRC). Ten bins were found associated with the cancer status, suggesting viruses may play important roles in CRC. CONCLUSIONS: Powered by deep learning and high throughput sequencing metagenomic data, DeepVirFinder significantly improved the accuracy of viral identification and will assist the study of viruses in the era of metagenomics.

This paper presents a policy-driven sequential image augmentation approach for image-related tasks. Our approach applies a sequence of image transformations (e.g., translation, rotation) over a training image, one transformation at a time, with the augmented image from the previous time step treated as the input for the next transformation. This sequential data augmentation substantially improves sample diversity, leading to improved test performance, especially for data-hungry models (e.g., deep neural networks). However, the search for the optimal transformation of each image at each time step of the sequence has high complexity due to its combination nature. To address this challenge, we formulate the search task as a sequential decision process and introduce a deep policy network that learns to produce transformations based on image content. We also develop an iterative algorithm to jointly train a classifier and the policy network in the reinforcement learning setting. The immediate reward of a potential transformation is defined to encourage transformations producing hard samples for the current classifier. At each iteration, we employ the policy network to augment the training dataset, train a classifier with the augmented data, and train the policy net with the aid of the classifier. We apply the above approach to both public image classification benchmarks and a newly collected image dataset for material recognition. Comparisons to alternative augmentation approaches show that our policy-driven approach achieves comparable or improved classification performance while using significantly fewer augmented images. The code is available at https://github.com/Paul-LiPu/rl_autoaug.

Background: Trastuzumab is associated with the risk of cardiotoxicity. Here, we aim to explore interventions for preventing trastuzumab-related cardiotoxic effects in breast cancer patients.Methods: A systematic review was performed including trials of breast cancer patients with intervention to prevent cardiotoxicity of trastuzumab. Trials were searched through databases including PubMed, Embase, and Cochrane Library.Results: Eight RCTs were included. Five trials reported the outcomes of short-duration interventions, including 6-month and 9-week durations, and only 9-week treatment has a significant difference from the 12-month group (OR 0.38; 95% CI 0.18–0.83) using cardiotoxicity as the outcome. However, 6-month treatment turned out to yield less occurrence of trastuzumab discontinuation (OR 0.32; 95% CI 0.24–0.42). Three trials reported interventions of cardioprotective drugs, and there is no significant difference shown in any cardioprotective group compared with placebo (cardiotoxicity outcome: angiotensin-converting enzyme inhibitor: OR 0.48; 95% CI 0.057–2.3; angiotensin receptor blocker: OR 1.3; 95% CI 0.12–14; β-blocker: OR 0.48; 95% CI 0.057–2.3; trastuzumab interruption outcome: angiotensin-converting enzyme inhibitor: OR 0.45; 95% CI 0.12–1.3; angiotensin receptor blocker: OR 0.87; 95% CI 0.15–4.8; β-blocker: OR 0.41; 95% CI 0.11–1.2).Conclusion: Only the 9-week group has a significant difference from the 12-month group using cardiotoxicity as the outcome. And 6-month treatment turned out to yield less occurrence of trastuzumab discontinuation. The use of cardioprotective drugs failed to prevent trastuzumab-related cardiotoxic effects in breast cancer patients.

High-frequency motions in the southeastern South China Sea (SCS) have rarely been investigated due to sparse field observations. The vertical distribution and temporal variation of internal tides (ITs) and near-inertial waves (NIWs) near the Nansha area in the southeastern SCS were studied using a mooring current dataset from December 2018 to June 2019 in this study. Results showed that ITs were mainly dominated by O1, K1, and M2. Tidal energy analysis indicates that the diurnal ITs were the most energetic components, followed by the semidiurnal ITs. Modal decomposition reveals that diurnal ITs were dominated by mode-3, rather than mode-1, as reported by previous studies. The horizontal kinetic energy (HKE) of diurnal ITs fluctuated within a limited range, almost unaffected by the background field. However, the HKE of semidiurnal ITs was dominated by mode-1 and more affected by the background field, especially at the end of March. Most observations showed the phase of the NIWs propagating upward and the energy propagating downward. During the northeast monsoon period, the near-inertial energy had a large magnitude due to strong wind forcing. In addition, the near-inertial energy peaked from the middle of March to the beginning of April because of the input of NIWs from afar. Overall, near-inertial energy was found concentrated above a 500–600 m depth in the southeastern SCS.

Background Peritoneal dialysis (PD) is a form of dialysis to replace the function of kidney, that uses the peritoneum as a dialysis membrane to remove metabolites and water retained in the body. Vitamin D deficiency is prevalent in patients treated with PD. This research investigated the correlation between serum 25-hydroxyvitamin D [25(OH)D] concentration and anemia, microinflammation, and oxidative stress in PD patients. Methods 62 PD patients and 56 healthy volunteers were recruited in this research. Serum concentrations of 25(OH)D and basic parameters of anemia were detected. The correlation between serum 25(OH)D concentration with anemia, oxidative stress, and microinflammatory state were analyzed. Results In the PD group, the concentration of 25(OH)D was lower than the healthy control (HC) group (p Conclusion In conclusion, reduced serum 25(OH)D concentrations in PD patients contribute to anemia, oxidative stress and microinflammatory state.

Radiation therapy is one of the most widely used therapies for cancer treatment. A critical step in radiation therapy planning is to accurately delineate all organs at risk (OARs) to minimize potential adverse effects to healthy surrounding organs. However, manually delineating OARs based on computed tomography images is time-consuming and error-prone. Here, we present a deep learning model to automatically delineate OARs in head and neck, trained on a dataset of 215 computed tomography scans with 28 OARs manually delineated by experienced radiation oncologists. On a hold-out dataset of 100 computed tomography scans, our model achieves an average Dice similarity coefficient of 78.34% across the 28 OARs, significantly outperforming human experts and the previous state-of-the-art method by 10.05% and 5.18%, respectively. Our model takes only a few seconds to delineate an entire scan, compared to over half an hour by human experts. These findings demonstrate the potential for deep learning to improve the quality and reduce the treatment planning time of radiation therapy. To keep radiation therapy from damaging healthy tissue, expert radiologists have to segment CT scans into individual organs. A new deep learning-based method for delineating organs in the area of head and neck performs faster and more accurately than human experts.

