Search Results (7,558)

Background/Objectives: In laparoscopic and robot-assisted surgery, bleeding may rapidly impair operative-field readability and procedural safety. In the broader Robin Heart teleoperation framework, interpretation of such events is relevant not only for scene understanding but also as a potential prerequisite for future safety-oriented supervisory functions under communication-degraded conditions. The aim of this study was to assess whether a deep learning model for blood segmentation could provide outputs suitable for preliminary image-level temporal characterization of visible blood-region behavior in laparoscopic video. Methods: A U-Net-based binary blood-segmentation model was implemented in-house in PyTorch and evaluated on three paired image–mask datasets: a simulated bleeding dataset prepared under controlled laboratory conditions, an internal operative laparoscopic dataset, and an external-domain subset derived from the public GynSurg dataset. Segmentation performance was assessed using 5-fold cross-validation and reported using the Dice coefficient and Intersection over Union (IoU). Training dynamics were analyzed using training and validation loss and Dice curves. Additional baseline comparisons were performed on the internal operative dataset using U-Net++ and DeepLabV3+. Temporal analysis was performed on selected video fragments, including a low-motion reference sequence without active bleeding progression, internal bleeding-related sequences, and external-domain sequences, using mask-derived descriptors and auxiliary optical-flow-based motion descriptors computed after camera-motion compensation within the detected blood-related ROI. Results: In 5-fold cross-validation, the U-Net-based model achieved Dice coefficient and IoU values of 0.915 ± 0.012 and 0.851 ± 0.019 on the simulated dataset, 0.856 ± 0.013 and 0.756 ± 0.025 on the internal operative dataset, and 0.707 ± 0.053 and 0.570 ± 0.056 on the external-domain GynSurg subset, respectively. On the internal operative dataset, the proposed model performed comparably to U-Net++ and slightly above DeepLabV3+ under the same cross-validation protocol. The temporal descriptor set differentiated low-motion reference behavior, more spatially coherent progression, rapid coherent expansion, and dynamic or motion-active progression profiles. Peak dA/dt reflected abrupt visible blood-area expansion, temporal IoU described mask stability over time, and optical-flow-based descriptors provided additional information on local motion activity within the detected blood-related ROI. Conclusions: The results support the feasibility of combining deep-learning-based blood segmentation with temporal and optical-flow-based descriptors for exploratory image-level characterization of visible blood-region behavior in laparoscopic video. Within the Robin Heart development pathway, such descriptors may, in the future, serve as candidate components of image-analysis support modules for safety-oriented teleoperative scenarios. At this stage, they should be interpreted as exploratory image-derived indicators rather than clinically validated markers of bleeding severity. Full article

Autonomous navigation in unstructured off-road environments presents fundamental challenges due to terrain heterogeneity, the absence of structured road markings, and the necessity for real-time traversability reasoning from raw sensory observations. We present TerrainFormer, a hierarchical framework that integrates a world model for terrain dynamics prediction with a temporal decision transformer for action selection. Our methodology employs a two-phase training paradigm: (1) self-supervised world model pretraining on LiDAR point clouds to learn terrain representations encompassing traversability, elevation, and semantic segmentation; (2) behavioral cloning of the decision transformer conditioned on frozen world model features with temporally derived goal directions. The world model processes raw 3D LiDAR point clouds through a PointPillars encoder for real-time bird’s-eye-view (BEV) projection, followed by a Vision Transformer backbone that produces latent terrain representations. A principal contribution is our cross-dataset generalization paradigm: the world model is trained on separate datasets while the decision transformer is trained on separate sequences, ensuring zero data overlap between training phases. We introduce automatic goal direction computation from vehicle pose trajectories, enabling the model to learn directionally conditioned navigation policies. To address the class imbalance inherent in off-road driving data, we employ focal loss with inverse-frequency class weighting and action-chunk supervision. Experimental evaluation on the RELLIS-3D dataset achieves 87.31% test accuracy with 0.7948 macro F1 across all 12 action classes. The world model’s predicted future frames produce only a 0.79% accuracy drop versus ground-truth observations, with 98.82% action agreement, demonstrating effective cross-dataset generalization for real-time off-road navigation. Full article

