Search Results (106)

Background: Missed fractures on initial imaging assessments remain a clinically significant source of diagnostic errors in orthopedic and trauma care. AI-assisted imaging tools are increasingly integrated into fracture detection workflows. However, their diagnostic benefits and safety vary substantially across anatomical regions, clinical contexts, and levels of reader experience. Purpose: To synthesize the current evidence on the diagnostic impact of AI-assisted fracture detection and to discuss evidence-informed principles for safe and selective clinical deployment. Methods: A structured narrative synthesis of meta-analyses, multi-reader, multi-case observer studies, and real-world implementation investigations was performed. Diagnostic performance patterns were examined across anatomical regions and levels of reader experience. No quantitative pooling or reanalysis of the primary data was performed. The findings were synthesized across anatomical regions, reader-experience groups, and implementation-relevant clinical contexts. Results: Across studies, AI-assisted interpretation was generally associated with moderate gains in sensitivity and lower missed-fracture rates compared with unaided human reading, while largely preserving specificity. The diagnostic benefit was greatest among less-experienced readers in high-volume emergency settings. Performance was strongly anatomy-dependent: consistent and clinically meaningful improvements were observed for hip and appendicular skeleton fractures; intermediate benefits with increased false-positive burden were reported for wrist and rib fractures; and inferior sensitivity relative to expert interpretation was documented for cervical and vertebral spine injuries. Conclusions: AI-assisted fracture detection improves diagnostic safety when implemented as a structured second-reader tool; however, its effectiveness depends heavily on anatomy. Available evidence supports selective, risk-stratified deployment, guided by anatomy-specific risk considerations and supervised clinical use, rather than indiscriminate or autonomous use, to maximize benefits and minimize patient safety risks in orthopedic and trauma imaging. Full article

Underground mine tunnels are often characterized by extremely uneven illumination, weak surface textures, and frequent dynamic interference, which severely undermine multi-view photometric consistency and easily induce floating artifacts and spatial divergence in conventional vision-based 3D Gaussian Splatting (3DGS). To address these issues, we propose a LiDAR-guided 3DGS framework for underground tunnel reconstruction based on dynamic-object removal and differentiable unsigned distance field (UDF) regularization. First, a dynamic foreground removal strategy with background restoration is introduced to remove transient foreground disturbances and restore static supervision consistency. Second, LiDAR point clouds are leveraged to initialize Gaussian primitives with a reliable geometric skeleton in weak-texture regions. More importantly, LiDAR priors are further converted into a differentiable UDF field and serve as a persistent geometric constraint. A dual-track mechanism is designed, where continuous geometric attraction pulls mildly deviated Gaussians back toward the physical surface and periodic out-of-bound culling removes severely drifting primitives. Experiments on real underground tunnel and chamber scenes show a clear scene-dependent behavior of the proposed method. In the tunnel scene, the method achieves the best SSIM together with competitive PSNR and LPIPS, while also reducing redundant out-of-bound primitives and improving geometric cleanliness. In the chamber scene, however, its advantages under global full-reference metrics are less evident. These results suggest that the proposed LiDAR-guided and differentiable UDF-regularized framework is particularly beneficial for weak-texture tunnel environments, while further improvement is still needed for chamber scenes with more complex appearance variations. Full article