The evolution of energy, energy flux and modal structure of the internal tides (ITs) in the northeastern South China Sea is examined using the measurements at two moorings along a cross-slope section from the deep continental slope to the shallow continental shelf. The energy of both diurnal and semidiurnal ITs clearly shows a ~14-day spring-neap cycle, but their phases lag that of barotropic tides, indicating that ITs are not generated on the continental slope. Observations of internal tidal energy flux suggest that they may be generated at the Luzon Strait and propagate west-northwest to the continental slope in the northwestern SCS. Because the continental slope is critical-supercritical with respect to diurnal ITs, about 4.6 kJ/m2 of the incident energy and 8.7 kW/m of energy flux of diurnal ITs are reduced from the continental slope to the continental shelf. In contrast, the semidiurnal internal tides enter the shelf because of the sub-critical topography with respect to semidiurnal ITs. From the continental slope to the shelf, the vertical structure of diurnal ITs shows significant variation, with dominant Mode 1 on the deep slope and dominant higher modes on the shelf. On the contrary, the vertical structure of the semidiurnal ITs is stable, with dominant Mode 1.

Strut collisions can occur in tensegrity structures and seriously change the general mechanical behaviors of the structures. For many years, there have been few investigations into this topic. This paper systematically addresses the issue of strut collision modeling in tensegrity mechanical analyses. A general and efficient rigid body model for strut collision analysis is presented. A two-strut collision model is formulated using contact force methods. The normal contact force and the tangential friction force are evaluated by using the Lankarani–Nikravesh model and the modified Coulomb's friction law, respectively. Exact formulations of the generalized force vectors and the related tangent stiffness and tangent damping matrices are analytically derived. Fundamental issues related to collision detection and stiffness coefficient determination, are also addressed. Two representative examples involving both classical and clustered tensegrity structures are presented to demonstrate the effectiveness of the general framework, as well as to make a detailed study into the influence of collisions effects. Outcomes presented in this work will broaden the research discipline on tensegrity systems.

Intelligent human detection based on WiFi is a technique that has recently attracted a significant amount of interest from research communities. The use of ubiquitous WiFi to detect the number of queuing persons can facilitate dynamic planning and appropriate service provisioning. In this article, we propose HFD, one of the first schemes to leverage WiFi signals to estimate the number of queuing persons by employing classifiers from machine learning in a device-free manner. In the proposed HFD scheme, we first utilize the sliding window method to filter and remove the outliers. We extract two characteristics, skewness and kurtosis, as the identification features. Then, we use the support vector machine (SVM) to classify these two features to estimate the number of people in the current queue. Finally, we combine our scheme with the latest Fresnel Zone model theory to determine whether someone is in or out, and thus dynamically adjust the detected value. We implement a proof-of-concept prototype upon commercial WiFi devices and evaluate it in both conference room and corridor scenarios. The experimental results show that the accuracy of our proposed HFD detection can be maintained at about 90 percent with high robustness.

Registration is a fundamental task in medical robotics and is often a crucial step for many downstream tasks such as motion analysis, intra-operative tracking and image segmentation. Popular registration methods such as ANTs and NiftyReg optimize objective functions for each pair of images from scratch, which are time-consuming for 3D and sequential images with complex deformations. Recently, deep learning-based registration approaches such as VoxelMorph have been emerging and achieve competitive performance. In this work, we construct a test-time training for deep deformable image registration to improve the generalization ability of conventional learning-based registration model. We design multi-scale deep networks to consecutively model the residual deformations, which is effective for high variational deformations. Extensive experiments validate the effectiveness of multi-scale deep registration with test-time training based on Dice coefficient for image segmentation and mean square error (MSE), normalized local cross-correlation (NLCC) for tissue dense tracking tasks. Two videos are in https://www.youtube.com/watch?v=NvLrCaqCiAE and https://www.youtube.com/watch?v=pEA6ZmtTNuQ, ICRA 2021; 8 pages, 4 figures, 2 big tables

Mechanical properties of cells are important features that are tightly regulated, and are dictated by various pathologies. Deformability cytometry allows for the characterization of mechanical properties of hundreds of cells per second, opening the way to differentiating cells via mechanotyping. A remaining challenge for detecting and classifying rare sub-populations is the creation of a combined experimental and analysis protocol that would assure classification accuracy approaching 100%. In order to maximize the accuracy, we designed a microfluidic channel that subjects each cell to repeated deformations and relaxations. We also track the shape dynamics of individual cells with high time resolution, and apply sequence-based deep learning models for feature extraction. HL60 cells with and without treatment with cytochalasin D (cytoD), a reagent previously shown to perturb the actin network, were used as a model system to understand the classification potential of our approach. Multiple recurrent and convolutional neural network architectures were trained using time sequences of cell shapes, and shown to achieve high classification accuracy based on cytoskeletal properties alone. The best model classified the two sub-populations of HL60 cells with an accuracy of 95%. This work establishes the application of sequence-based deep learning models to dynamic deformability cytometry.

Background: Compared to commonly-used green space indicators from downward-facing satellite imagery, street view-based green space may capture different types of green space and represent how environments are perceived and experienced by people on the ground, which is important to elucidate the underlying mechanisms linking green space and health. Objectives: This study aimed to evaluate machine learning models that can classify the type of vegetation (i.e., tree, low-lying vegetation, grass) from street view images; and to investigate the associations between street green space and socioeconomic (SES) factors, in Los Angeles County, California. Methods: SES variables were obtained from the CalEnviroScreen3.0 dataset. Microsoft Bing Maps images in conjunction with deep learning were used to measure total and types of street view green space, which were compared to normalized difference vegetation index (NDVI) as commonly-used satellite-based green space measure. Generalized linear mixed model was used to examine associations between green space and census tract SES, adjusting for population density and rural/urban status. Results: The accuracy of the deep learning model was high with 92.5% mean intersection over union. NDVI were moderately correlated with total street view-based green space and tree, and weakly correlated with low-lying vegetation and grass. Total and three types of green space showed significant negative associations with neighborhood SES. The percentage of total green space decreased by 2.62 [95% confidence interval (CI): −3.02, −2.21, p < 0.001] with each interquartile range increase in CalEnviroScreen3.0 score. Disadvantaged communities contained approximately 5% less average street green space than other communities. Conclusion: Street view imagery coupled with deep learning approach can accurately and efficiently measure eye-level street green space and distinguish vegetation types. In Los Angeles County, disadvantaged communities had substantively less street green space. Governments and urban planners need to consider the type and visibility of street green space from pedestrian's perspective.