Parallel multi-blade centrifugal fans present a challenge in simultaneously reducing aerodynamic noise and maintaining efficiency. This study presents a multi-objective optimization using a radial basis function (RBF)-assisted Bayesian optimization framework, with three volute parameters (tongue radius, tongue clearance, and axial gap) as design variables. Computational fluid dynamics (CFD) combined with the Ffowcs Williams–Hawkings (FW-H) acoustic analogy was employed to evaluate noise and total pressure efficiency. To reduce computational cost, an RBF surrogate model was constructed from 30 Latin hypercube samples, achieving leave-one-out cross-validation (LOOCV) R² values of 0.978 and 0.995 for noise and efficiency, respectively. A Bayesian search using the log expected hypervolume improvement (logEHVI) acquisition function was performed on the RBF response surfaces, converging to a hypervolume of approximately 0.72, consistent with an NSGA-II benchmark. Based on household fan requirements, a 70/30 noise-efficiency weighting was adopted, yielding RBF-predicted values of 59.04 dB and 0.545 for the selected low-noise-preference candidate. An independent CFD recalculation yielded 59.19 dB and 0.554. The SPL at the characteristic frequency of 2550 Hz was reduced by 9.9 dB. Flow field analysis revealed that the optimized tongue clearance weakened the impingement on the volute tongue and suppressed unsteady vortex shedding. This framework provides an efficient strategy for multi-objective aerodynamic and acoustic optimization of parallel centrifugal fan systems. Full article

Understanding advanced pre-tertiary mathematics, particularly three-dimensional vectors, demands robust spatial reasoning skills that many students find challenging to develop through traditional pedagogical methods. This study proposes and evaluates an innovative educational tool that leverages large vision models to automate the conversion of handwritten vector equations into accurate 3D graphical representations. By interpreting students’ handwritten input using advanced computer vision, the system provides immediate, interactive visual feedback to bridge the cognitive gap between abstract symbolic notation and tangible geometric concepts. We evaluated the system using a dataset of 1000 handwritten vector equations typical of the Singapore-Cambridge GCE ‘A’ Level H2 Mathematics syllabus. Our findings demonstrate that while GPT-4o serves as a capable baseline, achieving 84.6% accuracy with multi-shot prompting, newer variants such as GPT-4.1-mini offer superior performance, reaching 91.4% accuracy with significantly higher computational efficiency. The results confirm that AI-powered visualisation tools can effectively interpret complex spatial mathematical layouts when guided by optimal prompt engineering. Implementing such technology in educational settings presents a viable, scalable, and cost-effective method to democratise learning support, fostering independent study and enhancing students’ conceptual comprehension of spatial mathematics. Full article

Background: Precise implant placement is crucial during inferior alveolar nerve (IAN) bypass in the posterior mandible where bone height above the IAN is limited. This in vitro study compared the accuracy of static computer-assisted implant surgery (sCAIS) and dynamic computer-assisted implant surgery (dCAIS) for implant placement during IAN bypass. Methods: Two cone-beam computed tomography (CBCT) mandibular models with deficient bone height (<7 mm) above the IAN canal, classified as clinical scenario I and clinical scenario II, were used as an in vitro setting. Thirty models per clinical scenario were prepared, after which 60 dental implants were placed in the edentulous area of tooth no. 47. Software-based analysis compared planned and actual implant placements by postoperative CBCT. The two models were compared for deviation in distance to the inferior alveolar nerve (DIAN), entry-3D deviation, entry-2D deviation, apex-3D deviation, apex-vertical deviation, and angular deviation by comparative statistical analysis. Results: Both sCAIS and dCAIS showed less deviation from planned implant position in both scenarios. No statistically significant differences were detected except for angular deviation (sCAIS: 1.73° vs. dCAIS: 1.19°, p = 0.004), including clinical scenario I (sCAIS: 1.65° vs. dCAIS: 1.19°, p = 0.033) and II (sCAIS: 1.98° vs. dCAIS: 1.2°, p = 0.033). Conclusions: Both approaches showed minor deviation in both IAN bypass models, while dCAIS showed better angular control, requiring future in vitro and in vivo research in complex clinical environments. Full article

Rapid and non-contact surface inspection is essential for quality control in modern production. Optical coherence tomography (OCT) can image a surface without contact, but turning those images into roughness parameters usually requires specialized processing software. This study examined whether a multimodal large language model (LLM) could estimate roughness parameters directly from OCT B-scans as a screening tool. The study was designed as a controlled macro-scale proof of concept using periodic, analytically defined phantoms rather than as validation on stochastic industrial micro-roughness. Five test surfaces with exactly known geometries were designed, 3D-printed, and scanned with a spectral-domain OCT system. For each surface, roughness values were computed from the theoretical shape, extracted from the OCT image using MATLAB, and also estimated by the LLM from the same image. The repeatability of the LLM was checked by running the same prompt ten times per surface. On a sawtooth profile, the LLM estimates varied by 3.8% for Ra, 4.2% for Rq, 3.5% for Rp, 2.8% for Rv, and 3.1% for Rt. Across all five surfaces, the variation in Ra and Rq was around 3–5%, and for Rt, it stayed below 5%. The results show that a generative AI approach can produce repeatable roughness estimates that are useful for comparative screening. This method offers a flexible option for surface comparison and AI-assisted quality control when calibrated measurements are not required. Full article

