One-Stage Anchor-Free Online Multiple Target Tracking with Deformable Local Attention and Task-Aware Prediction.

The tracking-by-detection paradigm currently dominates multiple target tracking algorithms. It usually includes three tasks: target detection, appearance feature embedding, and data association. Carrying out these three tasks successively usually leads to lower tracking efficiency. In this paper, we propose a one-stage anchor-free multiple task learning framework which carries out target detection and appearance feature embedding in parallel to substantially increase the tracking speed. This framework simultaneously predicts a target detection and produces a feature embedding for each location, by sharing a pyramid of feature maps. We propose a deformable local attention module which utilizes the correlations between features at different locations within a target to obtain more discriminative features. We further propose a task-aware prediction module which utilizes deformable convolutions to select the most suitable locations for the different tasks. At the selected locations, classification of samples into foreground or background, appearance feature embedding, and target box regression are carried out. Two effective training strategies, regression range overlapping and sample reweighting, are proposed to reduce missed detections in dense scenes. Ambiguous samples whose identities are difficult to determine are effectively dealt with to obtain more accurate feature embedding of target appearance. An appearance-enhanced non-maximum suppression is proposed to reduce over-suppression of true targets in crowded scenes. Based on the one-stage anchor-free network with the deformable local attention module and the task-aware prediction module, we implement a new online multiple target tracker. Experimental results show that our tracker achieves a very fast speed while maintaining a high tracking accuracy.

Population-based prevalence of self-reported pediatric diabetes and screening for undiagnosed type 2 diabetes in Chinese children in years 2017-2019, a cross-sectional study.

Background: The worldwide geographical and temporal variation in the prevalence of diabetes represents a challenge, but also an opportunity for gaining etiological insights. Encompassing the bulk of East Asians, a large and distinct proportion of the world population, China can be a source of valuable epidemiological insights for diabetes, especially in early life, when pathophysiology begins. We carried out a nationwide, epidemiological survey of Prevalence and Risk of Obesity and Diabetes in Youth (PRODY) in China, from 2017 to 2019, to estimate the population-based prevalence of diagnosed pediatric diabetes and screen for undiagnosed pediatric type 2 diabetes (T2D). Methods: PRODY was a nation-wide, school population-based, cross-sectional, multicenter survey by questionnaire, fasting urine glucose test and simple oral glucose tolerance test (s-OGTT), among a total number of 193,801 general-population children and adolescents (covered a pediatric population of more than 96.8 million), aged 3-18, from twelve provinces across China. The prevalence of the self-reported pediatric diabetes, the proportion of subtypes, the crude prevalence of undiagnosed T2D and prediabetes in general juvenile population and the main risk factors of type 1 (T1D) and type 2 (T2D) diabetes had been analyzed in the study. Findings: The prevalence of all self-reported pediatric diabetes was estimated at 0.62/1000 (95% CI: 0.51-0.74), with T1D at 0.44/1000 (95% CI: 0.35-0.54) and T2D at 0.18/1000 (95% CI: 0.13-0.25). For undiagnosed T2D, the crude prevalence was almost ten-fold higher, at 1.59/1000, with an estimated extra 28.45/1000 of undiagnosed impaired glucose tolerance (IGT) and 53.74/1000 of undiagnosed impaired fasting glucose (IFG) by s-OGTT screening. Maternal diabetes history is the major risk factors for all subtypes of pediatric diabetes in China. Interpretation: The PRODY study provides the first population-based estimate of the prevalence of pediatric diabetes China and reveals a magnitude of the problem of undiagnosed pediatric T2D. We propose a practical screening strategy by s-OGTT to address this serious gap. Funding: The National Key Research and Development Programme of China, Key R&D Program of Zhejiang, the National Natural Science Foundation of China and the Zhejiang Provincial Key Disciplines of Medicine, Key R&D Program Projects in Zhejiang Province.

Trends and prospects in nursing care for diabetic foot: A bibliometric analysis from 2003 to 2023.