Dynamic resting state functional connectivity (RSFC) characterizes fluctuations that occurs over time in functional brain networks. Existing methods to extract dynamic RSFCs, such as sliding-window and clustering methods, have various limitations due to their inherent non-adaptive nature and high-dimensionality including an inability to reconstruct brain signals, insufficiency of data for reliable estimation, insensitivity to rapid changes in dynamics, and a lack of generalizability across multimodal functional imaging datasets. To overcome these deficiencies, we develop a novel and unifying time-varying dynamic network (TVDN) framework for examining dynamic resting state functional connectivity. TVDN includes a generative model that describes the relation between low-dimensional dynamic RSFC and the brain signals, and an inference algorithm that automatically and adaptively learns to detect dynamic state transitions in data and a low-dimensional manifold of dynamic RSFC. TVDN is generalizable to handle multimodal functional neuroimaging data (fMRI and MEG/EEG). The resulting estimated low-dimensional dynamic RSFCs manifold directly links to the frequency content of brain signals. Hence we can evaluate TVDN performance by examining whether learnt features can reconstruct observed brain signals. We conduct comprehensive simulations to evaluate TVDN under hypothetical settings. We then demonstrate the application of TVDN with real fMRI and MEG data, and compare the results with existing benchmarks. Results demonstrate that TVDN is able to correctly capture the dynamics of brain activity and more robustly detect brain state switching both in resting state fMRI and MEG data.

Estimating 3D hand pose directly from RGB images is challenging but has gained steady progress recently by training deep models with annotated 3D poses. However annotating 3D poses is difficult and as such only a few 3D hand pose datasets are available, all with limited sample sizes. In this study, we propose a new framework of training 3D pose estimation models from RGB images without using explicit 3D annotations, i.e., trained with only 2D information. Our framework is motivated by two observations: 1) Videos provide richer information for estimating 3D poses as opposed to static images; 2) Estimated 3D poses ought to be consistent whether the videos are viewed in the forward order or reverse order. We leverage these two observations to develop a self-supervised learning model called temporal-aware self-supervised network (TASSN). By enforcing temporal consistency constraints, TASSN learns 3D hand poses and meshes from videos with only 2D keypoint position annotations. Experiments show that our model achieves surprisingly good results, with 3D estimation accuracy on par with the state-of-the-art models trained with 3D annotations, highlighting the benefit of the temporal consistency in constraining 3D prediction models.

Estimating 3D hand poses from a single RGB image is challenging because depth ambiguity leads the problem ill-posed. Training hand pose estimators with 3D hand mesh annotations and multi-view images often results in significant performance gains. However, existing multi-view datasets are relatively small with hand joints annotated by off-the-shelf trackers or automated through model predictions, both of which may be inaccurate and can introduce biases. Collecting a large-scale multi-view 3D hand pose images with accurate mesh and joint annotations is valuable but strenuous. In this paper, we design a spin match algorithm that enables a rigid mesh model matching with any target mesh ground truth. Based on the match algorithm, we propose an efficient pipeline to generate a large-scale multi-view hand mesh (MVHM) dataset with accurate 3D hand mesh and joint labels. We further present a multi-view hand pose estimation approach to verify that training a hand pose estimator with our generated dataset greatly enhances the performance. Experimental results show that our approach achieves the performance of 0.990 in $\text{AUC}_{\text{20-50}}$ on the MHP dataset compared to the previous state-of-the-art of 0.939 on this dataset. Our datasset is public available. \footnote{\url{https://github.com/Kuzphi/MVHM}} Our datasset is available at~\href{https://github.com/Kuzphi/MVHM}{\color{blue}{https://github.com/Kuzphi/MVHM}}.

Ovarian cancer (OC), the third common gynecologic malignancy, contributes to the most cancer-caused mortality in women. However, 70% of patients with OC are diagnosed at an advanced stage, of which the 5-year survival is less than 30%. Long noncoding RNAs (long ncRNAs or lncRNA), a type of RNA with exceeding 200 nucleotides in length but no protein-coding capability, have been demonstrated to involve the pathogenesis of various cancers and show considerable potential in the diagnosis of OC. In this study, we found that the LINC00909 expression in tumor and serum specimens of OC patients was elevated, determined by real-time quantitative, and droplet digital PCR. In receiver operating characteristic (ROC) analysis, our results revealed that serum LINC00909 distinguished cancers from normal ovarian tissue with 87.8% of sensitivity and 69.6% of specificity (AUC, 81.2%) and distinguished serous ovarian cancer from normal ovarian tissue with 90.0% of sensitivity and 75.9% of specificity (AUC, 84.5%). Furthermore, we observed that the tumor and serum LINC00909 level was positively associated with the International Federation of Gynecology and Obstetrics (FIGO) stage and the Eastern Cooperative Oncology Group (ECOG) score (reflecting patients’ performance status). Also, patients with low serum LINC00909 level showed a longer overall (hazard ratio, HR = 1.874 , p = 0.0004 ) and progression-free ( HR = 1.656 , p = 0.0017 ) survival. Functional assays indicated that the elevation of LINC00909 expression contributes to cell proliferation, migration, and invasion capability of ovarian cancer cells. Besides, we demonstrated that LINC00909 functions as a competing endogenous RNA (ceRNA) of MRC2 mRNA by sponging miR-23-3p, and thereby promotes epithelial-to-mesenchymal transition (EMT) of ovarian cancer cells. Therefore, we highlight that the LINC00909/miR-23b-3p/MRC2 axis is implicated in the pathogenesis of ovarian cancer, and serum LINC00909 may be a promising biomarker for the diagnosis of OC.

Characterizing genome-wide binding profiles of transcription factor (TF) is essential for understanding many biological processes. Although techniques have been developed to assess binding profiles within a population of cells, determining binding profiles at a single cell level remains elusive. Here we report scFAN (Single Cell Factor Analysis Network), a deep learning model that predicts genome-wide TF binding profiles in individual cells. scFAN is pre-trained on genome-wide bulk ATAC-seq, DNA sequence and ChIP-seq data, and utilizes single-cell ATAC-seq to predict TF binding in individual cells. We demonstrate the efficacy of scFAN by studying sequence motifs enriched within predicted binding peaks and investigating the effectiveness of predicted TF peaks for discovering cell types. We develop a new metric “TF activity score” to characterize each cell, and show that the activity scores can reliably capture cell identities. The method allows us to discover and study cellular identities and heterogeneity based on chromatin accessibility profiles.

Estimating 3D hand poses from RGB images is essential to a wide range of potential applications, but is challenging owing to substantial ambiguity in the inference of depth information from RGB images. State-of-the-art estimators address this problem by regularizing 3D hand pose estimation models during training to enforce the consistency between the predicted 3D poses and the ground-truth depth maps. However, these estimators rely on both RGB images and the paired depth maps during training. In this study, we propose a conditional generative adversarial network (GAN) model, called Depth-image Guided GAN (DGGAN), to generate realistic depth maps conditioned on the input RGB image, and use the synthesized depth maps to regularize the 3D hand pose estimation model, therefore eliminating the need for ground-truth depth maps. Experimental results on multiple benchmark datasets show that the synthesized depth maps produced by DGGAN are quite effective in regularizing the pose estimation model, yielding new state-of-the-art results in estimation accuracy, notably reducing the mean 3D endpoint errors (EPE) by 4.7%, 16.5%, and 6.8% on the RHD, STB and MHP datasets, respectively.