Ankylosing spondylitis (AS) is a chronic immune-mediated inflammatory disease affecting the axial skeleton, characterized by progressive structural damage and functional impairment. Although biologic therapies targeting tumor necrosis factor and interleukin-17 have improved clinical outcomes, a substantial proportion of patients fail to achieve sustained disease control. Emerging evidence suggests that metabolic alterations may contribute to AS pathogenesis; however, systematic characterization of metabolism-related biomarkers and their regulatory networks remains limited, and the interplay between metabolic dysfunction and immune dysregulation in AS is poorly understood. Two whole-blood GEO datasets (GSE25101, GSE73754; n = 104) were integrated as the primary analytical cohort. A third dataset (GSE11886, n = 18; monocyte-derived macrophages) was included for exploratory cross-tissue analysis. Differential expression analysis identified 847 DEGs, which were refined to 16 metabolism-related genes through weighted gene co-expression network analysis (WGCNA) and GeneCards database filtering. Eleven machine learning algorithms with 5-fold cross-validation were applied to construct diagnostic models and identify hub genes. Validation analyses included immune cell infiltration estimation using CIBERSORT, metabolic pathway activity assessment via ssGSEA, single-cell transcriptomics from GSE268839, functional enrichment through GSEA/GSVA, and chromosomal localization analysis. A competing endogenous RNA (ceRNA) regulatory network was constructed to map post-transcriptional regulation. Natural compounds from 66 AS-treating traditional Chinese medicines were screened against hub genes using deep learning-based binding prediction. Multiple machine learning algorithms achieved comparable cross-validated performance (CV AUC range 0.741–0.836; top five models: 0.805–0.836) using the six hub genes (MFN2, SLC27A3, RHOB, SMG7, AKR1B1, LCOR) identified through SHAP-based feature importance analysis of the PLS model. Leave-one-dataset-out validation between the two whole-blood cohorts showed that all algorithms exceeded an AUC of 0.77 in Round 1 (validate: GSE73754, n = 72; best AUC 0.861), while Round 2 (validate: GSE25101, n = 32) yielded more modest performance (best AUC, 0.715) reflecting the smaller validation sample. Exploratory application to GSE11886 (macrophage-derived samples) showed near-chance performance, consistent with the tissue-source discrepancy. AS patients exhibited significant downregulation of oxidative phosphorylation, TCA cycle, and glycolysis pathways (p < 0.01), accompanied by elevated glutathione metabolism (p < 0.001). Immune cell deconvolution revealed reduced CD8+ T cell proportions correlating with MFN2 downregulation, and increased neutrophil frequencies correlating with SLC27A3 upregulation. Exploratory single-cell analysis indicated that RHOB expression was relatively enriched in border-associated macrophages and fibroblasts, while AKR1B1 was more prominently expressed in vascular endothelial cells and plasmacytoid dendritic cells. The ceRNA network identified 21 miRNAs and 65 lncRNAs forming 86 regulatory interactions, with four key regulatory axes (SATB1-AS1/miR-539-5p/LCOR, FAM95B1/miR-223-3p/RHOB, LINC01106/miR-106a-5p/MFN2, AATBC/miR-185-5p/SMG7) predicted to regulate hub gene expression. Compound screening identified betaine, pyruvic acid, citric acid, etc., as top-ranking candidates, with MFN2 showing the highest binding capacity among hub genes. This study provides an integrative framework linking metabolic reprogramming with immune dysfunction in AS. The six-gene diagnostic signature showed preliminary discriminatory ability in the available datasets, while the ceRNA regulatory network and natural compound screening results prioritize candidate regulatory pathways and compounds for future validation. These findings advance our understanding of AS pathogenesis and may guide future biomarker development and targeted intervention strategies. Full article

Over the past decade, an increasing number of functional microbial-like terpene synthases (MTPSLs) have been reported in non-seed plants. However, whether the traditional Chinese medicinal plant H. serrata harbors such enzymes and their corresponding functions remains unexplored. In this study, we mined the transcriptome of H. serrata and identified a microbial-like terpene synthase, HsMTPSL1, which produces multiple diterpene products. Following isolation and structural elucidation, seven distinct compounds were obtained, representing three skeletal types: spatane, prenylkelsoene-type, and biflorane. Among these, compound 7 is a novel biflorane diterpene. Structural analysis and subsequent mutagenesis revealed critical residues governing the formation of distinct skeletons, uncovering the multifunctional nature of this enzyme. Notably, the S224A mutation significantly enhanced the production of spatane diterpene compound 1 by 11.6-fold, demonstrating the potential for protein engineering to improve the yield of this bioactive marine-specific diterpene. Transcriptomic profiling revealed that HsMTPSL1 is highly expressed in sporangia, and co-expression analysis with cytochrome P450s identified the CYP781 subfamily as candidates potentially involved in the downstream modification of these skeletons. Collectively, we report the first MTPSL from H. serrata and characterize it as a multifunctional diterpene synthase. Through structure-guided mutagenesis, we uncovered the molecular basis of its functional versatility, with the S224A mutation providing a powerful tool for enhancing the yields of all three diterpene skeletons, thereby laying a foundation for future protein engineering and synthetic biology applications. Full article