BACKGROUNDS: Diabetic foot (DF) is a globally significant concern, with complications like diabetic foot ulcers (DFUs) posing major challenges despite medical advancements. Effective nursing strategies are crucial to preventing DF progression and reducing disability risk. However, nursing research in DF care is fragmented, necessitating a comprehensive bibliometric analysis to identify key trends, influential contributors, and critical research areas. PURPOSE: This study explored current trends in nursing methods for DF care and their impact on patient outcomes, utilizing CiteSpace, VOSviewer, and Bibliometrix to identify key contributors, influential countries, and noteworthy topics, aiming to provide valuable insights for healthcare professionals and researchers in the field. METHODS: Relevant publications from the Web of Science (WOS) Core Collection Science Citation Index Expanded were retrieved for the period between 2003 and 2023. We included peer-reviewed original articles or reviews related to diabetic foot (DF) and nursing. The following criteria were used for exclusion: ① conference abstracts or corrigendum documents, ② unpublished articles, ③ repeated publications, ④ unrelated articles, ⑤ case reports, and ⑥ qualitative studies. CiteSpace was employed to identify top authors, institutions, countries, keywords, co-cited authors, journals, references, and research trends. VOSviewer was used to generate a network of authors, journals, and references. Bibliometrix was utilized to create maps of cooperating countries and keyword frequency charts, as well as a Sankey diagram illustrating the relationship between authors, keywords, and countries. RESULTS: A total of 305 relevant articles were included in this study. The research pertaining to nursing aspects of diabetic foot care exhibited a noticeable upward trend. The analysis in this study revealed that "amputation" held the highest centrality, indicating a critical area of focus in nursing interventions to prevent severe outcomes. "Diabetic foot ulcer" ranked first in terms of citation rate, emphasizing the ongoing challenges in managing DFUs through nursing care. In recent years, there was a shift in focus towards keywords such as "pressure ulcers", "burden", and "chronic wound" highlighting the evolving priorities in nursing research to address complex wound care, patient burden, and long-term management strategies. CONCLUSIONS: The current primary research focuses in nursing care for diabetic foot (DF) include wound management, offloading techniques, sensory protection, anti-infective treatment, education and self-management, and multidisciplinary teamwork. Future research should prioritize developing innovative nursing interventions tailored to individual patient needs, integrating advanced technologies like telemedicine and wearable devices for continuous monitoring, and exploring the psychological aspects of DFU management to improve patient adherence and outcomes. Additionally, more longitudinal studies are needed to assess the long-term effectiveness of various nursing strategies on patient quality of life.

Synthesis of P-(NiCo)CO3 /TiO2 /Ti Self-Supported Electrode with High Catalytic Activity and Stability for Hydrogen Evolution.

Hydrogen is considered an ideal clean energy due to its high mass-energy density, and only water is generated after combustion. Water electrolysis is a sustainable method of obtaining a usable amount of pure hydrogen among the various hydrogen production methods. However, its development is still limited by applying expensive noble metal catalysts. Here, the dissolution-recrystallization process of TiO2 nanotube arrays in water with the hydrothermal reaction of a typical nickel-cobalt hydroxide synthesis process followed by phosphating to prepare a self-supported electrode with (NiCo)CO3 /TiO2 heterostructure named P-(NiCo)CO3 /TiO2 /Ti electrode is combined. The electrode exhibits an ultra-low overpotential of 31 mV at 10 mA  cm-2 with a Tafel slope of 46.2 mV dec-1 in 1 m KOH and maintained its stability after running for 500 h in 1 m KOH. The excellent catalytic activity can be attributed to the structure of nanotube arrays with high specific surface area, superhydrophilicity, and super aerophobicity on the electrode surface. In addition, the uniform (NiCo)CO3 /TiO2 heterostructure also accelerates the electron transfer on the electrode surface. Finally, DFT calculations demonstrate that phosphating also improves the ΔGH* and ΔGH2O of the electrode. The synthesis strategy also promotes the exploration of catalysts for other necessary electrocatalytic fields.

The LMIT: Light-mediated minimally-invasive theranostics in oncology.

Minimally-invasive diagnosis and therapy have gradually become the trend and research hotspot of current medical applications. The integration of intraoperative diagnosis and treatment is a development important direction for real-time detection, minimally-invasive diagnosis and therapy to reduce mortality and improve the quality of life of patients, so called minimally-invasive theranostics (MIT). Light is an important theranostic tool for the treatment of cancerous tissues. Light-mediated minimally-invasive theranostics (LMIT) is a novel evolutionary technology that integrates diagnosis and therapeutics for the less invasive treatment of diseased tissues. Intelligent theranostics would promote precision surgery based on the optical characterization of cancerous tissues. Furthermore, MIT also requires the assistance of smart medical devices or robots. And, optical multimodality lay a solid foundation for intelligent MIT. In this review, we summarize the important state-of-the-arts of optical MIT or LMIT in oncology. Multimodal optical image-guided intelligent treatment is another focus. Intraoperative imaging and real-time analysis-guided optical treatment are also systemically discussed. Finally, the potential challenges and future perspectives of intelligent optical MIT are discussed.

