Search

Traditional object detection methods usually underperform when locating tiny or small drones against complex backgrounds, since the appearance features of the targets and the backgrounds are highly similar. To address this, inspired by the magnocellular motion processing mechanisms, we proposed to utilize the spatial–temporal characteristics of the flying drones based on spiking neural networks, thereby developing the Magno-Spiking Neural Network (MG-SNN) for drone detection. The MG-SNN can learn to identify potential regions of moving targets through motion saliency estimation and subsequently integrates the information into the popular object detection algorithms to design the retinal-inspired spiking neural network module for drone motion extraction and object detection architecture, which integrates motion and spatial features before object detection to enhance detection accuracy. To design and train the MG-SNN, we propose a new backpropagation method called Dynamic Threshold Multi-frame Spike Time Sequence (DT-MSTS), and establish a dataset for the training and validation of MG-SNN, effectively extracting and updating visual motion features. Experimental results in terms of drone detection performance indicate that the incorporation of MG-SNN significantly improves the accuracy of low-altitude drone detection tasks compared to popular small object detection algorithms, acting as a cheap plug-and-play module in detecting small flying targets against complex backgrounds.

Abstract Magnetic resonance imaging (MRI) is a non‐interposition imaging technique that provides rich anatomical and physiological information. Yet it is limited by the long imaging time. Recently, deep neural networks have shown potential to significantly accelerate MRI. However, most of these approaches ignore the correlation between adjacent slices in MRI image sequences. In addition, the existing deep learning‐based methods for MRI are mainly based on convolutional neural networks (CNNs). They fail to capture long‐distance dependencies due to the small receptive field. Inspired by the feature similarity in adjacent slices and impressive performance of Transformer for exploiting the long‐distance dependencies, a novel multislice texture transformer network is presented for undersampled MRI reconstruction (TTMRI). Specifically, the proposed TTMRI is consisted of four modules, namely the texture extraction, correlation calculation, texture transfer and texture synthesis. It takes three adjacent slices as inputs, in which the middle one is the target image to be reconstructed, and the other two are auxiliary images. The multiscale features are extracted by the texture extraction module and their inter‐dependencies are calculated by the correlation calculation module, respectively. Then the relevant features are transferred by the texture transfer module and fused by the texture synthesis module. By considering inter‐slice correlations and leveraging the Transformer architecture, the joint feature learning across target and adjacent slices are encouraged. Moreover, TTMRI can be stacked with multiple layers to recover more texture information at different levels. Extensive experiments demonstrate that the proposed TTMRI outperforms other state‐of‐the‐art methods in both quantitative and qualitative evaluationsions.

In the real world, the large and complex nature of text increases the difficulty of tagging and results in a limited amount of tagged text. Few-shot Named Entity Recognition(NER) only uses a small amount of annotation data to identify and classify entities. It avoids the above problems. Few-shot learning methods usually use prior knowledge to achieve good results. However, prior knowledge may become a confounding factor affecting the relation between sample features and real labels. This problem leads to bias and difficulty accurately capturing class. To solve this problem, a new model, Few-shot Named Entity Recognition via Encoder and Class Intervention, is proposed based on causality. We show that we can steer the model to manufacture interventions on encoder and class, and reduce the interference of confounding factors. Specifically, while cross-sample attention perturbation is used in the encoder layer, a practical causal relation between feature and classification label is developed in the class layer. This way is an attempt of causal methodology in the Few-shot Named Entity Recognition task, which improves the discrimination ability of the NER classifier. Experimental results demonstrate that our model outperforms baseline models in both 5-way and 10-way on two NER datasets.

Interest in anti-UAV detection systems has increased due to growing concerns about the security and privacy issues associated with unmanned aerial vehicles (UAVs). Achieving real-time detection with high accuracy, while accommodating the limited resources of edge-computing devices poses a significant challenge for anti-UAV detection. Existing deep learning-based models for anti-UAV detection often cannot balance accuracy, processing speed, model size, and computational efficiency. To address these limitations, a lightweight and efficient anti-UAV detection model, DRBD-YOLOv8, is proposed in this paper. The model integrates several innovations, including the application of a Re-parameterization Cross-Stage Efficient Layered Attention Network (RCELAN) and a Bidirectional Feature Pyramid Network (BiFPN), to enhance feature processing capabilities while maintaining a lightweight design. Furthermore, DN-ShapeIoU, a novel loss function, has been established to enhance detection accuracy, and depthwise separable convolutions have been included to decrease computational complexity. The experimental results showed that the proposed model outperformed YOLOV8n in terms of mAP50, mAP95, precision, and FPS while reducing GFLOPs and parameter count. The DRBD-YOLOv8 model is almost half the size of the YOLOv8n model, measuring 3.25 M. Its small size, fast speed, and high accuracy combine to provide a lightweight, accurate device that is excellent for real-time anti-UAV detection on edge-computing devices.

