Search Results (340)

Adolescence is a developmental stage characterised by identity formation, increasing independence, and a strong need for belonging. For adolescents in foster care, these developmental tasks are often complicated by unresolved trauma arising from abuse, neglect, abandonment, or the loss of primary caregivers. Frequent placement disruptions and difficulty establishing secure attachments with foster parents may further exacerbate emotional distress and risk behaviours. Despite these challenges, little research has examined foster parents’ experiences of caring for adolescents exhibiting risk behaviours. This qualitative study explores foster parents’ perspectives on the challenges they encounter, the coping strategies they employ, and the support systems they utilise while caring for adolescents exhibiting risk behaviours. Face-to-face semi-structured in-depth interviews were conducted in English and Setswana with eight purposively selected foster parents receiving services at the Joe Morolong Local Municipality in the Northern Cape Province, South Africa. Conducting the interviews in the participants’ home languages enhanced rapport and generated richer, more nuanced data. Data collection continued until saturation was reached. The participants described managing a range of behavioural challenges, including truancy, absconding, aggression, defiance, dishonesty, stealing, association with negative peer groups, and refusal to attend to household responsibilities. These behaviours contributed to considerable emotional strain and caregiving stress. Foster parents also reported inadequate professional and psychosocial support, leading to feelings of isolation, being overwhelmed, and burnout. Although compassion and cultural values strengthened resilience, caregiving capacity was constrained by poverty, unemployment, and limited support services. The findings underscore the need for strengthened supervision of foster care placements and sustained professional support, including counselling and caregiver support groups, to enhance foster parents’ well-being, improve coping capacity, and promote stable foster care placements for adolescents. Full article

Anomaly detection in industrial automated visual inspection (AVI) is constrained by severe data sparsity and class imbalance, as defect samples are both scarce and morphologically heterogeneous in real production environments. Existing context-aware synthesis frameworks, such as CASDA, alleviate this limitation by modeling defect–background relationships. However, they depend on domain-specific handcrafted rules and manually constructed compatibility matrices that must be redesigned for each new dataset, limiting scalability. We propose AROMA (adaptive ROI-based morphology-aware augmentation), a framework whose central contribution is the automatic, data-driven re-derivation of the defect–background compatibility matrix that prior work constructed by hand. AROMA profiles defect-morphology and background-context distributions and, from their patch-level co-occurrence, learns a symmetric compatibility gate that scores placement locations without any per-dataset hand-tuning, automatically replacing the handcrafted compatibility matrices required by previous approaches. It further summarizes each dataset’s difficulty with a morphology complexity index (MCI) and a context complexity index (CCI); the CCI, in particular, characterizes when context-aware placement can help. To evaluate the placement mechanism independently of generative-model quality, we employ a training-free copy–paste compositing engine with gradient-domain or alpha blending, driven by this data-driven symmetric compatibility gate. Experiments on five industrial datasets spanning textiles (AITeX), metal commutators (Kolektor), steel (Severstal), magnetic tiles (MTD), and leather (MVTec) show that the effectiveness of context-aware ROI selection is not universal but jointly depends on two conditions: sufficient baseline headroom and adequate background-context complexity (CCI). Performance gains are consistently observed on datasets satisfying both conditions (AITeX: +9.4 pp mAP@0.5 over baseline and +2.8 pp over random placement; Severstal: +1.6 pp and +1.3 pp), whereas the advantage narrows to parity—without reversing—on datasets with near-ceiling baseline accuracy (Kolektor, MTD) or largely homogeneous backgrounds that make placement position inconsequential (MVTec Leather, where both placement strategies gain equally, about +9 pp). These findings indicate that context-aware ROI selection is beneficial when both improvement headroom and contextual diversity are present and is otherwise neutral rather than harmful, providing practical guidance on when adaptive ROI optimization is expected to outperform random placement. Full article

The structurally complex Cangdong fault zone controls the medium- to low-temperature geothermal potential of the Dongli Lake field in Tianjin, China. Previous assessments in Chinese sedimentary basins predominantly employed simplified volumetric methods that cannot capture transient system behavior under production–reinjection conditions, largely due to the scarcity of systematic dynamic monitoring data. This study addresses this gap by exploiting relatively abundant dynamic monitoring records in the study area, integrating borehole stratigraphy, geological cross-sections, and well logging data to construct a 3D geological model and coupled numerical model of the fault-controlled reservoir. Meticulous natural-state calibration against temperature and pressure logging monitoring data yields key reservoir parameters, achieving mean relative errors of 2.35–3.33% for temperature and 6.32–9.73% for pressure. These results are based on calibration against wells DL-19 and DL-40, which intersect the Jxw reservoir at depths of 1850–2300 m. This process quantifies thermal conductivity and reveals pronounced permeability anisotropy in the primary Jixian Wumishan (Jxw) reservoir, with horizontal permeability (3.00 × 10⁻¹³ m²) an order of magnitude higher than vertical (3.00 × 10⁻¹⁴ m²), directly evidencing the fault zone’s hydraulic dominance. Dynamic sensitivity analysis demonstrates that optimal well placement must target high-permeability zones adjacent to the main fault, and adaptive reinjection protocols are proposed for pressure maintenance and thermal drawdown mitigation. This research provides a calibrated, data-driven framework and scientifically grounded strategies for long-term sustainability of the Dongli Lake resource, with transferable implications for optimizing development in other fault-controlled sedimentary basin systems. Full article