Micro-Nano Water Film Enabled High-Performance Interfacial Solar Evaporation.

Interfacial solar evaporation holds great promise to address the freshwater shortage. However, most interfacial solar evaporators are always filled with water throughout the evaporation process, thus bringing unavoidable heat loss. Herein, we propose a novel interfacial evaporation structure based on the micro-nano water film, which demonstrates significantly improved evaporation performance, as experimentally verified by polypyrrole- and polydopamine-coated polydimethylsiloxane sponge. The 2D evaporator based on the as-prepared sponge realizes an enhanced evaporation rate of 2.18 kg m-2 h-1 under 1 sun by fine-tuning the interfacial micro-nano water film. Then, a homemade device with an enhanced condensation function is engineered for outdoor clean water production. Throughout a continuous test for 40 days, this device demonstrates a high water production rate (WPR) of 15.9-19.4 kg kW-1 h-1 m-2. Based on the outdoor outcomes, we further establish a multi-objective model to assess the global WPR. It is predicted that a 1 m2 device can produce at most 7.8 kg of clean water per day, which could meet the daily drinking water needs of 3 people. Finally, this technology could greatly alleviate the current water and energy crisis through further large-scale applications.

Ranking-Based Color Constancy With Limited Training Samples.

Computational color constancy is an important component of Image Signal Processors (ISP) for white balancing in many imaging devices. Recently, deep convolutional neural networks (CNN) have been introduced for color constancy. They achieve prominent performance improvements comparing with those statistics or shallow learning-based methods. However, the need for a large number of training samples, a high computational cost and a huge model size make CNN-based methods unsuitable for deployment on low-resource ISPs for real-time applications. In order to overcome these limitations and to achieve comparable performance to CNN-based methods, an efficient method is defined for selecting the optimal simple statistics-based method (SM) for each image. To this end, we propose a novel ranking-based color constancy method (RCC) that formulates the selection of the optimal SM method as a label ranking problem. RCC designs a specific ranking loss function, and uses a low rank constraint to control the model complexity and a grouped sparse constraint for feature selection. Finally, we apply the RCC model to predict the order of the candidate SM methods for a test image, and then estimate its illumination using the predicted optimal SM method (or fusing the results estimated by the top k SM methods). Comprehensive experiment results show that the proposed RCC outperforms nearly all the shallow learning-based methods and achieves comparable performance to (sometimes even better performance than) deep CNN-based methods with only 1/2000 of the model size and training time. RCC also shows good robustness to limited training samples and good generalization crossing cameras. Furthermore, to remove the dependence on the ground truth illumination, we extend RCC to obtain a novel ranking-based method without ground truth illumination (RCC_NO) that learns the ranking model using simple partial binary preference annotations provided by untrained annotators rather than experts. RCC_NO also achieves better performance than the SM methods and most shallow learning-based methods with low costs of sample collection and illumination measurement.

The Kinematics and Dynamics of Schizopygopsis malacanthus Swimming during Ucrit Testing.

The swimming kinematics (how fish move) and dynamics (how forces effect movement) of Schizopygopsis malacanthus were investigated during the determination of Ucrit by stepped velocity testing. A video tracking program was used to record and analyze the motion of five test fish in a Brett-type flume during each velocity step. The findings fell into three groups: (1) Even when flow was uniform, fish did not swim steadily, with speeds fluctuating by 2.2% to 8.4% during steady swimming. The proportion of unsteady swimming time increased with water velocity, and defining steady and unsteady swimming statistically, in terms of the definition of standard deviation of instantaneous displacements, may have higher accuracy. (2) In steady swimming, the forward velocity and acceleration of fish were correlated with body length (p < 0.05), but in unsteady swimming the correlations were not significant. The maximum swimming speed (1.504 m/s) and acceleration (16.54 m/s2) occurred during unsteady swimming, but these measurements may not be definitive because of tank space constraints on fish movement and the passive behavior of the test fish with respect to acceleration. (3) Burst-coast swimming in still water, investigated by previous scholars as an energy conserving behavior, is not the same as the gait transition from steady to unsteady swimming in flowing water. In this study, the axial force of fish swimming in the unsteady mode was significantly higher (×1.2~1.6) than in the steady mode, as was the energy consumed (×1.27~3.33). Thus, gait transition increases, rather than decreases, energy consumption. Our characterization of the kinematics and dynamics of fish swimming provides important new information to consider when indices of swimming ability from controlled tank testing are applied to fish passage design.

