With increasing urban underground space use, the adverse effects of static conditions-lacking natural light-thermal rhythms-on human health and performance are evident. Existing research focuses on isolated dynamic variables or static parameter combinations, while the dynamic synergy of light and thermal changes remains unclear. This study investigated four dynamic light-thermal conditions-Light Increase with Temperature Increase (LITI), Light Decrease with Temperature Decrease (LDTD), Light Increase with Temperature Decrease (LITD), and Light Decrease with Temperature Increase (LDTI)-and a static condition (SE) control. We measured physiological responses (heart rate variability: RMSSD, HF/LF), subjective perceptions (mood, fatigue, environmental acceptability), and cognitive performance (2-back, NASA-TLX). Results showed: (1) Dynamic light-thermal conditions significantly enhanced parasympathetic activity (RMSSD increased by 64.3%-151.9%) and autonomic balance (HF/LF ratio up 202.4%-349.2%), while improving positive affect (17.3%-28.5%) and environmental acceptability (e.g., LITI: +42.5%, LDTD: +51.1%) over SE. (2) Benefits were strongly modulated by light-thermal synergy: synergistic patterns (LITI, LDTD) consistently outperformed conflicting ones (LITD, LDTI) in task performance (LDTD was 34.4% faster than LITD) and subjective experience. The two synergistic patterns yielded statistically comparable benefits across most outcomes. Integrating light-thermal dynamicity and synergy into environmental design thus offers a promising strategy for enhancing human health and performance in underground spaces.

Existing image defogging methods have made significant advancements. However, these approaches are primarily optimized for land scenes, resulting in suboptimal performance when applied to overwater images due to the distinct characteristics of overwater scenes. In this paper, we propose an inverted dark channel Prior-Guided Cycle-consistent Generative Adversarial Network (PG-CycleGAN) for overwater image defogging. Specifically, an inverted dark channel prior map is designed to suppress the sky and highlight objects over the water. Building on this prior map, we develop a prior encoder to extract object-related features. Additionally, we propose a Prior-Guided Residual Block (PGRB) and a Prior-Guided TriUpsample (PGTU) module, which effectively integrate the extracted prior features for both feature encoding and upsampling. This integrated approach enhances the network's ability to accurately restore overwater objects, leading to improved defogging performance. Furthermore, we develop a prior map-guided GAN loss and a prior map-guided cycle-consistency loss, which guide the network to recover objects with greater fidelity while minimizing unnecessary restoration of the sky region. Through extensive experimental comparisons, our method demonstrates superior performance over existing state-of-the-art approaches in terms of qualitative analysis, quantitative metrics, and improvements in object detection.

This paper aims to investigate the wear evolution of contact interfaces and their impact on the nonlinear dynamics of anisotropic single-crystal blades. A modeling method for anisotropic blades is proposed, accompanied by a numerical simulation method that examines contact, friction, and wear behavior. Furthermore, a multi-scale dynamics solver is developed to analyze the behavior of anisotropic blades experiencing fretting wear. This method is applied to a shrouded single-crystal blade system, effectively predicting the wear behavior and its influence on the nonlinear response. Ultimately, the contributions of distinct anisotropic crystal orientations in single-crystal turbine blades to wear behavior and nonlinear dynamics are elucidated. It is determined that the nonlinear response, accounting for wear, is predominantly influenced by the angle beta.

Due to illumination variations and complex background interference, robust object detection in aerial imagery remains highly challenging. A single modality often fails to adapt to diverse scenarios, while the complementary properties of RGB and infrared (IR) modalities provide new opportunities for robust object detection. However, effectively exploiting their complementary information remains difficult. To address this issue, we propose a Thermal gradient prior and Illumination-aware Guided Network (TIGNet), which explicitly incorporates thermal gradient priors and illumination awareness into cross-modality feature fusion. Specifically, we design a Thermal Gradient-guided Cross-modality Interaction Module (TGCIM) to enhance geometric structure features and facilitate effective cross-modality interaction. Additionally, we introduce an Illumination-aware Cross-modality Fusion Module (ICFM), which generates a spatially varying weight map to adaptively balance the contributions of RGB and IR features under different illumination conditions. By combining these two modules, TIGNet achieves more flexible and robust feature fusion, effectively alleviating modality imbalance and illumination sensitivity. Experimental results on the DroneVehicle and RGBT-Tiny datasets demonstrate that the proposed TIGNet outperforms existing RGB-IR object detection approaches. The code is available at https://github.com/GrimreaperZyq/TIGNet.