Objectives: The pandemic of the coronavirus disease 2019 (COVID-19) continuously poses a serious threat to public health, highlighting an urgent need for simple and efficient early detection and prediction. Methods: We comprehensively investigated and reanalyzed the published indexes and models for predicting severe illness among COVID‑19 patients in our dataset, and validated them on an independent dataset. Results: 696 COVID-19 cases in the discovery stage and 337 patients in the validation stage were involved. The AuROC of neutrophil to lymphocyte ratio (NLR) (0.782) was significantly higher than that of the other 11 independent risk indexes in severe outcome prediction. The combination of NLR and oxygen saturation (SaO2) (NLR+SaO2) showed the biggest AuROC calculations with a value of 0.901; with a cut-off value of 0.532, it exhibited 84.2% sensitivity, 88.4% specificity and 86.8% correct classification ratio. Moreover, we first identified that principal component analysis (PCA) is an effective tool to predict the severity of COVID-19. We obtained 86.5% prediction accuracy with 86% sensitivity when PCA was applied to predict severe illness. In addition, to evaluate the performance of NLR+SaO2 and PCA, we compared them with currently published predictive models in the same dataset. Conclusions: It showed that NLR+SaO2 is an appropriate and promising method for predicting severe illness, followed by PCA. We then validated the results on an independent dataset and revealed that they remained robust accuracy in outcome prediction. This study is significant for early treatment, intervention, triage and saving limited resources.

Education informatization in China has been greatly promoted in recent years, along with the popularity of the network technology and the smart terminal, and the breakthrough in big data and cloud computing technology, K12 online education grows rapidly, it’s especially necessary to comprehensively know about the development history of domestic K12 online education, research status and hotspots, thus further explore the development direction of K12 online education. In this paper, Citespace software is used to visualize the journal papers and master's degree papers collected in CNKI database and draw K12 keywords co-occurrence and the research hotspot in the research of the K12 online education change sequential graph, and we found K12 online education research focused on the "basic education", "Internet +", "online learning space", "online school" and "business model", etc., indicating that K12 online education develops with the national education informatization and, and turns the way of Market-oriented Development.

Evaluation metrics play an important role in the batch evaluation of IR systems. Based on a user model that describes how users interact with the rank list, an evaluation metric is defined to link the relevance scores of a list of documents to an estimation of system effectiveness and user satisfaction. Therefore, the validity of an evaluation metric has two facets: whether the underlying user model can accurately predict user behavior and whether the evaluation metric correlates well with user satisfaction. While a tremendous amount of work has been undertaken to design, evaluate, and compare different evaluation metrics, few studies have explored the consistency between these two facets of evaluation metrics. Specifically, we want to investigate whether the metrics that are well calibrated with user behavior data can perform as well in estimating user satisfaction. To shed light on this research question, we compare the performance of various metrics with the C/W/L Framework in estimating user satisfaction when they are optimized to fit observed user behavior. Experimental results on both self-collected and public available user search behavior datasets show that the metrics optimized to fit users' click behavior can perform as well as those calibrated with user satisfaction feedback. We also investigate the reliability in the calibration process of evaluation metrics to find out how much data is required for parameter tuning. Our findings provide empirical support for the consistency between user behavior modeling and satisfaction measurement, as well as guidance for tuning the parameters in evaluation metrics.

Anhedonia is a core characteristic of depression, the amelioration of which accounts for depressive symptom improvement. Electroconvulsive therapy (ECT) has been shown remarkable antidepressive effect, however, less is known about the effect of ECT on anhedonia and its underlying neural mechanism. Herein, we investigated local and global intrinsic brain functional alterations during the resting state in 46 patients with pre- and post-ECT major depressive disorder using the amplitude of low-frequency fluctuations (ALFF) and degree centrality (DC) approach. Functional connectivity (FC) was also calculated between nodes with significant local and global intrinsic brain functional alterations. The severity of anhedonia and depression was assessed with the Temporal Experience of Pleasure Scale and Hamilton Depression Rating Scale, respectively. The relationship between the change in anhedonia and depressive symptoms and brain functional alterations was determined. Increased ALFF and DC were observed in the bilateral dorsal medial prefrontal cortex (dmPFC), right dorsal lateral prefrontal cortex (dlPFC), left orbitofrontal cortex, and right orbitofrontal cortex (ROFC) after ECT. Correlational analysis between the change in anhedonia and ALFF had positive results in the dmPFC. Similarly, there was a positive correlation between the change in anhedonia and change in DC in the dmPFC, right dlPFC, ROFC, and middle frontal gyrus. Furthermore, there was a significant relationship between the change in anhedonia and altered dmPFC-dlPFC FC. These results revealed that amelioration of anhedonia may be associated with intrinsic neural activity alteration in the reward-related circuit and executive control network following ECT.

Background The rapid spread of coronavirus disease 2019 (COVID-19) revealed significant constraints in critical care capacity. In anticipation of subsequent waves, reliable prediction of disease severity is essential for critical care capacity management and may enable earlier targeted interventions to improve patient outcomes. The purpose of this study is to develop and externally validate a prognostic model/clinical tool for predicting COVID-19 critical disease at presentation to medical care. Methods This is a retrospective study of a prognostic model for the prediction of COVID-19 critical disease where critical disease was defined as ICU admission, ventilation, and/or death. The derivation cohort was used to develop a multivariable logistic regression model. Covariates included patient comorbidities, presenting vital signs, and laboratory values. Model performance was assessed on the validation cohort by concordance statistics. The model was developed with consecutive patients with COVID-19 who presented to University of California Irvine Medical Center in Orange County, California. External validation was performed with a random sample of patients with COVID-19 at Emory Healthcare in Atlanta, Georgia. Results Of a total 3208 patients tested in the derivation cohort, 9% (299/3028) were positive for COVID-19. Clinical data including past medical history and presenting laboratory values were available for 29% (87/299) of patients (median age, 48 years [range, 21–88 years]; 64% [36/55] male). The most common comorbidities included obesity (37%, 31/87), hypertension (37%, 32/87), and diabetes (24%, 24/87). Critical disease was present in 24% (21/87). After backward stepwise selection, the following factors were associated with greatest increased risk of critical disease: number of comorbidities, body mass index, respiratory rate, white blood cell count, % lymphocytes, serum creatinine, lactate dehydrogenase, high sensitivity troponin I, ferritin, procalcitonin, and C-reactive protein. Of a total of 40 patients in the validation cohort (median age, 60 years [range, 27–88 years]; 55% [22/40] male), critical disease was present in 65% (26/40). Model discrimination in the validation cohort was high (concordance statistic: 0.94, 95% confidence interval 0.87–1.01). A web-based tool was developed to enable clinicians to input patient data and view likelihood of critical disease. Conclusions and relevance We present a model which accurately predicted COVID-19 critical disease risk using comorbidities and presenting vital signs and laboratory values, on derivation and validation cohorts from two different institutions. If further validated on additional cohorts of patients, this model/clinical tool may provide useful prognostication of critical care needs.