Few-shot Named Entity Recognition (NER) is a challenging task that involves identifying new entity types using a limited number of labeled instances for training. Currently, the majority of Few-shot NER methods are based on span, which pay more attention to the boundary information of the spans as candidate entities and the entity-level information. However, these methods often overlook token-level semantic information, which can limit their effectiveness. To address this issue, we propose a novel Joint Span and Token (JST) framework that integrates both the boundary information of an entity and the semantic information of each token that comprises an entity. The JST framework employs span features to extract the boundary features of the entity and token features to extract the semantic features of each token. Additionally, to reduce the negative impact of the Other class, we introduce a method to separate named entities from the Other class in semantic space, which helps to improve the distinction between entities and the Other class. In addition, we used GPT to do data augmentation on the support sentences, generating similar sentences to the original ones. These sentences increase the diversity of the sample and the reliability of our model. Our experimental results on the Few-NERD11 https://ningding97.github.io/fewnerd/. and SNIPS22 https://github.com/AtmaHou/FewShotTagging. datasets demonstrate that our model outperforms existing methods in terms of performance.

Abstract Background Causal inference is a crucial element within medical decision-making. There have been many methods for investigating potential causal relationships between disease and treatment options developed in recent years, which can be categorized into two main types: observational studies and experimental studies. However, due to the nature of experimental studies, financial resources, human resources, and patients' ethical considerations, researchers cannot fully control the exposure of the research participants. Furthermore, most existing observational research designs are limited to determining causal relationships and cannot handle observational data, let alone determine the dosages needed for medical research. Results This paper presents a new experimental strategy called quasi-intervention for quantifying the causal effect between disease and treatment options in observed data by using a causal inference method, which converts the potential effect of different treatment options on disease into computing differences in the conditional probability. We evaluated the accuracy of the quasi-intervention by quantifying the impact of adjusting Chinese patients’ neutrophil-to-lymphocyte ratio (NLR) on their overall survival (OS) (169 lung cancer patients and 79 controls).The results agree with the literature in this study, consisting of nine papers on cohort studies on the NLR and the prognosis of lung cancer patients, proving that our method is correct. Conclusion Taken together, the results imply that quasi-intervention is a promising method for quantifying the causal effect between disease and treatment options without clinical trials, and it could improve confidence about treatment options' efficacy and safety.

The rapid expansion of nuclear technology across various sectors due to global economic growth has led to a substantial rise in the transportation of radioactive materials. The International Atomic Energy Agency (IAEA) estimates that approximately 20 million shipments of radioactive materials occur annually. In this context, ensuring the safety and security of radioactive material transportation is of significant importance. IAEA’s “Security of Radioactive Materials in Transport” (Nuclear Security Series No. 9-G) mandates that an effective transport security system should provide immediate detection of any unauthorized removal of the packages. In the present study, an innovative Adam-optimized BP neural network model is developed for detecting unauthorized movements of radioactive material packages. To analyze the performance of the proposed algorithm, numerous experiments were conducted. The results demonstrate that the proposed method achieves a 99.17% accuracy rate in detecting unauthorized movements of radioactive materials, with a missed alarm rate of 0.72% and a false alarm rate of 0.1%. This method also enables real-time detection of unauthorized removal of radioactive materials and effectively enhances the security of radioactive materials during transport to reduce the risks of theft, loss, diversion, or sabotage.

Mosaicking is a crucial step in the application of remote sensing images. The amount of remote sensing image data has grown rapidly, along with the expansion of observed areas and increased image resolution. As a result, traditional serial mosaicking techniques are facing significant challenges. In recent times, various studies have utilized high-performance computing to hasten image mosaicking and attain favorable outcomes. Nevertheless, the current research only accelerates mosaicking through external technology, without optimizing from the perspective of algorithm flow, which introduces unnecessary data I/O and slows down the mosaicking. This paper introduces a rapid parallel remote sensing image mosaicking algorithm utilizing read filtering. To begin with, the target images are divided into blocks and stored in a distributed file system. Subsequently, the image blocks are read and filtered based on a designated input format. Finally, the overlapping and non-overlapping areas are read and processed asynchronously, reducing the data I/O and computing overhead, thereby improving the efficiency of parallel computing. The experiments indicate that the mosaicking algorithm introduced in this paper enhances throughput and speedup by an average of 1.38 MB/S and 0.87 relative to the current techniques, respectively, concerning various datasets and cores. This study provides a theoretical foundation and novel ideas for processing remote sensing images on cluster platforms.