Implementing innovative digital teaching methods in higher education is a particularly complex endeavor that calls for adaptive and hybrid implementation strategies. This study describes the integration of digital simulations into a higher education nursing faculty. Kotter’s change model was used as an organizational framework to structure and report the process, while a design approach guided day-to-day practice. This combination was chosen to balance accountability with the flexibility required in a complex higher education context. This four-year single case study draws on annual project reports and descriptive student questionnaire data. The findings show that neither the problem definition nor the solution remained stable across project cycles: each cycle surfaced new challenges and revealed new contextual constraints. Tangible outcomes include a growing library of co-created simulation scenarios, a range of implementation tools, and the “simulatie wasstraat”, a curated arrangement of scenarios tailored to students’ placement contexts. Students reported positive experiences regarding perceived usefulness, motivation, and the perceived educational value of the simulations. Persistent challenges around organizational embedding remain. The findings nonetheless suggest that a hybrid implementation strategy was practically viable in this institutional context. Full article

Background: Flow diverters (FDs) are widely used to treat intracranial aneurysms. Although effective, FD-related thrombosis or progressive stenosis may lead to acute or delayed cerebral hypoperfusion with significant neurological morbidity. Management options are limited when medical or endovascular rescue strategies fail. This study aims to evaluate the role of extracranial–intracranial bypass (EC-IC bypass) as a rescue strategy across the spectrum of FD-related cerebral hypoperfusion. Methods: We retrospectively reviewed all patients who underwent EC-IC bypass for FD-related cerebral hypoperfusion over a 10-year period in our institution. Both acute and chronic presentations were included. Clinical status was assessed using the modified Rankin Scale (mRS). Imaging evaluation included digital subtraction angiography (DSA), computed tomography angiography (CTA), magnetic resonance angiography (MRA) and quantitative magnetic resonance imaging for bypass flow assessment. Surgical technique, antiplatelet therapy, timing of intervention and clinical outcomes were analysed. Results: Five patients with FD-related cerebral hypoperfusion who underwent superficial temporal artery–middle cerebral artery bypass (STA-MCA bypass) were identified. Two distinct patterns emerged. Three patients developed early hypoperfusion within the first month of FD placement, presenting with acute neurological deficits and treated by emergent or urgent bypass under dual antiplatelet therapy. Two patients developed late hypoperfusion years after the procedure, presenting with progressive haemodynamic symptoms and treated by elective flow-augmentation bypass under aspirin monotherapy. At discharge, only one patient had an mRS ≥ 3. At one-year follow-up, a favourable outcome (mRS ≤ 2) was observed in all patients with available data. Follow-up data were missing for one patient. Conclusions: EC-IC bypass represents a feasible rescue strategy for FD-related cerebral hypoperfusion when medical or endovascular treatments fail. In our experience, though limited, STA-MCA bypass can provide sufficient cerebral reperfusion, even in urgent settings and under dual antiplatelet therapy, leading to favourable neurological outcomes. Full article

Objectives: To evaluate temporal trends in ocular biometrics, patient demographics and surgical strategies for intraocular lens (IOL) localization in pediatric cataracts. Methods: This retrospective study included 291 eyes of 148 pediatric patients operated at a tertiary referral center between 2019 and 2025, comprising 219 operated eyes and 72 unaffected fellow eyes serving as intra-individual controls. Eyes were stratified into three chronological cohorts (2019–2020, 2021–2022, 2023–2025). Axial length (AL), white-to-white (WTW) diameter and central corneal thickness were analyzed in relation to the anatomical IOL placement. A generalized linear mixed model with a random intercept per patient accounted for the within-patient correlation between fellow eyes. Results: Mean AL increased from 18.91 ± 2.83 mm to 20.32 ± 3.19 mm (p = 0.004), whereas the increase in WTW diameter was not significant (p = 0.450). The age at surgery increased significantly across the periods (median 1.08, 1.33 and 3.13 years; p < 0.001), a general temporal effect that was not modified by the site of implantation (p = 0.32). Underlying etiology strongly influenced the approach: all eyes with Marfan syndrome received an iris-fixated IOL, whereas 90% of uveitic eyes remained aphakic. Femtosecond laser-assisted cataract surgery (FLACS) was used in 13 eyes and confined to the most recent period. Conclusions: Pediatric cataract surgery shows a progressive increase in the age at surgery that is not specific to any IOL localization, while the surgical approach is frequently determined by the underlying etiology. The increasing use of FLACS represents an institutional trend; the small number of cases and absence of outcome data preclude conclusions regarding its safety or feasibility. Full article