The high irregularity of multiple sclerosis (MS) lesions in sizes and numbers often proves difficult for automated systems on the task of MS lesion segmentation. Current State-of-the-art MS segmentation algorithms employ either only global perspective or just patch-based local perspective segmentation approaches. Although global image segmentation can obtain good segmentation for medium to large lesions, its performance on smaller lesions lags behind. On the other hand, patch-based local segmentation disregards spatial information of the brain. In this work, we propose SynergyNet, a network segmenting MS lesions by fusing data from both global and local perspectives to improve segmentation across different lesion sizes. We achieve global segmentation by leveraging the U-Net architecture and implement the local segmentation by augmenting U-Net with the Mask R-CNN framework. The sharing of lower layers between these two branches benefits end-to-end training and proves advantages over simple ensemble of the two frameworks. We evaluated our method on two separate datasets containing 765 and 21 volumes respectively. Our proposed method can improve 2.55% and 5.0% for Dice score and lesion true positive rates respectively while reducing over 20% in false positive rates in the first dataset, and improve in average 10% and 32% for Dice score and lesion true positive rates in the second dataset. Results suggest that our framework for fusing local and global perspectives is beneficial for segmentation of lesions with heterogeneous sizes.

Hand pose estimation is more challenging than body pose estimation due to severe articulation, self-occlusion and high dexterity of the hand. Current approaches often rely on a popular body pose algorithm, such as the Convolutional Pose Machine (CPM), to learn 2D keypoint features. These algorithms cannot adequately address the unique challenges of hand pose estimation, because they are trained solely based on keypoint positions without seeking to explicitly model structural relationship between them. We propose a novel Nonparametric Structure Regularization Machine (NSRM) for 2D hand pose estimation, adopting a cascade multi-task architecture to learn hand structure and keypoint representations jointly. The structure learning is guided by synthetic hand mask representations, which are directly computed from keypoint positions, and is further strengthened by a novel probabilistic representation of hand limbs and an anatomically inspired composition strategy of mask synthesis. We conduct extensive studies on two public datasets - OneHand 10k and CMU Panoptic Hand. Experimental results demonstrate that explicitly enforcing structure learning consistently improves pose estimation accuracy of CPM baseline models, by 1.17% on the first dataset and 4.01% on the second one. The implementation and experiment code is freely available online. Our proposal of incorporating structural learning to hand pose estimation requires no additional training information, and can be a generic add-on module to other pose estimation models., The paper has be accepted and will be presented at 2020 IEEE Winter Conference on Applications of Computer Vision (WACV). The code is freely available at https://github.com/HowieMa/NSRMhand

Previous studies have shown that the CareStart™ G6PD Deficiency rapid diagnostic test has high diagnostic accuracy on G6PD deficiency in Africa and Thailand, but not in China. As there are regional differences of G6PD genotype distribution, we are attending to verify the effectiveness of the kit in Chinese population. The study cohort included 247 newborns admitted to our hospital for jaundice. The quantitative detection of G6PD enzyme activity and G6PD gene mutations analysis was used to classify the status of G6PD deficiency. The performance of CareStart™ assays was verified by calculating the sensitivity, specificity and area under the receiver operating characteristic curve (AUC) based on the corrected G6PD deficiency status. In male newborns, the sensitivity of the CareStart™ assay was 98.9%, the specificity was 94.2% and the AUC was 0.97. In female newborns, the sensitivity was 58.5% when the cutoff value of residual enzyme activity was 100%; however, the sensitivity was 100% when the cutoff value was 60%. Therefore, the CareStart™ test can effectively screen G6PD deficiency in male newborns and female infants with less than 60% residual enzyme activity, female infants with residual enzyme activities of 60–100% are more likely to be missed diagnosed among Chinese newborns.

In this paper, we propose a new architecture named Rotation-invariant Mixed Graphical Model Network (R-MGMN) to solve the problem of 2D hand pose estimation from a monocular RGB image. By integrating a rotation net, the R-MGMN is invariant to rotations of the hand in the image. It also has a pool of graphical models, from which a combination of graphical models could be selected, conditioning on the input image. Belief propagation is performed on each graphical model separately, generating a set of marginal distributions, which are taken as the confidence maps of hand keypoint positions. Final confidence maps are obtained by aggregating these confidence maps together. We evaluate the R-MGMN on two public hand pose datasets. Experiment results show our model outperforms the state-of-the-art algorithm which is widely used in 2D hand pose estimation by a noticeable margin., 2020 IEEE Winter Conference on Applications of Computer Vision (WACV)

Organs-at-risk (OAR) delineation in computed tomography (CT) is an important step in Radiation Therapy (RT) planning. Recently, deep learning based methods for OAR delineation have been proposed and applied in clinical practice for separate regions of the human body (head and neck, thorax, and abdomen). However, there are few researches regarding the end-to-end whole-body OARs delineation because the existing datasets are mostly partially or incompletely annotated for such task. In this paper, our proposed end-to-end convolutional neural network model, called \textbf{AttentionAnatomy}, can be jointly trained with three partially annotated datasets, segmenting OARs from whole body. Our main contributions are: 1) an attention module implicitly guided by body region label to modulate the segmentation branch output; 2) a prediction re-calibration operation, exploiting prior information of the input images, to handle partial-annotation(HPA) problem; 3) a new hybrid loss function combining batch Dice loss and spatially balanced focal loss to alleviate the organ size imbalance problem. Experimental results of our proposed framework presented significant improvements in both S{\o}rensen-Dice coefficient (DSC) and 95\% Hausdorff distance compared to the baseline model., Comment: accepted by ISBI 2020 (4 pages, 2 figures)