Aortic stenosis is predominantly treated through transcatheter bioprosthetic heart valve implantation. However, the materials used in these devices are prone to premature failure. Polymer heart valves provide an alternative to current commercial devices, offering materials with greater durability and customisation through fibre reinforcement. Given the wide range of available materials and structures, there is a need for a systematic and efficient approach to designing and optimising novel bioinspired polymeric leaflets. This work presents a framework that employs computational modelling and Design of Experiments (DOE) tools to optimise bioinspired, 3D-printed, fibre-reinforced polymer leaflets made using melt electrowriting (MEW). Here, finite element (FE) models are created to represent MEW fibre-reinforced polymer leaflets for application in a transcatheter aortic heart valve. The behaviour of this valve under physiological loading conditions is modelled to predict valve performance and leaflet material response. These models were first used to investigate the impact of fibre orientation on valve performance and leaflet response, thereby demonstrating the benefits of a bioinspired fibre reinforcement structure. Using a DOE approach, the structural combination of MEW fibre reinforcement and an elastomeric matrix was optimised to improve valve performance and reduce leaflet stress and strain. Overall, the framework offers an efficient and versatile methodology for optimising fibre-reinforced polymer leaflets using an in silico approach, thereby reducing the need for physical prototyping and testing of these next-generation devices during early product development. Full article

The processing of two-dimensional (2D) spectral images constitutes a critical and multifaceted discipline in contemporary astronomical data analysis. As spectroscopic instruments evolve towards higher multiplexing, resolution, and sensitivity, the raw 2D data captured by detectors present increasingly complex challenges that transcend simple one-dimensional extraction. This review provides a systematic and comprehensive examination of the methodological evolution in this field over the past two decades. It gathered relevant studies by searching mainstream academic repositories and general search engines with the core keyword ‘2D Spectral Image’, and selected qualified references according to accessibility and research relevance. We categorize the landscape into three major paradigms: (1) physics-based modeling and algorithmic correction techniques for geometric distortion, scattered light, and sky background; (2) data-driven machine learning and deep learning approaches for image correction, spectral classification, and faint signal detection; and (3) the development of open-source software pipelines that democratize advanced processing. A central contribution of this review is a detailed comparative analysis of the performance metrics, underlying assumptions, and practical limitations of prominent algorithms. We highlight the transformative impact of convolutional neural networks (CNNs) and vision transformers (ViTs) on tasks such as celestial object classification and exoplanet detection, while also acknowledging the enduring importance of robust physical models for calibration and uncertainty quantification. The discussion culminates in an assessment of persistent challenges—including computational scalability, model generalizability, and interpretability—and outlines promising future directions at the intersection of AI, statistical inference, and large-scale survey science. Full article

Complex networks have become a fundamental paradigm for modeling real-world systems. Synchronization of such networks, particularly under hyperchaotic dynamics, presents a significant control challenge due to the high-dimensional state space and multiple positive Lyapunov exponents. This paper addresses the driver node selection problem in a 4D Hyperchaotic Lorenz complex network, formulating it as a constrained binary optimization task. We evaluate a pool of advanced metaheuristics, including the quantum genetic algorithm (QGA), seahorse optimizer (SHO), and artificial bee colony (ABC), across multiple network experiments conducted over 30 independent runs to guarantee statistical validity. The performance of these solvers is rigorously benchmarked against traditional topological heuristics, a random selection baseline comprising 600 feasible configurations, and verified through Wilcoxon statistical testing. Furthermore, addressing computational sustainability, we introduce a “Green-Artificial Intelligence” architecture based on dual-tier structured query language memoization (SQL-memoization) and provide a detailed runtime comparison evaluating its efficiency. The empirical results indicate that swarm-intelligence methods such as ABC and SHO exhibit robust competitive performance in minimizing synchronization errors while the Green-AI framework consistently and drastically reduces the computation of the repetitive simulations. Full article