Self-supervised speckle noise reduction of optical coherence tomography without clean data.

Optical coherence tomography (OCT) is widely used in clinical diagnosis due to its non-invasive, real-time, and high-resolution characteristics. However, the inherent speckle noise seriously degrades the image quality, which might damage the fine structures in OCT, thus affecting the diagnosis results. In recent years, supervised deep learning-based denoising methods have shown excellent denoising ability. To train a deep denoiser, a large number of paired noisy-clean images are required, which is difficult to achieve in clinical practice, since acquiring a speckle-free OCT image requires dozens of repeated scans and image registration. In this research, we propose a self-supervised strategy that helps build a despeckling model by training it to map neighboring pixels in a single noisy OCT image. Adjacent pixel patches are randomly selected from the original OCT image to generate two similar undersampled images, which are respectively used as the input and target images for training a deep neural network. To ensure both the despeckling and the structure-preserving effects, a multi-scale pixel patch sampler and corresponding loss functions are adopted in our practice. Through quantitative evaluation and qualitative visual comparison, we found that the proposed method performs better than state-of-the-art methods regarding despeckling effects and structure preservation. Besides, the proposed method is much easier to train and deploy without the need for clean OCT images, which has great significance in clinical practice.

Prevalence and phenotypic features of diabetes due to recessive, non-syndromic WFS1 mutations.

OBJECTIVE: Recessive WFS1 mutations are known to cause Wolfram syndrome, a very rare systemic disorder. However, they were also found in non-syndromic diabetes in Han Chinese misdiagnosed with type 1 diabetes (T1D), a molecular cause that appears to be considerably more common than the fully expressed syndrome. We aimed to better define the incidence and clinical features of non-syndromic diabetes due to recessive WFS1 mutation. DESIGN: We analyzed the genotype and phenotype of 320 consecutive incident Chinese pediatric diabetic patients diagnosed from 2016 to 2019 to search for non-syndromic diabetic cases due to recessive WFS1 mutation. METHODS: A cohort of 105 pancreatic autoantibody-negative patients were recruited for exome sequencing. All patients tested positive for pathogenic diallelic WFS1 mutations were examined for phenotypic features (fundoscopy, audiogram, and urine density). RESULTS: We found three cases of non-syndromic diabetes due to recessive WFS1 mutations (incidence = 0.94% (95% CI: 0.25-2.7%)). All three cases only had mild diabetes when diagnosed. All patients had well-conserved fasting C-peptide when diagnosed but one of them progressed to T1D-like insulin deficiency. In addition, we found a fourth case with previously undetected features of Wolfram syndrome. CONCLUSIONS: Non-syndromic diabetes due to WFS1 mutation may be common among Chinese pediatric patients with diabetes. It is important to differentiate it from other maturity-onset diabetes in the young subtypes with similar phenotype by molecular diagnosis because of different prognosis and, potentially, therapy.

Robust Texture-Aware Computer-Generated Image Forensic: Benchmark and Algorithm.

With advances in rendering techniques and generative adversarial networks, computer-generated (CG) images tend to be indistinguishable from photographic (PG) images. Revisiting previous works towards CG image forensic, we observed that existing datasets are constructed years ago and limited in both quantity and diversity. Besides, current algorithms only consider the global visual features for forensic, ignoring finer differences between CG and PG images. To mitigate these problems, we first contribute a Large-Scale CG images Benchmark (LSCGB), and then propose a simple yet strong baseline model to address the forensic task. On the one hand, the introduced benchmark has three superior properties, 1) large-scale: the benchmark contains 71168 CG and 71168 PG images with the corresponding expert-annotated labels. It is orders of magnitude bigger than previous datasets. 2) high diversity: we collect CG images from 4 different scenes generated by various rendering techniques. The PG images are varied in terms of image content, camera models, and photographer styles. 3) small bias: we carefully filter the collected images to ensure that the distributions of color, brightness, tone and saturation between CG and PG images are close. Furthermore, inspired by an empirical study on texture difference between CG and PG images, an effective texture-aware network is proposed to improve forensic accuracy. Concretely, we first strengthen texture information of multilevel features extracted from a backbone. Then, the relations among feature channels are explored by learning its gram matrix. Each feature channel represents a specific texture pattern. The gram matrix is thus able to embed the finer texture differences. Experimental results demonstrate that this baseline surpasses the existing methods. The benchmark is publically available at https://github.com/wmbai/LSCGB.