An integrated analysis of human microarray and rat transcriptomic data was performed to illustrate the pathologic effects of cigarette smoking (CS) on the gene expression profile and identify new gene candidates for chronic obstructive pulmonary disease (COPD) progression. Small airway epithelial samples of Human with different smoking status using fiberoptic bronchoscopy and corresponding rat lung tissues following 0, 3, and 6 months of CS exposure were obtained. Differential expression analysis was performed to identify differentially expressed genes (DEGs). Using qRT-PCR, the expression of the significant overlapping genes between human and rats were confirmed in 16HBE cells. 8 relevant studies comprising 293 individuals including phenotypically normal non-smokers, normal smokers, and COPD smokers were involved by searching the public databases. The integrated bioinformatic analysis of 8 human GEO datasets and rat transcriptome databases revealed 13 overlapping genes between humans and rats in response to CS exposure during COPD progression. Of these, 5 genes (AKR1C3, ERP27, AHRR, KCNMB2, and MRC1) were consistently identified in both the human and rat and validated by qRT-PCR. Among them, ERP27, KCNMB2, and MRC1 were newly identified candidate genes for COPD progression in response to CS. In addition, we also found that DEGs obtained from each expression profile dataset was better than combined analysis as more genes could be identified. The findings of the present study illustrated that integration of human microarray and rat transcriptome data might facilitate the discovery of new DEGs candidate of COPD progression in response to smoking. Funding Statement: This study was supported by grants from the National Key R&D Program of China (2016YFC0903700), National Natural Science Foundation of China (81520108001, 81770043 and 81700043), Guangzhou Department of Education (1201620007), Local Innovative and Research Teams Project of Guangdong Pearl River Talents Program (2017BT01S155), the National Key R&D Program of China (2018YFC1311900), the Guangdong Natural Science Foundation (2020A1515010076), and the Research Projects of SKLRD (OP-201808, QN-201706, QN-201917, OP-201912). Declaration of Interests: The authors have no conflict of interest to declare. Ethics Approval Statement: The study was conducted according to the criteria set by the declaration of Helsinki, and written informed consent was obtained from all participants before data collection. The study was approved by the Institutional Review Board (GZMC 2009-08-1336).

Multi-organ segmentation is one of most successful applications of deep learning in medical image analysis. Deep convolutional neural nets (CNNs) have shown great promise in achieving clinically applicable image segmentation performance on CT or MRI images. State-of-the-art CNN segmentation models apply either 2D or 3D convolutions on input images, with pros and cons associated with each method: 2D convolution is fast, less memory-intensive but inadequate for extracting 3D contextual information from volumetric images, while the opposite is true for 3D convolution. To fit a 3D CNN model on CT or MRI images on commodity GPUs, one usually has to either downsample input images or use cropped local regions as inputs, which limits the utility of 3D models for multi-organ segmentation. In this work, we propose a new framework for combining 3D and 2D models, in which the segmentation is realized through high-resolution 2D convolutions, but guided by spatial contextual information extracted from a low-resolution 3D model. We implement a self-attention mechanism to control which 3D features should be used to guide 2D segmentation. Our model is light on memory usage but fully equipped to take 3D contextual information into account. Experiments on multiple organ segmentation datasets demonstrate that by taking advantage of both 2D and 3D models, our method is consistently outperforms existing 2D and 3D models in organ segmentation accuracy, while being able to directly take raw whole-volume image data as inputs., Comment: Accepted WACV 2021

BACKGROUND The colon plays a vital role in the disposal of nitrogenous waste products. Therefore, the colon may provide a therapeutic target for managing chronic kidney disease (CKD). AIM To evaluate the efficacy of a simplified model of colonic dialysis with hemodialysis solutions (SCD) to delay the progression of stages 3-5 CKD. DESIGN Retrospective study. METHODS We retrospectively analyzed 178 stages 3-5 CKD patients who did or did not receive SCD (SCD group, n = 88; control group, n = 90). The follow-up was 36 months. The outcome of CKD progression was defined as a decrease in 50% or more in estimated glomerular filtration rate, starting hemodialysis or peritoneal dialysis or undergoing transplantation. The Kaplan-Meier analysis was used to compare CKD progression between SCD and control groups as well as between subgroups at different CKD stages. Cox proportional hazard models adjusted for patients' characteristics were used to examine the association between SCD and the outcome. RESULTS For all patients, the outcome was significantly better in SCD group compared to control group (P < 0.05). The results were similar in the subgroups of patients at stage 4 (P = 0.001) and stage 5 (P = 0.000), but not in the subgroup of patients at stage 3 (P = 0.121). For all patients, SCD was associated with a lower risk of CKD progression after adjusted for patients' characteristics (adjusted hazard ratio, 0.373; 95% confidence interval, 0.201-0.694; P 0.002). CONCLUSION SCD is an effective supplementary therapy to delay the progression of stages 4-5 CKD.

In recent years, the rapid development of urbanization has posed enormous challenges to transportation, security management, quality of life, and so on, which makes the research and development of smart city important. As an information-driven project, strong communication infrastructures are required for connecting smart objects, people, and sensors together. As a consequence, the optimization of the wireless network is the primary premise to support and improve the quality of smart services. The challenge lies in the difficulty to achieve the quantitative description of networks due to the complexity and variability of wireless environments. It is too coarse-grained to express characteristics of networks simply by the strength of received signals through mobile devices. To extract more fine-grained network characteristics, we dig into the PHY layer collecting CSI for network descriptions and extract three signal characteristics, i.e., Rician-K, delay spread and spectral width, from real-world wireless channels. The K-means clustering algorithm is implemented in this article for good performance of network partitioning, and experiments are conducted based on the data sets collected from real scenarios. Simulation results verify the feasibility of our scheme.

Smart healthcare, instead of traditional healthcare, has attracted tremendous attention all over the world, and calls for integrated design bringing together interdisciplinary technological approaches and solutions with the aim of supporting affordable and high-quality patient care. With the increasing number of visually impaired people, including the blind, the elderly, and patients with eye diseases, it is vital to help them explore their outside environment with the assistance of smart technologies so that they can adapt to the environment more easily. The smart indoor anti-collision system is one typical application. Previous works normally use video monitors or deployed sensors to provide outside environment information for visually impaired people. However, these methods suffer from privacy disclosure and inconvenience. In this article, we present IAAS, a smart indoor anti-collision system based on RFID, which identifies and tracks passive RFID tags by analyzing the received backscatter signals. We extract RSSI based on the LWLR algorithm and phase profiles as fingerprints to help the user guide from obstacles without observations of eyes. Experiments are conducted to verify our system, and results show that IAAS could achieve high accuracy of 94 percent in obstacle avoidance.

