Search

Abstract The application of deep learning models to precision medical diagnosis often requires the aggregation of large amounts of medical data to effectively train high-quality models. However, data privacy protection mechanisms make it difficult to perform medical data collection from different medical institutions. In autism spectrum disorder (ASD) diagnosis, automatic diagnosis using multimodal information from heterogeneous data has not yet achieved satisfactory performance. To address the privacy preservation issue as well as to improve ASD diagnosis, we propose a deep learning framework using multimodal feature fusion and hypergraph neural networks for disease prediction in federated learning (FedHNN). By introducing the federated learning strategy, each local model is trained and computed independently in a distributed manner without data sharing, allowing rapid scaling of medical datasets to achieve robust and scalable deep learning predictive models. To further improve the performance with privacy preservation, we improve the hypergraph model for multimodal fusion to make it suitable for autism spectrum disorder (ASD) diagnosis tasks by capturing the complementarity and correlation between modalities through a hypergraph fusion strategy. The results demonstrate that our proposed federated learning-based prediction model is superior to all local models and outperforms other deep learning models. Overall, our proposed FedHNN has good results in the work of using multi-site data to improve the performance of ASD identification.

Wound healing is a complex process that is critical for maintaining the barrier function of the skin. However, when a large quantity of microorganisms invade damaged skin for an extended period, they can cause local and systemic inflammatory responses. If left untreated, this condition may lead to chronic infected wounds. Infected wounds significantly escalate wound management costs worldwide and impose a substantial burden on patients and healthcare systems. Recent clinical trial results suggest that the utilization of effective antimicrobial wound dressing could represent the simplest and most cost-effective strategy for treating infected wounds, but there has hitherto been no comprehensive evaluation reported on the efficacy of antimicrobial wound dressings in promoting wound healing. Therefore, this review aims to systematically summarize the various types of antimicrobial wound dressings and the current research on antimicrobial agents, thereby providing new insights for the innovative treatment of infected wounds.

Association rule mining utilizing metaheuristic algorithms is a prominent area of study in the field of data mining. However, when working with extensive data, conventional metaheuristic algorithms exhibit limited search efficiency and face challenges in deriving high-quality rules in multi-objective association rule mining. In order to tackle this issue, a novel approach called the adaptive Weibull distribution sparrow search algorithm is introduced. This algorithm leverages the adaptive Weibull distribution to improve the traditional sparrow search algorithm’s capability to escape local optima and enhance convergence during different iterations. Secondly, an enhancement search strategy and a multidirectional learning strategy are introduced to expand the search range of the population. This paper empirically evaluates the proposed method under real datasets and compares it with other leading methods by using three association rule metrics, namely, support, confidence, and lift, as the fitness function. The experimental results show that the quality of the obtained association rules is significantly improved when dealing with datasets of different sizes.

In ship detection through remote sensing images, optical images often provide rich details and texture information; however, the quality of such optical images can be affected by cloud and fog interferences. In contrast, Synthetic Aperture Radar (SAR) provides all-weather and all-day imaging capabilities; however, SAR images are susceptible to interference from complex sea clutter. Cooperative ship detection combining the advantages of optical and SAR images can enhance the detection performance of ships. In this paper, by focusing on the slight shift of ships in a small neighborhood range in the prior and later temporal images, we propose a method for cooperative ship detection based on neighborhood saliency in multisource heterogeneous remote sensing images, including optical and SAR data. Initially, a sea-land segmentation algorithm of optical and SAR images is applied to reduce interference from land regions. Next, single-source ship detection from optical and SAR images is performed using the RetinaNet and YOLOv5s models, respectively. Then, we introduce a multisource cooperative ship target detection strategy based on the neighborhood window opening of single-source detection results in remote sensing images and secondary detection of neighborhood salient ships. This strategy further leverages the complementary advantages of both optical and SAR heterogeneous images, reducing the possibility of missing ship and false alarms to improve overall detection performance. The performance of the proposed method has been validated using optical and SAR remote sensing data measured from Yantai, China, in 2022. Compared with existing ship detection methods, our method improves detection accuracy AP50 by ≥1.9%, demonstrating its effectiveness and superiority.

BackgroundThis research seeks to formulate a prognostic model for forecasting prostate cancer recurrence by examining the interaction between mitochondrial function and programmed cell death (PCD).MethodsThe research involved analyzing four gene expression datasets from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) using univariate Cox regression. These analyses identified genes linked with mitochondrial function and PCD that correlate with recurrence prognosis. Various machine learning algorithms were then employed to construct an optimal predictive model.ResultsA key outcome was the creation of a mitochondrial-related programmed cell death index (mtPCDI), which effectively predicts the prognosis of prostate cancer patients. It was observed that individuals with lower mtPCDI exhibited higher immune activity, correlating with better recurrence outcomes.ConclusionThe study demonstrates that mtPCDI can be used for personalized risk assessment and therapeutic decision-making, highlighting its clinical significance and providing insights into the biological processes affecting prostate cancer recurrence.

Spiking neural networks (SNNs) offer a promising energy-efficient alternative to artificial neural networks (ANNs), in virtue of their high biological plausibility, rich spatial-temporal dynamics, and event-driven computation. The direct training algorithms based on the surrogate gradient method provide sufficient flexibility to design novel SNN architectures and explore the spatial-temporal dynamics of SNNs. According to previous studies, the performance of models is highly dependent on their sizes. Recently, direct training deep SNNs have achieved great progress on both neuromorphic datasets and large-scale static datasets. Notably, transformer-based SNNs show comparable performance with their ANN counterparts. In this paper, we provide a new perspective to summarize the theories and methods for training deep SNNs with high performance in a systematic and comprehensive way, including theory fundamentals, spiking neuron models, advanced SNN models and residual architectures, software frameworks and neuromorphic hardware, applications, and future trends.