Augmented reality-based autostereoscopic surgical visualization system for telesurgery.

PURPOSE: The visualization of remote surgical scenes is the key to realizing the remote operation of surgical robots. However, current non-endoscopic surgical robot systems lack an effective visualization tool to offer sufficient surgical scene information and depth perception. METHODS: We propose a novel autostereoscopic surgical visualization system integrating 3D intraoperative scene reconstruction, autostereoscopic 3D display, and augmented reality-based image fusion. The preoperative organ structure and the intraoperative surface point cloud are obtained from medical imaging and the RGB-D camera, respectively, and aligned by an automatic marker-free intraoperative registration algorithm. After registration, preoperative meshes with precalculated illumination and intraoperative textured point cloud are blended in real time. Finally, the fused image is shown on a 3D autostereoscopic display device to achieve depth perception. RESULTS: A prototype of the autostereoscopic surgical visualization system was built. The system had a horizontal image resolution of 1.31 mm, a vertical image resolution of 0.82 mm, an average rendering rate of 33.1 FPS, an average registration rate of 20.5 FPS, and average registration errors of approximately 3 mm. A telesurgical robot prototype based on 3D autostereoscopic display was built. The quantitative evaluation experiments showed that our system achieved similar operational accuracy (1.79 ± 0.87 mm) as the conventional system (1.95 ± 0.71 mm), while having advantages in terms of completion time (with 34.11% reduction) and path length (with 35.87% reduction). Post-experimental questionnaires indicated that the system was user-friendly for novices and experts. CONCLUSION: We propose a 3D surgical visualization system with augmented instruction and depth perception for telesurgery. The qualitative and quantitative evaluation results illustrate the accuracy and efficiency of the proposed system. Therefore, it shows great prospects in robotic surgery and telesurgery.

Intelligent optical diagnosis and treatment system for automated image-guided laser ablation of tumors.

PURPOSE: For tumor resections near critical structures, accurate identification of tumor boundaries and maximum removal are the keys to improve surgical outcome and patient survival rate, especially in neurosurgery. In this paper, we propose an intelligent optical diagnosis and treatment system for tumor removal, with automated lesion localization and laser ablation. METHODS: The proposed system contains a laser ablation module, an optical coherence tomography (OCT) unit, and a robotic arm along with a stereo camera. The robotic arm can move the OCT sample arm and the laser ablation front-end to the suspected lesion area. The corresponding diagnosis and treatment procedures include computer-aided lesion segmentation using OCT, automated ablation planning, and laser control. The ablation process is controlled by a deflectable mirror, and a non-common-path ablation planning algorithm based on the transformation from lesion positions to mirror deflection angles is presented. RESULTS: Phantom and animal experiments are carried out for system verification. The robot could reach the planned position with high precision, which is approximately 1.16 mm. Tissue classification with OCT images achieves 91.7% accuracy. The error of OCT-guided automated laser ablation is approximately 0.74 mm. Experiments on mouse brain tumors show that the proposed system is capable of clearing lesions efficiently and precisely. We also conducted an ex vivo porcine brain experiment to verify the whole process of the system. CONCLUSION: An intelligent optical diagnosis and treatment system is proposed for tumor removal. Experimental results show that the proposed system and method are promising for precise and intelligent theranostics. Compared to conventional cancer diagnosis and treatment, the proposed system allows for automated operations monitored in real-time, with higher precision and efficiency.

Insulin Injection Knowledge, Attitudes, and Practices of Nurses in China: A Cross-Sectional Nationwide Study.