MotivationFor many RNA molecules, the secondary structure is essential for the correct function of the RNA. Predicting RNA secondary structure from nucleotide sequences is a long-standing problem in genomics, but the prediction performance has reached a plateau over time. Traditional RNA secondary structure prediction algorithms are primarily based on thermodynamic models through free energy minimization, which imposes strong prior assumptions and is slow to run.ResultsHere we propose a deep learning-based method, called UFold, for RNA secondary structure prediction, trained directly on annotated data without any thermodynamic assumptions. UFold improves substantially upon previous models, with approximately 10~30% improvement over traditional thermodynamic models and 14% improvement over other learning-based methods. It achieves an F1 score of 0.91 on base pair prediction accuracy on an RNA structure prediction benchmark dataset. UFold is also fast with an inference time about 160ms per sequence up to 1600bp length. We provide an online web server that implements UFold for RNA structure prediction and is made freely available.AvailabilityAn online web server running UFold is available at https://ufold.ics.uci.edu. Code is available at https://github.com/uci-cbcl/UFold.Contactxhx@uci.edu

Many biological processes are governed by protein-ligand interactions. One such example is the recognition of self and nonself cells by the immune system. This immune response process is regulated by the major histocompatibility complex (MHC) protein which is encoded by the human leukocyte antigen (HLA) complex. Understanding the binding potential between MHC and peptides can lead to the design of more potent, peptide-based vaccines and immunotherapies for infectious autoimmune diseases. We apply machine learning techniques from the natural language processing (NLP) domain to address the task of MHC-peptide binding prediction. More specifically, we introduce a new distributed representation of amino acids, name HLA-Vec, that can be used for a variety of downstream proteomic machine learning tasks. We then propose a deep convolutional neural network architecture, name HLA-CNN, for the task of HLA class I-peptide binding prediction. Experimental results show combining the new distributed representation with our HLA-CNN architecture achieves state-of-the-art results in the majority of the latest two Immune Epitope Database (IEDB) weekly automated benchmark datasets. We further apply our model to predict binding on the human genome and identify 15 genes with potential for self binding., Comment: 12 pages, 4 figures, 4 tables

Background: Improvements to prognostic models in metastatic castration-resistant prostate cancer have the potential to augment clinical trial design and guide treatment strategies. In partnership with Project Data Sphere, a not-for-profit initiative allowing data from cancer clinical trials to be shared broadly with researchers, we designed an open-data, crowdsourced, DREAM (Dialogue for Reverse Engineering Assessments and Methods) challenge to not only identify a better prognostic model for prediction of survival in patients with metastatic castration-resistant prostate cancer but also engage a community of international data scientists to study this disease. Methods Data from the comparator arms of four phase 3 clinical trials in first-line metastatic castration-resistant prostate cancer were obtained from Project Data Sphere, comprising 476 patients treated with docetaxel and prednisone from the ASCENT2 trial, 526 patients treated with docetaxel, prednisone, and placebo in the MAINSAIL trial, 598 patients treated with docetaxel, prednisone or prednisolone, and placebo in the VENICE trial, and 470 patients treated with docetaxel and placebo in the ENTHUSE 33 trial. Datasets consisting of more than 150 clinical variables were curated centrally, including demographics, laboratory values, medical history, lesion sites, and previous treatments. Data from ASCENT2, MAINSAIL, and VENICE were released publicly to be used as training data to predict the outcome of interestnamely, overall survival. Clinical data were also released for ENTHUSE 33, but data for outcome variables (overall survival and event status) were hidden from the challenge participants so that ENTHUSE 33 could be used for independent validation. Methods were evaluated using the integrated time-dependent area under the curve (iAUC). The reference model, based on eight clinical variables and a penalised Cox proportional-hazards model, was used to compare method performance. Further validation was done using data from a fifth trialENTHUSE M1in which 266 patients with metastatic castration-resistant prostate cancer were treated with placebo alone. Findings 50 independent methods were developed to predict overall survival and were evaluated through the DREAM challenge. The top performer was based on an ensemble of penalised Cox regression models (ePCR), which uniquely identified predictive interaction effects with immune biomarkers and markers of hepatic and renal function. Overall, ePCR outperformed all other methods (iAUC 0·791; Bayes factor >5) and surpassed the reference model (iAUC 0·743; Bayes factor >20). Both the ePCR model and reference models stratified patients in the ENTHUSE 33 trial into high-risk and low-risk groups with significantly different overall survival (ePCR: hazard ratio 3·32, 95% CI 2·394·62, p, European Union within the ERC grant LatentCauses supported the work of C.F and I.K. German Research Foundation (DFG) within the Collaborative Research Centre 1243, subproject A17 awarded to C.F. German Federal Ministry of Education and Research (BMBF) through the Research Consortium e:AtheroMED (Systems medicine of myocardial infarction and stroke) under the auspices of the e:Med Programme (grant # 01ZX1313C) supported the work of D.P.A., P.D., C.F., C.K., I.K., N.K., M.L., H.S. and J.F.S. at the Institute of Computational Biology. NIH Grants RR025747-01, MH086633 and 1UL1TR001111, and NSF Grants SES-1357666, DMS-14-07655 and BCS0826844 supported the work of C.H., J.I., E.L., Y.W., H.Y., H.Z. and J.Z. NSFC Grant Nos. 61332013, 61572139 supported the work of X.L, Y.L, Y.Z., and S.Z. National Natural Science Foundation of China grants [Nos. 61422309, 61379092] was awarded to S.Z. The Patrick C. Walsh Prostate Research Fund and the Johns Hopkins Individualized Health Initiative supported the work of R.Y.C., D.D., Y.D., Z.J., K.R., Z.W. and Y.Z. FCT Ph.D. Grant SFRH/BD/80925/2011 was awarded to S.C. Clinical Persona Inc., East Palo Alto, CA supported the work of L.B. and D.K. The Finnish Cultural Foundation and the Drug Research Doctoral Programme (DRDP) at the University of Turku supported T.D.L. The National Research Foundation Singapore and the Singapore Ministry of Education, under its Research Centres of Excellence initiative, supported the work of J.G. and K.T. A grant from the Russian Science Foundation 14-24-00155 was awarded to M.D.K. A*MIDEX grant (no. ANR-11-IDEX-0001-02) was awarded to P.J.B. NSERC supported the work of R.G. The Israeli Centers of Research Excellence (I-CORE) program (Center No. 4/11) supported the work of E.G. Academy of Finland (grants 292611, 269862, 272437, 279163, 295504), National Cancer Institute (16X064), and Cancer Society of Finland supported the work of T.A. Academy of Finland (grant 268531) supported the work of T.M

Image-based virtual try-on systems with the goal of transferring a desired clothing item onto the corresponding region of a person have made great strides recently, but challenges remain in generating realistic looking images that preserve both body and clothing details. Here we present a new virtual try-on network, called VTNFP, to synthesize photo-realistic images given the images of a clothed person and a target clothing item. In order to better preserve clothing and body features, VTNFP follows a three-stage design strategy. First, it transforms the target clothing into a warped form compatible with the pose of the given person. Next, it predicts a body segmentation map of the person wearing the target clothing, delineating body parts as well as clothing regions. Finally, the warped clothing, body segmentation map and given person image are fused together for fine-scale image synthesis. A key innovation of VTNFP is the body segmentation map prediction module, which provides critical information to guide image synthesis in regions where body parts and clothing intersects, and is very beneficial for preventing blurry pictures and preserving clothing and body part details. Experiments on a fashion dataset demonstrate that VTNFP generates substantially better results than state-of-the-art methods.