Hepatocellular carcinoma (HCC) early identification is crucial for improving patient outcomes. Current screening methods are often complex and costly. This study developed a simplified, cost-effective HCC screening model using serum marker data. A diverse study population from two Chinese hospitals was recruited, including cancer patients, hospital patients, and healthy individuals. A two-stage screening model was created: LASSO logistic regression for preliminary screening, followed by logistic regression incorporating alpha-fetoprotein (AFP). The model's performance was evaluated in multiple cohorts. Across five populations, the model showed strong performance with AUC-ROC ranging from 0.868 to 0.907, accuracy between 87.43% and 96.96%, and sensitivity over 75% with specificity above 90%. Compared with solely AFP models, the second-stage model improved HCC risk estimates in healthy populations, with significantly higher AUC (0.930 vs. 0.827) and net reclassification improvement (NRI) up to 56.2%. This two-stage model offers a practical, cost-efficient tool for early HCC detection, addressing a significant public health need.

Multilabel pathological tissue segmentation is a vital task in computational pathology that aims to semantically segment different tissues within pathological images. Fully and weakly supervised models have demonstrated impressive performances in this regard. However, weakly supervised models still face challenges, such as the poor performance of nondominant samples and limited effectiveness of aggregation functions in conveying supervisory signals. To address these issues, we propose two key contributions: the introduction of a graph attention network(GAT) module to establish contextual relationships between pixels within patches and generate high-quality pseudo-labels, and the development of a novel global classified max pooling(GCMP) aggregation function that effectively transmits the supervision signal from weakly annotated labels and improves the model's classification accuracy. The experimental results show that our method improved the MIoU scores by 3.3 and 3 for the nondominant samples, necrosis(NEC) and lymphocytes(LYM), respectively, in the LUAD-HistoSeg test set. This led to an overall MIoU of 0.774, which is a 1.8 increase in the state-of-the-art(SOTA) performance. Similarly, our approach improved MIoU scores by 5.7 and 2 on the NEC and LYM samples, respectively, in the Breast Cancer Semantic Segmentation(BCSS) test set, resulting in an overall MIoU of 0.721. This represents a 1.6 increase in SOTA performance. In summary, our work addresses the issues of poor performance on nondominant samples and the suboptimal performance of aggregation functions. We propose a novel approach to achieve a significant performance improvement. This is extremely significant in reducing the workload of manual annotation and promoting the development of computational pathologies.

This paper aims to propose a time–frequency combination method to predict steady responses of a multi-freedom nonlinear vibration system with multi-interface connections. Firstly, the non-linear mathematical model of a multi-freedom was established, and a time–frequency combination method was proposed in the solving process, in which the non-linear force was obtained by the time-domain method, and the steady response was acquired through Incremental Harmonic Balance Method (IHBM). Then, the bolted cylindrical shell was selected as an object for the application. Considering the non-linear mechanical properties of the bolt-connection interface, we presented the non-linear dynamic model and solving method in detail. The validity and accuracy were verified. This shows the present method can be utilized to obtain steady responses of bolted shells considering the friction behaviors in the bolted joints. The accuracy of the results of vibration response was analyzed through the cases with different reserved harmonic orders in IHBM, and it indicates more harmonic should be considered for getting more accurate results. Also, in the time-domain progress, the friction force and locus figure at the position of the bolt can be observed. Finally, several cases were investigated. These analyses show the proposed method is appropriate to study the dynamics of the multi-freedom nonlinear system as the presented bolted cylindrical shell here.