Millimeter-wave (mmWave) radar point clouds offer a privacy-preserving solution for Human Activity Recognition (HAR), but their inherent sparsity and noise limit single-modal performance. While multimodal fusion mitigates this issue, existing methods often suffer from severe negative transfer during visual degradation and incur high computational costs, unsuitable for edge devices. To address these challenges, we propose Tac-Mamba, a lightweight cross-modal state space model. First, we introduce a topology-guided distillation scheme that uses a Spatial Mamba teacher to extract structural priors from visual skeletons. These priors are then explicitly distilled into a Point Transformer v3 (PTv3) radar student with a modality dropout strategy. We also developed a Trust-Aware Cross-Modal Attention (TACMA) module to prevent negative transfer. It evaluates the reliability of visual features through a SiLU-activated cross-modal bilinear interaction, smoothly degrading to a pure radar-driven fallback projection when visual inputs are corrupted. Finally, a Lightweight Temporal Mamba Block (LTMB) with a Zero-Parameter Cross-Gating (ZPCG) mechanism captures long-range kinematic dependencies with linear complexity. Experiments on the public MM-Fi dataset under strict cross-environment protocols demonstrate that Tac-Mamba achieves competitive accuracies of 95.37% (multimodal) and 87.54% (radar-only) with only 0.86M parameters and 1.89 ms inference latency. These results highlight the model’s exceptional robustness to modality missingness and its feasibility for edge deployment. Full article

This study explores the potential of biomimetic reasoning to inform early-stage design thinking, with a focus on enhancing the consideration of material utilization and textile waste. While sustainability efforts within the field of textiles are often focused on recycling and end-of-life management strategies, it is important to recognize that a substantial proportion of final waste-related outcomes are determined during the conceptual design stage and the initial prototyping iterations. This study investigates the potential of organizational principles derived from natural systems to inform the definition of problems, the generation of ideas, and early conceptual prototyping. This is achieved by the introduction of ecological constraints and material life-cycle awareness in conjunction with user-centered requirements. To address the conceptual gap between biological forms and manufacturing, biomimicry is approached as a mode of systemic reasoning, utilizing topological skeletonization as a tool for logic extraction rather than formal imitation, with emphasis placed on continuity, modularity, and adaptive organization. This computational proof-of-concept employs a Particle Swarm Optimization (PSO) framework, utilizing biological venation as a topological guide to demonstrate how distinct organizational logics influence pattern configuration while incorporating manufacturing-inspired constraints (such as path continuity and density) as optimization penalties. The findings are exploratory in nature and are confined to the computational domain; while the study utilizes proxy indicators to simulate potential textile behaviors, it acknowledges the lack of direct experimental validation of physical fabrication as a current limitation. By framing waste as an outcome of upstream design choices, this paper contributes a methodological perspective. This perspective places biomimetic design thinking as a reflective tool within sustainable and regenerative design practice. It also supports earlier engagement with ecological considerations in textile development. Full article

This study addresses a gap in the literature by explicitly linking responsive web design frameworks to concrete cybersecurity vulnerabilities, moving beyond traditional discussions of usability and device compatibility to incorporate security-by-design principles in contemporary frontend development. The research adopts a qualitative comparative approach and considers five widely used responsive design frameworks: Bootstrap, Tailwind CSS, Foundation, Pure CSS, and Skeleton. These frameworks were selected based on criteria such as maturity, adoption, and architectural diversity. Three research questions guide the analysis: the identification of cybersecurity risks associated with responsive design frameworks, the extent to which these risks vary across frameworks, and the mitigation strategies required to address them. The findings confirm that most critical vulnerabilities originate outside the frontend layer, reinforcing the separation between presentation and backend logic. However, the results demonstrate that frameworks significantly influence the security risk profile, particularly regarding cross-site scripting, dependency management, and configuration practices. Modern utility-first frameworks shift security concerns toward the build pipeline and toolchain, while minimalistic and abandoned frameworks introduce risks related to obsolescence and unpatched “forever-day” vulnerabilities. The study concludes that frontend security depends less on framework choice alone and more on governance, continuous maintenance, and the systematic adoption of secure development and DevSecOps practices. Full article