Social deficits are features of autism and highly heritable traits. A common variant in autism-related CNTNAP2 gene, rs2710102, has been linked with social performance, but the neural substrates are largely unknown. We investigated variations in social performance and functional connectivity (static and dynamic) in the subregions of right temporoparietal junction (RTPJ), a key node of brain social network, using resting-state magnetic resonance imaging (n = 399) by genotype at rs2710102 in healthy volunteers. Social performance was evaluated using the social domain of the Autism-Spectrum Quotient (AQ-social; n = 641) and fixation time on eye areas during an eye-tracking task (n = 32). According to previous evidence that the A-allele is the risk allele for social dysfunction, we classified participants into GG and A-allele carriers (AA/AG) groups. The A-allele carriers showed poor social performance (high AQ-social and short fixation time on eye areas) compared with the GG carriers. In the A-allele carriers, decreased stationary functional connectivity between the orbitofrontal cortex and posterior RTPJ (pRTPJ), and decreased dynamic functional connectivity (dFC) between the medial prefrontal cortex (mPFC) and pRTPJ were observed. The fixation time at eye areas positively were correlated with the pRTPJ-mPFC dFC. These findings provided insight for genetic effect on social behavior and its potential neural substrate.

Echocardiography has become routinely used in the diagnosis of cardiomyopathy and abnormal cardiac blood flow. However, manually measuring myocardial motion and cardiac blood flow from echocardiogram is time-consuming and error-prone. Computer algorithms that can automatically track and quantify myocardial motion and cardiac blood flow are highly sought after, but have not been very successful due to noise and high variability of echocardiography. In this work, we propose a neural multi-scale self-supervised registration (NMSR) method for automated myocardial and cardiac blood flow dense tracking. NMSR incorporates two novel components: 1) utilizing a deep neural net to parameterize the velocity field between two image frames, and 2) optimizing the parameters of the neural net in a sequential multi-scale fashion to account for large variations within the velocity field. Experiments demonstrate that NMSR yields significantly better registration accuracy than state-of-the-art methods, such as advanced normalization tools (ANTs) and VoxelMorph, for both myocardial and cardiac blood flow dense tracking. Our approach promises to provide a fully automated method for fast and accurate analyses of echocardiograms., Comment: Blood tracking video: https://youtu.be/pEA6ZmtTNuQ Muscle tracking video: https://youtu.be/NvLrCaqCiAE

Pulmonary lobe segmentation is an important task for pulmonary disease related Computer Aided Diagnosis systems (CADs). Classical methods for lobe segmentation rely on successful detection of fissures and other anatomical information such as the location of blood vessels and airways. With the success of deep learning in recent years, Deep Convolutional Neural Network (DCNN) has been widely applied to analyze medical images like Computed Tomography (CT) and Magnetic Resonance Imaging (MRI), which, however, requires a large number of ground truth annotations. In this work, we release our manually labeled 50 CT scans which are randomly chosen from the LUNA16 dataset and explore the use of deep learning on this task. We propose pre-processing CT image by cropping region that is covered by the convex hull of the lungs in order to mitigate the influence of noise from outside the lungs. Moreover, we design a hybrid loss function with dice loss to tackle extreme class imbalance issue and focal loss to force model to focus on voxels that are hard to be discriminated. To validate the robustness and performance of our proposed framework trained with a small number of training examples, we further tested our model on CT scans from an independent dataset. Experimental results show the robustness of the proposed approach, which consistently improves performance across different datasets by a maximum of $5.87\%$ as compared to a baseline model., 2019 IEEE 16th International Symposium on Biomedical Imaging (ISBI 2019)

Early detection of pulmonary nodules in computed tomography (CT) images is essential for successful outcomes among lung cancer patients. Much attention has been given to deep convolutional neural network (DCNN)-based approaches to this task, but models have relied at least partly on 2D or 2.5D components for inherently 3D data. In this paper, we introduce a novel DCNN approach, consisting of two stages, that is fully three-dimensional end-to-end and utilizes the state-of-the-art in object detection. First, nodule candidates are identified with a U-Net-inspired 3D Faster R-CNN trained using online hard negative mining. Second, false positive reduction is performed by 3D DCNN classifiers trained on difficult examples produced during candidate screening. Finally, we introduce a method to ensemble models from both stages via consensus to give the final predictions. By using this framework, we ranked first of 2887 teams in Season One of Alibaba's 2017 TianChi AI Competition for Healthcare., 2018 IEEE 15th International Symposium on Biomedical Imaging (ISBI 2018)

Herein, a magnetic graphene field-effect transistor biosensor was prepared through the transfer of a chemical vapor deposition graphene film onto a glass substrate to produce a sensing film and conductive channel. By fixing 1-pyrenebutanoic acid succinimidyl ester onto graphene film as an anchor, a probe aptamer was immobilized on the graphene film in order to capture magnetically labeled complementary single-stranded DNA. Our experiments showed that, within a periodic magnetic field, the biosensor impedance exhibited a periodic oscillation, the amplitude of which was correlated to the complementary DNA concentration. Based on this principle, the magnetic graphene field-effect transistor was utilized to detect single-stranded DNA with detection limition of 1 pM. The results were rationalized using a model wherein the magnetic force causes the DNA strand to bend, thereby resulting in magnetic nanobeads/DNA modulation of the double conductive layer of graphene transistors. Furthermore, since a periodic magnetic field could be introduced to produce a periodic impedance changes of MGFETs, sampling integration could be used to improve the signal-to-noise ratio efficiently by increasing the number of periods of the external magnetic field. Therefore, a novel biosensor for DNA detection with high sensitivity has been presented in this work. Based on the detection principle, this system may also be a potential tool for detecting other bio-molecules, cells, etc.