For loss and theft in the transport of radioactive materials by a single vehicle, this article summarizes the characteristics of “illegal movement” and establishes a security system that senses its inter-relationship and responds though the group network relationship. The security system reminds the vehicle crew through on-site response and linkage response. A failure detection method for on-site response is proposed, that is, the push model is used first, and when the measurement results are suspected, the pull model is used to further confirm the failure. The failure detection for linkage response adopts the push model. According to the different security requirements of the basic and enhanced transportation, the principle of setting the timeout threshold in the failure detection algorithm is proposed. In the enhanced type, the value is smaller, otherwise the value is larger. A specific timeout threshold quantification scheme is proposed. Experiments show that the method proposed in this article is effective.

The loss and theft of radioactive material in transport can be attributed to the illegal movement. In order to distinguish it from the movement caused by the turbulence of the transportation vehicle, this paper proposes the criterion of “illegal movement” as the movement of radioactive materials outside the transportation compartment. Since the interior of the compartment is generally a metal environment, this paper proposes wireless signal strength data as a sensing method. The wireless signal strength data is filtered and converted into distance data. We construct a spatial triangle perpendicular to the top and sides of the compartment based on the distance data. When the radioactive material is inside the compartment, the angle between its corresponding point and the top plane of the compartment is less than 90°. Once it moves out of the compartment, the angle will be greater than 90°. Based on this, a sensing method of “illegal movement” based on spatial triangles is proposed. The simulation research shows that the scheme proposed in this paper is feasible.

This paper proposes to design security measures based on the radioactive material package as the basic unit. The principle of four-layer defense in depth is put forward. Based on the concept of self-security intelligence, combined with out-of-vehicle monitoring, in-vehicle monitoring, and Beidou positioning technology, a security system for transport of radioactive materials was designed. It realized the perception, early warning, delay, and alarm functions and greatly improved the security.

Abstracts The large-scale loach (Paramisgurnus dabryanus; Cypriniformes: Cobitidae) is primarily distributed in East Asia. It is an important economic fish species characterized by fast growth, temperature-dependent sex determination and the ability to breathe air. Currently, molecular mechanism studies related to some aspects such as sex determination, toxicology, feed nutrition, growth and genetic evolution have been conducted. However, the lack of a high-quality reference genome has hindered further research. In this study, we performed PacBio HiFi and Hi-C sequencing on a female (ZW) specimen and assembled the first high-quality chromosome-level genome of the large-scale loach. The assembled genome comprises 24 chromosomes, with a total length of 1.04 Gb, a scaffold N50 of 41.7 Mb, and a BUSCO completeness of 95.8%, including 28,311 protein-coding genes. These findings not only provide new insights into the genome structure of the large-scale loach but also establish a crucial reference point for omics studies and serve as an essential genomic resource for breeding programs in this species.

Based on the diversity of transmission routes and host heterogeneity of some infectious diseases, a dynamical model with multi-age-structured, asymptomatic infections, as well as horizontal and vectorial transmission, is proposed. First, the existence and uniqueness of the global positive solution of this model is discussed and the exact expression of the basic reproduction number $ \mathcal{R}_0 $ is obtained using the linear approximation method. Further, we deduce that the disease-free steady state $ \mathcal{E}^0 $ is globally asymptotically stable for $ \mathcal{R}_0 < 1 $, the endemic steady state $ \mathcal{E}^* $ exists and the disease is persistent for $ \mathcal{R}_0 > 1 $. In addition, the locally asymptotically stability of $ \mathcal{E}^* $ is also obtained under some certain conditions. Next, our model is extended to a control problem and the existence and uniqueness of the optimal control by using the Gateaux derivative. Finally, numerical simulations are used to explain the main theoretical results and discuss the impact of age-structured parameters and control strategies on the prevention and control of vector-borne infectious diseases.