Infrared ship images have low resolution and limited recognizable features, especially for small targets, leading to low accuracy and poor generalization of traditional detection methods. To address this, we design a prior knowledge auxiliary loss for leveraging the unique brightness distribution of infrared ship images, we construct a joint feature extraction module that sufficiently captures context awareness, channel differentiation, and global information, and then we propose a prior-knowledge- and joint-feature-extraction-based YOLO (PJ-YOLO) for use in detecting infrared ships. Additionally, a residual deformable attention module is designed to integrate multi-scale information, enhancing detail capture. Experimental results on the SFISD and InfiRray Ships datasets demonstrate that the proposed PJ-YOLO achieves state-of-the-art detection performance for infrared ship targets. In particular, PJ-YOLO achieves improvements of 1.6%, 5.0%, and 2.8% in mAP50, mAP75, and mAP50:95 on the SFISD dataset, respectively.

Maritime environments often face visibility challenges due to haze which significantly impacts detection models. However, existing maritime object detection algorithms often neglect haze conditions or the unique characteristics of the maritime environment, resulting in decreased effectiveness in hazy weather. In this article, we propose a prior knowledge-driven maritime image dehazing and object detection framework (MDD), which consists of a detection network and a restoration network. Leveraging the characteristics of the highlighted ships in the inverted dark channel prior (IDCP), the detection network incorporates a prior subnetwork to learn ship-related features, which are subsequently merged into the backbone network through an IDCP cross-attention module. During training, the restoration network is integrated to improve the clarity of the features learned by the detection network. In addition, a ship-haze enrichment strategy is implemented to emphasize ship regions in the training samples, along with a ship-aware reconstruction loss to enhance the network's ability to learn dehazed features. Moreover, we establish a maritime object recognition with haze levels (MORHL) data set to evaluate object detector performance in maritime hazy conditions. It includes 13 280 annotated images across six categories: cargo ship, container ship, fishing boat, passenger ship, island, and buoy, with haze levels categorized as light, medium, and heavy. Comprehensive experiments on the MORHL and SMD data sets demonstrate that the proposed MDD framework outperforms the state-of-the-art detectors and various combinations of dehazing and detection methods.

The magnetic levitation system is a typical non-self-stabilizing and strongly nonlinear system, subject to vibration phenomena due to changes in the operating point. This paper proposed combining nonlinear stiffness feedforward compensation control and deviation rectification control (FCC-DRC) to improve the system's vibration suppression control performance and load resistance. Firstly, using the kinetic decoupling method, the decentralized control of multiple points was decoupled to three degrees of freedom: z, alpha, and beta. Secondly, the nonlinear negative stiffness of the electromagnetic system was counteracted by introducing a feedforward compensation controller (FCC). Then, a deviation rectification controller (DRC) was designed to automatically correct the tilt angle of the platform caused by eccentric load. Finally, the control algorithm is coupled to the electromagnetic system to establish a dynamical coupling model of the magnetic levitation closed-loop system, which verifies that the FCC-DRC can provide alpha, beta-degreeof-freedom control forces while counteracting the nonlinear negative stiffness of the magnetic levitation system. The simulation and experimental results showed that the control system designed in this paper helped to reduce the amplitude of the first-order vibration of the system and effectively suppressed the modal vibration of the magnetic levitation platform. The system achieved offset-free tracking in the z degree of freedom and achieved the effect of vibration suppression. Under the action of load, the vibration period was shortened by 40 %, and the displacement variation in the Z-axis direction decreased by 27 %, indicating the system's enhanced stability.

Accurate prognostics of thermal boundaries are essential for improving the precision of temperature calculations in complex turbine rotor structures. However, existing methods often rely on extensive experimental data or demand substantial computational resources to determine thermal boundaries. This makes it difficult to balance computational efficiency and accuracy, especially when dealing with complex rotor geometries or fluctuating operating conditions. To address these challenges, this study proposes a feasible and efficient approach that integrates numerical, statistical, and iterative techniques to predict uncertain thermal boundaries in complex turbine rotors. Specifically, a precise finite element model of the assembled turbine rotor's temperature field is constructed using numerical methods. Subsequently, the temperature sensitivity of uncertain boundary parameters is evaluated by statistical methods. The key boundary parameters with high sensitivity are identified via the white shark optimizer. Based on these key parameters, the temperature distribution of the turbine rotor is predicted. Experimental results validate the high accuracy (error < 3 %) of this sensitivity-driven approach, highlighting its viability in practical scenarios where experimental data are limited but both efficiency and accuracy are essential.