INTRODUCTION: To evaluate insulin injection knowledge, attitudes, and practices of nurses across China in order to provide reference for the formulation of a national unified standard of insulin injection practice and the targeted implementation of standardized training on insulin injection for nurses. METHODS: We enrolled nurses who worked and injected insulin at grassroot hospitals including community health service centers and township clinics, secondary and tertiary care hospitals across China between July 28, 2019 and August 30, 2019. A nurse insulin injection knowledge, attitude, and practice questionnaire was used to evaluate the knowledge, attitude, and practice level of nurses. RESULTS: A total of 223,368 nurses were included in the study. The mean knowledge score was 13.70 ± 3.30 and 35.19% had a poor knowledge score. The mean attitude score was 17.18 ± 2.69 for the study nurses; merely 3.15% had a poor attitude score. The mean practice score of the study population was 83.03 ± 8.16 and only 0.88% had a poor practice score. Pearson correlation analysis showed significant correlation between the knowledge score and the attitude score (r = 0.29; P < 0.001), the knowledge score and the practice score (r = 0.27; P < 0.001), and between the attitude score and the practice score (r = 0.56; P < 0.001). A multivariate analysis revealed that nurses with higher knowledge scores were also more likely to have higher attitude scores and practice scores, and nurses with higher attitude scores were also more likely to have higher practice scores. CONCLUSION: Chinese nurses have a good attitude and behavior towards insulin injection, while their knowledge of insulin injection is insufficient. It is also revealed that knowledge of insulin injection can directly or indirectly affect insulin injection behavior through attitude, indicating that hospitals should formulate unified insulin injection norms and regularly organize relevant training and assessment so as to improve nurses' knowledge, attitude, and behavior of insulin injection.

EDP: An Efficient Decomposition and Pruning Scheme for Convolutional Neural Network Compression.

Model compression methods have become popular in recent years, which aim to alleviate the heavy load of deep neural networks (DNNs) in real-world applications. However, most of the existing compression methods have two limitations: 1) they usually adopt a cumbersome process, including pretraining, training with a sparsity constraint, pruning/decomposition, and fine-tuning. Moreover, the last three stages are usually iterated multiple times. 2) The models are pretrained under explicit sparsity or low-rank assumptions, which are difficult to guarantee wide appropriateness. In this article, we propose an efficient decomposition and pruning (EDP) scheme via constructing a compressed-aware block that can automatically minimize the rank of the weight matrix and identify the redundant channels. Specifically, we embed the compressed-aware block by decomposing one network layer into two layers: a new weight matrix layer and a coefficient matrix layer. By imposing regularizers on the coefficient matrix, the new weight matrix learns to become a low-rank basis weight, and its corresponding channels become sparse. In this way, the proposed compressed-aware block simultaneously achieves low-rank decomposition and channel pruning by only one single data-driven training stage. Moreover, the network of architecture is further compressed and optimized by a novel Pruning & Merging (PM) module which prunes redundant channels and merges redundant decomposed layers. Experimental results (17 competitors) on different data sets and networks demonstrate that the proposed EDP achieves a high compression ratio with acceptable accuracy degradation and outperforms state-of-the-arts on compression rate, accuracy, inference time, and run-time memory.

tRNA methyltransferase 10 homologue A (TRMT10A) mutation in a Chinese patient with diabetes, insulin resistance, intellectual deficiency and microcephaly.

INTRODUCTION: Loss-of-function mutations in tRNA methyltransferase 10 homologue A (TRMT10A), a tRNA methyltransferase, have recently been described as a monogenic cause of early-onset diabetes with microcephaly, epilepsy and intellectual disability. RESEARCH DESIGN AND METHODS: We report a Chinese young patient who was diagnosed with diabetes mellitus as a result of a TRMT10A mutation. RESULTS: A homozygous mutation c.496-1G>A in TRMT10A was identified using targeted next-generation sequencing and confirmed by PCR/Sanger sequencing. In addition to being diagnosed with diabetes, the patient also has microcephaly and intellectual deficiency. The diabetes was due to marked insulin resistance and responded very well to metformin treatment. CONCLUSION: Our case is the first report in the Asian population. It adds to current knowledge of TRMT10A related with young-onset non-insulin-dependent diabetes and confirms the a single previous report of insulin resistance in this syndrome. Genomic testing should be considered in children with non-insulin-dependent diabetes with intellectual disability and microcephaly. A clear genetic diagnosis is helpful for early detection and treatment addressing insulin resistance.

Anisotropic Convolution for Image Classification.