A lightweight composite that simultaneously exhibits high strength and excellent thermal insulation is of great interest for thermal protection applications. In this study, dimensionally stable needled quartz fiber felt-reinforced phenolic aerogel composites were prepared using vacuum impregnation, sol–gel, and ambient pressure drying. The composites exhibit a multiscale porous structure formed by interconnected nanometer polymer skeletons and micronscale fibers. By regulating the thermoplastic phenolic resin concentration in the precursor solution, the pore structure of the material was refined; the average particle diameter reduced from 99.76 nm to 38.91 nm, and the average pore diameter decreased from 216.79 nm to 49.53 nm. At a phenolic resin concentration of 25%, the composite exhibits outstanding thermal insulation and mechanical properties: a low thermal conductivity of 0.0646 W·m⁻¹·K⁻¹ at room temperature, with a mere 19.5 °C temperature rise on the sample backside after 1800 s heating at 200 °C, and compressive strengths of 7.70 MPa in the XY-direction and 3.87 MPa in the Z-direction (at 10% strain). X-ray micro-CT characterized the internal structural evolution during loading, revealing a failure mechanism dominated by fiber buckling. Theoretical models and experimental data were used to analyze and quantify the contribution rates of gas and solid heat conduction in NQF/PR aerogel composites, with solid conduction accounting for over 80%. Combined with microstructural evolution, the mechanism for the high thermal insulation efficiency of NQF/PR aerogel composites was elucidated. This study prepared NQF/PR aerogel composites with promising application potential. By systematically evaluating their compressive behavior and quantifying the respective contributions of gas and solid conduction, this work provides a methodological framework to guide the rational design of similar aerogel composites. Full article

Soil–steel interface shear governs load transfer and long-term serviceability in piles, retaining systems, and buried infrastructure; yet the large-displacement interface mechanics of fibre-reinforced sands remain poorly resolved, limiting sustainable design. This study couples large-displacement ring-shear testing with physics-guided hybrid AI to quantify and predict the peak and residual resistance of sand–polypropylene fibre mixtures sliding on smooth and rough steel. Two quartz sands with contrasting particle morphology were tested under 25–200 kPa normal stress and 0–1.0% fibre content, producing a design-oriented database that captures post-peak evolution and residual states. The experiments reveal a strongly nonlinear reinforcement law: an optimum fibre range enhances dilation, stabilises the shear band, suppresses post-peak softening, and increases residual strength, whereas excessive fibres disrupt the granular skeleton and reduce mobilisation efficiency. Roughness and confinement act as amplifiers, intensifying fibre-driven dilation and asperity interlock. To translate mechanisms into prediction, three strategies were benchmarked: a deep neural network (DNN), the Physics-Guided Neural Additive Model (PG-NAM++), and the physics-anchored Residual-DNN that learns only the correction to a mechanical baseline. Residual-DNN achieved the tightest agreement and the highest physical consistency for both peak and residual strength, enabling robust parameter selection with reduced uncertainty and overdesign. The combined experimental–AI framework advances the United Nations Sustainable Development Goals (SDGs) by supporting SDG 9 through resilient, innovation-led infrastructure design and contributing to SDG 12 by enabling optimised (rather than maximal) use and reuse of reinforcement materials within circular ground-improvement practice. Full article