This study presents a comprehensive computational evaluation of radiofrequency (RF) ablation efficacy and the spatial formation of thermal ablation zones within a 3D model of a liver tumor. By systematically comparing these configurations, the study aims to elucidate the physical mechanisms governing electromagnetic (EM) energy dissipation in hepatic tissue and to provide clear engineering guidelines for optimizing RF applicator selection and treatment planning in clinical practice. To reliably simulate the biophysical phenomena of the RF ablation procedure, a coupled electro-thermal model based on the finite element method and the Pennes bioheat equation was implemented. The research investigates six distinct applicator variants: conventional needle-type applicators and advanced expandable umbrella-type RF applicators equipped with four- and eight-tine electrodes, each evaluated in both monopolar and bipolar configurations. Numerical simulations were conducted for a standard 10 min ablation procedure at varying applied voltages to assess the specific absorption rate (SAR) distribution, transient heating dynamics, and the exact volumes of the resulting coagulation necrosis which were quantified using rigorous isotherms and the cumulative equivalent minutes at 43 °C (CEM₄₃) thermal dose index. Volumetric analysis of the ablation zones revealed that bipolar multi-tine electrodes induce highly localized heat concentration. Conversely, monopolar multi-tine setups strongly disperse EM energy. The results demonstrated that, for conventional needle applicators, the monopolar configuration generated significantly larger necrosis zones than the bipolar operating mode. The RF applicator geometry and its operating mode directly dictate the spatial extent of liver tissue necrosis. Moreover, advanced numerical treatment planning is essential for optimizing SAR and CEM₄₃ distributions and ensuring safe and complete hepatocellular carcinoma eradication. Full article

The selective detection of small, highly hydrophilic metabolites differing only in stereochemistry represents a major challenge in biosensor development. Here, we present a computational investigation to elucidate the molecular basis of the experimentally observed selectivity of a protein-based electrochemical biosensor toward myo-inositol over D-chiro-inositol. Although the two stereoisomers differ only in the orientation of a single hydroxyl group, they induce distinct dynamic effects on the protein recognition element. Molecular docking revealed comparable binding regions and similar affinity scores, indicating that selectivity does not arise from differences in binding site or docking energy. To investigate dynamic contributions, all-atom molecular dynamics simulations were performed in triplicate (3 × 100 ns) using the AMBER99SB force field and explicit TIP3P water. Trajectory analyses showed that myo-inositol forms a more persistent hydrogen bond network, resulting in reduced residue-level flexibility, more stable ligand–protein interactions, and enhanced local structural stabilization. Overall, these findings support a dynamic model of stereoselective recognition in which ligand-induced modulation of protein conformational ensembles, rather than static affinity, governs biosensor performance. This work highlights the value of molecular dynamics simulations in the rational design of biosensors targeting structurally similar analytes. Full article

Background: Coronary bifurcation lesions represent one of the most technically demanding scenarios in coronary artery disease (CAD), associated with higher procedural complexity, restenosis, and periprocedural complications. Recent advances in coronary computed tomography angiography (CCTA) have markedly improved its ability to visualize complex coronary anatomy, assess plaque morphology, and guide revascularization. Objectives: This review summarizes (1) technological advances in CCTA over the last decade, (2) its role in evaluating bifurcation stenosis, (3) assessment of plaque morphology and distribution, (4) quantification of bifurcation geometry, and (5) emerging evidence supporting its application in revascularization planning and guidance. Findings: Modern wide-detector and dual-source CT systems, iterative and deep-learning reconstruction algorithms, and photon-counting CT (PCCT) have significantly improved temporal and spatial resolution, reduced blooming artifacts, and lowered radiation dose. CCTA now reliably quantifies bifurcation stenosis and plaque distribution, characterizes high-risk plaque features, and accurately measures bifurcation angles. The integration of CT-derived fractional flow reserve (FFR-CT) and artificial intelligence (AI)-based plaque quantification further strengthens its diagnostic and prognostic performance. CCTA-derived bifurcation scores and 3D modelling support procedural strategy selection, stent sizing, and side-branch (SB) protection. Conclusions: CCTA has evolved into a comprehensive tool for non-invasive diagnosis, physiological assessment, and pre-procedural planning of bifurcation disease. With the advent of PCCT and AI-enhanced quantitative tools, CCTA is poised to become a central component of revascularization decision-making in complex coronary bifurcations. Full article