Semantic segmentation of remote sensing images (SSRSIs), which aims to assign a category to each pixel in remote sensing images, plays a vital role in a broad range of applications, such as environmental monitoring, urban planning, and land resource utilization. Recently, with the successful application of deep learning in remote sensing, a substantial amount of work has been aimed at developing SSRSI methods using deep learning models. In this survey, we provide a comprehensive review of SSRSI. First, we review the current mainstream semantic segmentation models based on deep learning. Next, we analyze the main challenges faced by SSRSI and comprehensively summarize the current research status of deep-learning-based SSRSI, especially some new directions in SSRSI are outlined, including semisupervised and weakly-supervised SSRSI, unsupervised domain adaption in SSRSI, multimodal data-fusion-based SSRSI, and pretrained models for SSRSI. Then, we examine the most widely used datasets and metrics and review the quantitative results and experimental performance of some representative methods of SSRSI. Finally, we discuss promising future research directions in this area.

Background: Artesunate (ART) has the potential to modulate the nuclear factor kappa B (NF-κB) and Notch1/Hes1 signaling pathways, which play crucial roles in the pathogenesis of osteoporosis. This study aims to explore whether ART participates in the progression of osteoporosis by regulating these signaling pathways. Methods: In the in vitro experiments, we treated bone marrow mesenchymal stem cells (BMSCs) with different concentrations of ART (0, 3, 6, 12 µM) and evaluated osteogenic differentiation using alkaline phosphatase staining (ALP) and alizarin red S staining (ARS) staining. The expression levels of osteocalcin (OCN), RUNT-related transcription factor 2 (RUNX2), osteoprotegerin (OPG), and receptor activator of the nuclear factor kappa ligand (RANKL) were detected by real-time quantitative PCR (RT-qPCR). The effects of ART on NF-κB p65 and Notch1 protein expression were analyzed by Western blot (WB) and immunofluorescence (IF). In the in vivo experiments, a postmenopausal osteoporosis rat model was established via ovariectomy. Bone tissue pathological injury was evaluated using hematoxylin eosin (HE) staining. Serum ALP levels were measured using a kit, bone density was determined by dual-energy X-ray absorptiometry, and serum levels of bone gla protein (BGP), OPG, RANKL, tumor necrosis factor-alpha (TNF-α), interleukin 6 (IL-6), and IL-1β were measured by enzyme-linked immunosorbent assay (ELISA). Additionally, the expression of NF-κB p65 and Notch1 in tissues was assessed by immunohistochemistry. Results: In vitro experiments revealed that compared to the control group, ART dose-dependently promoted BMSCs proliferation and enhanced their osteogenic differentiation capability. The expression of OCN, RUNX2, and OPG significantly increased in the ART-treated group, while RANKL expression decreased significantly (p < 0.05). ART significantly inhibited the expression of NF-κB p65 and Notch1/Hes1 signaling pathway proteins (p < 0.05). Compared to ART treatment alone, combined treatment with ART and phorbol myristate acetate (PMA) or valproic acid (VPA) resulted in increased expression of NF-κB p65 and Notch1 proteins and decreased osteogenic differentiation capability (p < 0.05). In vivo experiments showed that in rats treated with ART, bone damage was significantly reduced, bone density and mineral content were restored considerably, and the expression of inflammatory factors (TNF-α, IL-6, IL-1β) decreased significantly (p < 0.05). Additionally, ART treatment significantly reduced the expression of NF-κB p65 and Notch1 proteins, increased OPG expression, and decreased BGP and RANKL levels (p < 0.05). Conclusion: In summary, ART facilitates the osteogenic differentiation of BMSCs by inhibiting the NF-κB and Notch1/Hes1 signaling pathways, thereby exerting significant protective effects against osteoporosis.

IntroductionSpiking Neural Networks (SNNs), inspired by brain science, offer low energy consumption and high biological plausibility with their event-driven nature. However, the current SNNs are still suffering from insufficient performance.MethodsRecognizing the brain's adeptness at information processing for various scenarios with complex neuronal connections within and across regions, as well as specialized neuronal architectures for specific functions, we propose a Spiking Global-Local-Fusion Transformer (SGLFormer), that significantly improves the performance of SNNs. This novel architecture enables efficient information processing on both global and local scales, by integrating transformer and convolution structures in SNNs. In addition, we uncover the problem of inaccurate gradient backpropagation caused by Maxpooling in SNNs and address it by developing a new Maxpooling module. Furthermore, we adopt spatio-temporal block (STB) in the classification head instead of global average pooling, facilitating the aggregation of spatial and temporal features.ResultsSGLFormer demonstrates its superior performance on static datasets such as CIFAR10/CIFAR100, and ImageNet, as well as dynamic vision sensor (DVS) datasets including CIFAR10-DVS and DVS128-Gesture. Notably, on ImageNet, SGLFormer achieves a top-1 accuracy of 83.73% with 64 M parameters, outperforming the current SOTA directly trained SNNs by a margin of 6.66%.DiscussionWith its high performance, SGLFormer can support more computer vision tasks in the future. The codes for this study can be found in https://github.com/ZhangHanN1/SGLFormer.