Chondrosarcoma of the maxillofacial skeleton is a rare malignant tumor characterized by cartilaginous differentiation and locally invasive growth. Diagnosis is particularly challenging in low-grade tumors because histological features often overlap with those of benign chondroma. We describe a 62-year-old woman with a recurrent cartilaginous tumor of the hard palate. After previous resections in 2013 and 2022, a third recurrence was detected. MRI showed a lobulated lesion at the anterior hard palate contiguous with the nasal septum. A two-staged treatment was performed, starting with a minimal invasive access Brown class 2a maxillectomy guided by a patient-specific cutting guide. Pending histological confirmation, an obturator prosthesis was placed to seal the oroantral communication. Histopathology confirmed a low-grade chondrosarcoma with clear margins of at least 5 mm. A second-stage reconstruction was performed a year later using a posterior pedicle lateral nasal wall flap (inferior turbinate flap) and palatal rotation flap restored nasal lining and oral mucosa. This approach achieved oncologic clearance with excellent functional outcomes. The case highlights the value of image-guided maxillectomy and staged regional flap reconstruction. Full article

Analyzing three-dimensional (3D) phenotypic parameters of maize seedlings is of significant importance for maize cultivation and selection. However, existing methods often struggle to balance cost, efficiency, and accuracy, particularly when capturing the complex morphology of seedlings characterized by slender stems. To address these issues, this study proposes a novel end-to-end automated framework for extracting phenotypes using only consumer-grade RGB cameras. The pipeline initiates with Instant-NGP to rapidly reconstruct dense point clouds, establishing the 3D data foundation for phenotypic extraction. Subsequently, we formulate a directed topological graph-based mechanism. By mathematically defining bifurcation constraints via vector analysis, this mechanism guides a depth-first traversal strategy to explicitly disentangle stem and leaf skeletons. Building upon these decoupled skeletons, organ-level point cloud segmentation is achieved through constraint-based expansion, followed by density-based spatial clustering (DBSCAN) to detect individual leaves. Algorithms combining point cloud geometry with 3D Euclidean distance are also implemented to calculate key phenotypes including plant height and stem width. Finally, single-leaf skeleton fitting is used to estimate leaf length, and principal component analysis (PCA) is adopted to determine the stem–leaf angle, realizing the comprehensive automatic extraction of maize seedling phenotypes. Experiments show that the proposed method achieves high accuracy in extracting key phenotypic parameters. The mean relative errors for plant height, stem width, leaf length, stem-leaf angle, and leaf area are 0.76%, 2.93%, 1.26%, 2.13%, and 3.33%, respectively. Compared with existing methods as far as we know, the proposed method significantly improves extraction efficiency by reducing the processing time per plant to within 5 min while maintaining such high accuracy. Full article

Phase unwrapping is the decisive factor for achieving dimensional accuracy in phase-shifting profilometry, yet unavoidable phase jumps occur at discontinuities. Existing dual-frequency heterodyne techniques suffer from a narrow measurement range and overly coarse projected fringes due to grating superposition requirements, leading to large errors when scanning objects with hole-like features. To address these issues, this paper proposes an edge-oriented phase-unwrapping error-compensation method based on fringe projection phase profilometry. First, the wrapped phase of the measured object is acquired via phase-shifting profiling. The wrapped phase map is then smoothed at multiple scales using Gaussian filters, and parallel Canny edge detection combined with phase gradient thresholding is applied to comprehensively capture both coarse and fine discontinuities. Morphological closing fills in breakpoints, followed by skeleton thinning and connectivity reconstruction to generate an edge band of defined width. Within this band, edge-preserving smoothing is performed using guided filtering or bilateral filtering, and the result is fused with the original phase through Gaussian weighting based on the distance to the skeleton. Finally, an ordered multi-frequency heterodyne unwrapping restores the absolute phase, maximally preserving true discontinuities while effectively correcting noise and detection errors. Experiments show that this method overcomes edge-induced phase jumps—with jump-error correction rates exceeding 96.7%—exhibits strong noise resilience under various conditions, and achieves measurement precision better than 0.06 mm. Full article