A reliable system for dynamic prognostication and management of hepatocellular carcinoma (HCC) is urgently needed but currently unavailable. In our previous work, we developed a machine learning algorithm termed "Survival Path" (raw-SP) to enhance prognostication with longitudinal survival data. However, the previous model was limited to intermediate stage HCC patients, and it faced the risk of overfitting due to path proliferation. In this study, we developed a novel framework incorporating nodal fusion techniques to mitigate the risk of overfitting, and expanded the model's applicability to all stages of HCC patients. A post-fusion survival map (fusion-SP) containing 14 different paths was built, which demonstrated superior or non-inferior accuracy in dynamic prognosis prediction for HCC patients compared with raw-SP, as well as traditional staging systems within the first 15 months since initial diagnosis in large-scale derivation, internal and external validation cohorts. Subgroup analysis showed the fusion-SP demonstrated superior performance in dynamic prognostication compared to other models among patients with BCLC stage C disease and initial tumor burden above up-to-seven criteria. Under the framework of fusion-SP, novel and distinct optimal combination treatment strategies for advanced-stage HCC patients at different key nodes were uncovered, where traditional staging frameworks fall short. The fusion-SP framework could serve as a robust tool for facilitating dynamic prognosis prediction and treatment planning for HCC. Moreover, our streamlined methodology holds the potential to be applied across various types of cancers.

This paper develops a novel 3D brick element by Nodal Position Finite Element Method (NPFEM) to effectively model rotating solids. It uses nodal positions instead of nodal displacements to formulate element’s strain and kinetic energies. This approach effectively avoids computational errors caused by spurious strains induced by large rigid-body rotations and can automatically account for stiffening effects arising from centrifugal forces. By directly solving for the positions of rotating elastic solids using Hamiltonian canonical equations, the new 3D NPFEM brick element allows elastic deformation to be efficiently and accurately extracted by subtracting the rigid-body motion from these positions. Additionally, the ability of the new 3D NPFEM brick element to model bending deformation is enhanced by directly introducing incompatible modes into the element shape functions. Numerical validation shows that the new 3D NPFEM brick element accurately models and analyzes the elastic deformation of rotating blades. It automatically captures nonlinear frequency responses of rotating solids without requiring special boundary and loading condition treatments commonly used in classic FEM. This advancement offers significant advantages by avoiding errors when modeling complex rotating solids or machines, thereby improving computational efficiency and accuracy.

In real-world overwater scenarios, detecting occluded or distant objects is common challenges. In this paper, we initially construct a novel dataset SeaShips24790 for evaluating the performance of overwater object detectors, which includes 24,790 diverse overwater object annotations, especially focusing on small-scale objects. Subsequently, anew deep-learning network that integrates gated mechanism and complex-valued convolutions, termed IGC-Net, is proposed to tackle the challenges of object occlusion and small object detection in overwater scenarios. It employs the gating mechanism to selectively enhance or suppress features and incorporates complex-valued modules, including complex-valued convolutions, for fusing multi-scale feature maps. Additionally, a two-stage multi-scale feature fusion is used, comprising pre-fusion and post- fusion stages. Experimental results demonstrate that our proposed IGC-Net achieves state-of-the-art (SOTA) performance across several overwater object detection datasets. The SeaShips24790 dataset will be made available as requested.