Convolutional neural networks are built upon simple but useful convolution modules. The traditional convolution has a limitation on feature extraction and object localization due to its fixed scale and geometric structure. Besides, the loss of spatial information also restricts the networks' performance and depth. To overcome these limitations, this paper proposes a novel anisotropic convolution by adding a scale factor and a shape factor into the traditional convolution. The anisotropic convolution augments the receptive fields flexibly and dynamically depending on the valid sizes of objects. In addition, the anisotropic convolution is a generalized convolution. The traditional convolution, dilated convolution and deformable convolution can be viewed as its special cases. Furthermore, in order to improve the training efficiency and avoid falling into a local optimum, this paper introduces a simplified implementation of the anisotropic convolution. The anisotropic convolution can be applied to arbitrary convolutional networks and the enhanced networks are called ACNs (anisotropic convolutional networks). Experimental results show that ACNs achieve better performance than many state-of-the-art methods and the baseline networks in tasks of image classification and object localization, especially in classification task of tiny images.

High Prevalence of a Monogenic Cause in Han Chinese Diagnosed With Type 1 Diabetes, Partly Driven by Nonsyndromic Recessive WFS1 Mutations.

It is estimated that ∼1% of European ancestry patients clinically diagnosed with type 1 diabetes (T1D) actually have monogenic forms of the disease. Because of the much lower incidence of true T1D in East Asians, we hypothesized that the percentage would be much higher. To test this, we sequenced the exome of 82 Chinese Han patients clinically diagnosed with T1D but negative for three autoantibodies. Analysis focused on established or proposed monogenic diabetes genes. We found credible mutations in 18 of the 82 autoantibody-negative patients (22%). All mutations had consensus pathogenicity support by five algorithms. As in Europeans, the most common gene was HNF1A (MODY3), in 6 of 18 cases. Surprisingly, almost as frequent were diallelic mutations in WFS1, known to cause Wolfram syndrome but also described in nonsyndromic cases. Fasting C-peptide varied widely and was not predictive. Given the 27.4% autoantibody negativity in Chinese and 22% mutation rate, we estimate that ∼6% of Chinese with a clinical T1D diagnosis have monogenic diabetes. Our findings support universal sequencing of autoantibody-negative cases as standard of care in East Asian patients with a clinical T1D diagnosis. Nonsyndromic diabetes with WSF1 mutations is not rare in Chinese. Its response to alternative treatments should be investigated.

Context-Dependent Random Walk Graph Kernels and Tree Pattern Graph Matching Kernels with Applications to Action Recognition.

Graphs are effective tools for modeling complex data. Setting out from two basic substructures, random walks and trees, we propose a new family of context-dependent random walk graph kernels and a new family of tree pattern graph matching kernels. In our context-dependent graph kernels, context information is incorporated into primary random walk groups. A multiple kernel learning algorithm with a proposed l1,2-norm regularization is applied to combine context-dependent graph kernels of different orders. This improves the similarity measurement between graphs. In our tree-pattern graph matching kernel, a quadratic optimization with a sparse constraint is proposed to select the correctly matched tree-pattern groups. This augments the discriminative power of the tree-pattern graph matching. We apply the proposed kernels to human action recognition, where each action is represented by two graphs which record the spatiotemporal relations between local feature vectors. Experimental comparisons with state-of-the-art algorithms on several benchmark datasets demonstrate the effectiveness of the proposed kernels for recognizing human actions. It is shown that our kernel based on tree-pattern groups, which have more complex structures and exploit more local topologies of graphs than random walks, yields more accurate results but requires more runtime than the context-dependent walk graph kernel.

Tumor immunotherapy: New aspects of natural killer cells.

A group of impressive immunotherapies for cancer treatment, including immune checkpoint-blocking antibodies, gene therapy and immune cell adoptive cellular immunotherapy, have been established, providing new weapons to fight cancer. Natural killer (NK) cells are a component of the first line of defense against tumors and virus infections. Studies have shown dysfunctional NK cells in patients with cancer. Thus, restoring NK cell antitumor functionality could be a promising therapeutic strategy. NK cells that are activated and expanded ex vivo can supplement malfunctional NK cells in tumor patients. Therapeutic antibodies, chimeric antigen receptor (CAR), or bispecific proteins can all retarget NK cells precisely to tumor cells. Therapeutic antibody blockade of the immune checkpoints of NK cells has been suggested to overcome the immunosuppressive signals delivered to NK cells. Oncolytic virotherapy provokes antitumor activity of NK cells by triggering antiviral immune responses. Herein, we review the current immunotherapeutic approaches employed to restore NK cell antitumor functionality for the treatment of cancer.