Cancer metastasis, the spread of tumour cells from the primary site to distant organs, is responsible for over 90% of cancer deaths, yet effective treatments remain elusive due to incomplete understanding of the molecular drivers involved. Podocalyxin (PODXL), a protein overexpressed in many aggressive cancers, links the cell membrane to the internal skeleton through its interaction with Ezrin, an actin cytoskeleton cross-linker. Despite its therapeutic relevance, the PODXL–Ezrin interface remains structurally uncharacterised and pharmacologically intractable. Here, we employed an integrated computational approach combining protein–protein docking, molecular dynamics (MD) simulations, and virtual screening to investigate the structural basis of the PODXL–Ezrin interaction. Using AlphaFold-predicted structures, we modelled PODXL and Ezrin complexes, revealing that PODXL’s cytoplasmic domain stabilises upon Ezrin binding, with Arg495 mediating temporally distinct electrostatic interactions essential for initial complex assembly. Particularly, we characterised the R495W missense mutation in PODXL’s Ezrin-binding domain, demonstrating that substitution of arginine with bulky, hydrophobic tryptophan may allosterically destabilise Ezrin’s dormant conformation. This mutation slightly increases the intramolecular distance between the F3 subdomain and C-terminal domain from 2.59 Å to 3.40 Å, thus leading to potential partial unmasking of the Thr567 phosphorylation site that could plausibly prime Ezrin for activation. Molecular dynamics simulations in the WT state with a total simulation time of 100 ns revealed enhanced structural rigidity and reduced radius of gyration fluctuations in the mutant complex, consistent with a potential “locked,” activation-prone state that amplifies oncogenic signalling. Through virtual screening, we identified NSC305787 as a selective destabiliser of the R495W mutant complex by disrupting key Trp495–pre-C-terminal loop Ezrin interactions and causing steric hindrance to PIP2 recruitment. Our findings identified mutation-dependent changes in drug binding that can guide the development and repurposing of compounds for targeting PODXL-related cancers and improve patient outcomes in PODXL-altered malignancies. Full article

The craniofacial skeleton is a complex three-dimensional structure, and major reconstructive cases remain challenging. We describe a synergistic approach combining intra-operative navigation, three-dimensionally (3D) printed skull models, and mixed reality (MR) to improve predictability in surgical outcomes. A patient with previously repaired bilateral cleft lip and palate, significant midfacial retrusion, and a large maxillary alveolar gap underwent segmental Le Fort I osteotomy and advancement. Preoperative virtual planning was performed, and reference templates were uploaded onto MR glasses. Intra-operatively, the MR glasses projected the templates as holograms onto the patient’s skull, guiding osteotomy line marking and validating bony segment movement, which was confirmed with conventional navigation. The 3D-printed skull model facilitated dissection and removal of intervening bony spicules. Preoperative planning proceeded seamlessly across software platforms. Osteotomy lines marked with MR showed good concordance with conventional navigation, and final segment positioning was accurately validated. Postoperative outcomes were satisfactory, with re-established occlusion and closure of the maxillary alveolar gap. The combined use of conventional navigation, 3D-printed models, and MR is feasible and allows safe integration of MR into complex craniofacial reconstruction while further validation of the technology is ongoing. Full article

Federated learning (FL) for skeleton-based action recognition remains underexplored, particularly under strong client heterogeneity where regular FedAvg tends to cause client drift and unstable convergence. We introduce Clustered Federated Spatio-Temporal Graph Attention Networks (CF-STGAT), a clustered FL framework that leverages attention-derived spatio-temporal statistics from local STGAT models to dynamically group clients and perform attention-weighted inter-cluster fusion that gently align cluster models. Concretely, the server periodically extracts multi-head parameter-based attention descriptors, normalizes and projects them via PCA, and applies K-means to form clusters; a global reference is then computed by attention–similarity weighting and used to regularize each cluster model with a lightweight fusion step. On NTU RGB+D 60/120(NTU 60/120), CF-STGAT consistently outperforms strong FL baselines with the STGAT backbone, yielding absolute top-1 gains of +0.84/+4.09 (NTU 60, X-Sub/X-Setup) and +7.98/+4.18 (NTU 120, X-Sub/X-Setup) over FedAvg, alongside smoother per-client trajectories and lower terminal test loss. Ablations indicate that attention-guided clustering and inter-cluster fusion are complementary: clustering reduces within-group variance whereas fusion limits cross-cluster divergence. The approach keeps local training unchanged and adds only server-side statistics and clustering. Full article