BACKGROUND:To investigate the feasibility of detecting local recurrent nasopharyngeal carcinoma (rNPC) using unenhanced magnetic resonance images (MRI) and optimize a layered management strategy for follow-up with a deep learning model. METHODS:Deep learning models based on 3D DenseNet or ResNet frames using unique sequence (T1WI, T2WI, or T1WIC) or a combination of T1WI and T2WI sequences (T1_T2) were developed to detect local rNPC. A deep-learning-assisted recurrent NPC detecting simultaneous tactic (DARNDEST) utilized DenseNet was optimized by superimposing the T1WIC model over the T1_T2 model in a specific population. Diagnostic efficacy (accuracy, sensitivity, specificity) and examination cost of a single MR scan were compared among the conventional method, T1_T2 model, and DARNDEST using McNemar's Z test. RESULTS:No significant differences in overall accuracy, sensitivity, and specificity were found between the T1WIC model and T1WI, T2WI, or T1_T2 models in both test sets (all P > 0.0167). The DARNDEST had higher accuracy and sensitivity but lower specificity than the T1_T2 model in both the internal (accuracy, 85.91% vs. 84.99%; sensitivity, 90.36% vs. 84.26%; specificity, 82.20% vs. 85.59%) and external (accuracy, 86.14% vs. 84.16%; sensitivity, 90.32% vs. 84.95%; specificity, 82.57% vs. 83.49%) test sets. The cost of a single MR examination using DARNDEST was $330,724 (internal) and $328,971 (external) with a hypothetical cohort of 1,000 patients, relative to $313,250 of the T1_T2 model and $340,865 of the conventional method. CONCLUSIONS:Detecting local rNPC using unenhanced MRI with deep learning is feasible and DARNDEST-driven follow-up management is efficient and economic.

This study proposes a method to study blades with multiple friction interfaces, to reveal the coupling effect of multi-interface blades on nonlinear vibrations. The dynamics of a pre-twisted tenon jointed blade with under-platform dampers in a thermal environment is modeled, considering the combined effects of thermal effects and nonlinear loads. Subsequently, a friction model of the tenon and under-platform dampers contact interface is developed. While studying the coupling effect of multiple friction interface, the effects of natural frequency, temperature, rotational speed and pre-twisted angle on blade vibration characteristics and nonlinear characteristics are analyzed.

Maritime traffic community has paid a huge amount of focuses to establish maritime intelligent transportation infrastructure for the purpose of enhancing maritime traffic safety and efficiency. Maritime surveillance video is considered as a type of fundamental data sources for establishing intelligent maritime transportation infrastructure towards smart ship era. To that end, the study proposes an aggregated deep learning model-supported ship imaging trajectory extraction framework. The proposed framework starts by detecting ships from maritime images via a novel You Only Look Once (YOLO) model. More specifically, the proposed ship trajectory extraction framework obtains ship positions in a frame-by-frame manner via the proposed poly-YOLO module. Then, the proposed model maps ship positions in neighboring consecutive maritime images via an Enhanced Deep Sort (EDS) module. Experimental results suggest that the proposed ship trajectory extraction model achieves satisfactory performance due to that the average values of index multiple-object tracking accuracy (MOTa), recall rate (Rid) and index aggregated detection accuracy (Aggid) are larger than 89% (which outperform the comparison algorithms). The study can help varied maritime traffic participants obtain accurate on-site traffic situations in the smart ship era.

In this paper, an improved LuGre model was established based on the micro-convex assumption, Hertz contact theory, and thermal conditions. The displacement-tangential force and velocity-tangential force hysteresis curves under different temperature conditions were obtained by the dry friction testing experiment. Further, this paper constructed an objective function for the proposed friction model, identified the parameters through the experimental data based on an improved quantum genetic algorithm, and verified the effectiveness and superiority of the model identification results. Finally, this paper analyzed the matching degree between the model identification results and experimental data, confirmed the accuracy of parameter identification results with temperature changes, and put out the specific parameters of the improved model in different thermal environments. This study can provide theoretical guidance for accurately characterizing the friction characteristics of contact interfaces under different temperature conditions.

This paper proposes a method that combines the finite element method with an improved energy wear model to establish a three-dimensional (3D) ball-plane contact structures fretting wear model. The correctness of the model was verified by Hertz contact theory and fretting wear experiments. Then, the fretting wear dynamic behaviors of 3D ball-plane contact structures were investigated in detail. The results show that when the contact structure is in the partial slip regime (PSR), the cross-sectional wear scar profile changes from the “W” shape to the “U” shape along the fretting direction. When the contact structure is in the gross slip regime (GSR), the cross-sectional wear scar profile always maintains a “U” shape along the fretting direction. In addition, it was also found in the study that when the contact structure is at GSR under a low normal load or PSR under a high normal load, it will help reduce wear. This study can provide significant theoretical guidance for reducing and protecting fretting wear in contact structures.