High dynamic range (HDR) image reconstruction from a single low dynamic range (LDR) input remains an important problem for computational photography, particularly when practical deployment on consumer-grade hardware is considered. With the increasing availability of hardware supporting HDR, public demand for capturing and viewing HDR images has grown significantly. Recent research has explored deep learning-based approaches to reconstruct HDR images from low dynamic range (LDR) inputs by extracting regional pixel features or leveraging the camera response function (CRF) for model training. Many of these approaches employ Convolutional Neural Network (CNN) architectures and utilize skip connections to preserve learned information. Nevertheless, the configuration-level effects of data augmentation in HDR reconstruction remain insufficiently discussed. Existing CNN-based approaches, such as HDRCNN, HDRUNet, and ExpandNet, have demonstrated promising reconstruction ability, but they may involve a heavy backbone architecture, a long training time, or a limited discussion of how preprocessing configurations affect reconstruction performance. This study presents an engineering-oriented HDR reconstruction framework derived from HDRCNN, focusing on practical efficiency, structural fidelity, and training feasibility. The proposed framework introduces three modifications: (1) a configuration-level comparison of composite data augmentation settings, including unsharp masking, denoising, Gaussian blur, and brightness–contrast adjustment; (2) the replacement of the original VGG16 backbone with a ResNet50-based encoder enhanced with attention blocks and squeeze-and-excitation (SE) blocks for improved multi-scale feature extraction and channel-wise recalibration; and (3) the integration of mixed-precision training with cosine annealing learning-rate scheduling to reduce computational cost. Experimental results on the SI-HDR dataset show that the best composite augmentation configuration improves PSNR from 19.05 dB to 22.10 dB and SSIM from 0.6444 to 0.7714 without increasing the training time. Compared with the original VGG16-based HDRCNN setting, the ResNet50-based model reduces training time while improving SSIM from 0.2705 to 0.8512. Under the adopted comparison protocol, the proposed model achieves the shortest training time and slightly higher PSNR than HDRUNet, while HDRUNet retains a higher SSIM. This indicates a trade-off among pixel-wise fidelity, structural similarity, and computational efficiency. The current evaluation is limited by a small test setting, composite rather than operation-level augmentation analysis, and the use of PSNR and SSIM only; therefore, future work should include full benchmark evaluation, additional perceptual/HDR-specific metrics, and controlled component-level ablation studies. Full article

The steel surface defect detection is crucial for steel quality and usage safety. The high computational cost and low detection accuracy are still the main issues in current steel detection models. To efficiently address the issues above, this paper proposes a new steel surface defect detection model named DIDW-YOLOv11. In the proposed DIDW-YOLOv11, the YOLOv11 C3k2 module is first innovatively improved by C3K2-DIMB, which integrates C3K2 and DIMB by introducing DynamicInceptionDWConv2d (DIDW) to sufficiently strengthen the detailed feature extraction for tiny defects and weak-texture defects, improving the matching degree of multi-scale receptive fields. Then the YOLOv11 SPPF module is enhanced by integrating the IDWFSPPF module for optimizing the fusion of local and global information, which combines average pooling and max pooling to enhance the model’s multi-scale feature fusion capability. An auxiliary detection head (ADH) is finally proposed with an additional coarse loss function to process shallow feature information into the model, which uses extra supervision for shallow features to suppress background noise and reduce false detections. Experimental results on the NEU-DET and GC10-DET datasets show that DIDW-YOLOv11 achieves 4.9% and 3.8% improvements in mAP@0.5 compared to the baseline model YOLOv11s. Our research indicates that DIDW-YOLOv11 exhibits stronger recognition ability and robustness in complex and diverse defect detection, providing an effective solution for steel defect detection in industrial production. In addition, experimental results show that our model offers improved performance over the baseline methods. Full article

Traditional 3D bioprinting of corneal constructs relies on planar slicing, which often results in a significant stairstep effect and the loss of anatomical curvature. Curvilinear layering has emerged as a promising alternative to address these limitations. The presented methodology, based on non-planar layer integration, ensures a smoother surface finish. The model’s surface is identified via vertex normals and reconstructed using the Poisson method. Finally, surface parametrization is applied to generate spatially curved trajectories. To validate the algorithm, corneal constructs were printed using a planar and the proposed non-planar approach. Quantitative evaluation of micro-Computed Tomography data revealed that the non-planar approach achieved significantly higher morphological fidelity, successfully replicating the intended parabolic profile of the human cornea. Furthermore, the non-planar constructs demonstrated adequate functional performance, characterized by high optical transparency. Thereby, the feasibility of printing non-planar layers using the proposed novel approach is successfully demonstrated. Furthermore, the comparative analysis confirms the method’s potential for corneal biofabrication when compared to traditional planar methods. Full article