Abstract Objective The study aimed to identify distinct molecular subtypes of endometrial cancer (EC) by immunohistochemistry and to analyze their pathological characteristics, independent prognostic factors, and patient survival outcomes for potential clinical applications. Method 576 patients with preoperative EC confined to the uterus were divided into three subgroups based on the immunohistochemical detection method: MMR-deficiency (MMRd), P53 wild type (P53wt) and P53 abnormal (P53abn). These subgroups were retrospectively analyzed, and their pathological characteristics, prognostic factors and survival outcomes were compared. Results We identified 401 (69.6%), 123 (21.4%), and 52 (9%) cases of P53wt, MMRd, and P53abn subgroups, respectively. A significant difference was observed in the median age of onset, tumor stage, high-grade tumor differentiation, non-endometrioid carcinoma, myometrial invasion, lymphovascular invasion, the incidence of lymph node metastasis postoperative, and expression of ER and PR receptors among the three groups. Pathological type, lymphovascular invasion, ER and PR expression were identified as independent prognostic factors for disease-free survival (DFS). Additionally, pathological type, lymphovascular invasion, myometrial invasion, and PR expression were recognized as independent prognostic factors for overall survival (OS) in the study cohort. However, the survival outcome for P53abn was the worst, with lymphovascular invasion identified as an independent prognostic factor for DFS. Lymph node status, FIGO stage, and ER expression were identified as independent prognostic factors for OS. Conclusion The study concludes that immunohistochemical detection-based subtyping of EC holds clinical practicality and can be employed to explore both pathological and clinical prognoses for EC patients.

Abstract Background Bipolar disorder is a complex polygenic disorder that is characterized by recurrent episodes of depression and mania, the heterogeneity of which is likely complicated by epigenetic modifications that remain to be elucidated. Methods We performed transcriptomic analysis of peripheral blood RNA from monozygotic (MZ) twins discordant for bipolar disorder to identify disease-associated differentially expressed long noncoding RNAs (DE-lncRNAs), which were further validated in the PsychENCODE brain RNA-seq dataset. We then performed behavioral tests, electrophysiological assays, chromatin immunoprecipitation, and PCR to investigate the function of DE-lncRNAs in the mouse and cell models. Statistical analyses were performed using GraphPad Prism 9.0 or SPSS. Results We identified a bipolar disorder-associated upregulated long non-coding RNA (lncRNA), AP1AR-DT. We observed that overexpression of AP1AR-DT in the mouse medial prefrontal cortex (mPFC) resulted in a reduction of both the total spine density and the spontaneous excitatory postsynaptic current (sEPSC) frequency of mPFC neurons as well as depressive and anxiety-like behaviors. A combination of the results of brain transcriptome analysis of AP1AR-DT overexpressing mice brains with the known genes associated with bipolar disorder revealed that NEGR1, which encodes neuronal growth regulator 1, is one of the AP1AR-DT targets and is reduced in vivo upon gain of AP1AR-DT in mice. We further demonstrated that overexpression of recombinant Negr1 in the mPFC neurons of AP1AR-DT OE mice ameliorates depressive and anxiety-like behaviors and normalizes the reduced excitatory synaptic transmission induced by the gain of AP1AR-DT. We finally identified that AP1AR-DT reduces NEGR1 expression by competing for the transcriptional activator NRF1 in the overlapping binding site of the NEGR1 promoter region. Conclusions The epigenetic and pathophysiological mechanism linking AP1AR-DT to the modulation of depressive and anxiety-like behaviors and excitatory synaptic function provides etiological implications for bipolar disorder.

Abstract Purpose This study aimed to explore the prognostic factors and survival patterns based on the histological type for the perimenopausal endometrial carcinoma (PIPEC) patients treated with hysterectomy. Methods The PIPEC patients were selected from the Surveillance, Epidemiology, and End Results (SEER) database. Methods of random survival forest (RSF) and Cox regression were used to identify the possible prognostic factors of PIPEC patients. Then overall survival (OS) and cancer-specific survival (CSS) of PIPEC data were analyzed by histological types with regional lymph nodes status and SEER-stage to investigate the survival patterns of the PIPEC patients. Results A total of 14,178 PIPEC patients were included in the study. We found tumor size, grade, histology, SEER-stage, AJCC-stage, AJCC-T stage, metastasis to distant organs and regional lymph nodes status had a significant survival outcome for PIPEC both for OS and CSS (all p

Abstract Objective This study aims to assess the prognostic and diagnostic value of inflammatory indexes related to gestational diabetes mellitus (GDM) from the second trimester to the third trimester of pregnancy. Materials and methods In this study, we randomly selected 65 pregnant women diagnosed with GDM at our hospital from December 2022 to June 2023 to form the GDM group (n = 65). Additionally, 65 pregnant women at the same gestational weeks without GDM were selected as the Normal group (n = 65). We collected gestational information and serum samples at 24 and 36 weeks of gestation from the participants. The levels of NLRP3, IL-1Ra, and TBP-2 were determined using enzyme-linked immunosorbent assay (ELISA) to explore their changes during pregnancy. Further, this study analyzed the changes in the levels of NLRP3, IL-1Ra, and TBP-2 at 24 and 36 weeks of gestation in GDM patients and their correlation with gestational diabetes mellitus. Results The study showed that pre-pregnancy body mass index (BMI), neonatal weight, gestational hypertension, and macrosomia are significantly associated with the occurrence of GDM (P