Carbon fiber-reinforced polymer (CFRP) composites are increasingly used in aerospace, rail transportation, and energy engineering owing to their high specific strength and corrosion resistance. However, their complex and interacting damage mechanisms, including delamination and matrix cracking, present significant challenges for reliable structural health monitoring. Fiber Bragg grating (FBG) sensors offer distinct advantages for monitoring composite structures because of their compact size, immunity to electromagnetic interference, embeddability, and capability for distributed strain measurement. Nevertheless, the effectiveness of an FBG sensing network depends strongly on the spatial distribution of the sensing points. This study proposes a finite-element-assisted framework for evaluating and improving FBG sensor layouts for strain-field reconstruction and structural feature characterization of composite plates. The framework first reconstructs the spatial strain field from limited sensing data using interpolation and least-squares fitting methods, and then evaluates the performance of existing and candidate sensor layouts based on reconstruction errors and spatial coverage of structurally important regions. A strain-gradient-informed heuristic strategy is subsequently developed to improve sensor placement by combining high-gradient region identification, spatially uniform coverage, minimum-distance constraints, and predefined support-region monitoring requirements. The Fourier least-squares fitting method provides the lowest reconstruction error among the investigated approaches and is therefore adopted for subsequent layout evaluation and improvement. Finite-element simulations and experimental measurements are used to assess the reconstruction performance and identify the advantages and limitations of different sensor layouts under static and dynamic loading conditions. The results demonstrate that the proposed framework can effectively evaluate existing FBG layouts and provide a systematic basis for their improvement, while also revealing the trade-off between local strain-gradient resolution and global spatial coverage. The proposed framework provides practical guidance for the performance-oriented design and improvement of FBG sensor networks for structural health monitoring of composite structures. Full article

This paper presents a unified technical survey of Large Language Models (LLMs), connecting three layers of the modelling pipeline that existing surveys address in isolation: internal architecture, distributed training optimisation, and downstream adaptation. Its organising principle is the dependency between these layers—how a choice at one constrains what remains feasible at the next. The survey examines fundamental mechanisms (tokenisation, scaled dot-product attention, activation functions, and normalisation, including RMSNorm and pre- versus post-normalisation placement), then the engineering of training at scale: data, tensor, and pipeline parallelism, hybrid schemes, mixed-precision training with BF16 and FP8, ZeRO-Offload memory management, activation checkpointing, and compute-optimal scaling laws together with the conditions under which they fail. The adaptation section covers supervised and instruction fine-tuning, a comparison of parameter-efficient methods (LoRA, QLoRA, adapters, prefix and prompt tuning), Reinforcement Learning from Human Feedback with its reward-hacking failure mode, alternatives including DPO, KTO and Constitutional AI, Retrieval-Augmented Generation beyond the basic pipeline, and decoding strategies. Practical configuration guidance is given for 7B, 70B and trillion-parameter regimes. Dedicated treatments of Mixture-of-Experts architectures, long-context modelling, and hardware-aware co-design close the survey, with open challenges classified by origin and severity. Full article

The family Limoniidae is the most species-rich within the order Diptera. The classification of this family is still unresolved and is based largely on morphological studies and the opinions of a few specialists. The rapid advancement of molecular techniques has created new opportunities to address long-standing systematic challenges. Nevertheless, relatively few studies have used molecular data to refine our understanding of Limoniidae and its relationships with allied lineages. In the present study, we reviewed publications that applied molecular approaches to improve the classification of Limoniidae and clarify its relationships with related groups. Relevant studies were identified through the literature section of the Catalogue of the Craneflies of the World using the keywords “classification”, “systematics”, and “molecular”. We retained only studies that employed molecular data to investigate the classification of Limoniidae, the phylogenetic position of the family within Tipuloidea, and the placement of Tipulomorpha within the dipteran tree. The studies reviewed consistently support the conclusion that Limoniidae is not monophyletic. In contrast, Tipulomorpha is consistently recovered as a monophyletic lineage, although its position within the dipteran tree remains controversial. Consequently, a satisfactory taxonomic framework has yet to be achieved. Recent advances in high-throughput sequencing and the generation of large-scale molecular datasets provide unprecedented opportunities to resolve these systematic uncertainties. We therefore advocate expanding both taxon sampling and the number of genetic markers included in phylogenetic analyses. Ultimately, a coordinated effort among research groups and institutions is likely to represent the most effective strategy for resolving the taxonomy and evolutionary relationships of Limoniidae. Full article

Objectives: To describe the surgical setup, docking strategy, and perioperative outcomes of retroperitoneal renal surgery performed with the Hugo™ RAS system, reporting, to our knowledge, the first dedicated surgical series. Methods: Between February and August 2025, 30 consecutive patients underwent retroperitoneal renal surgery with the Hugo™ RAS system at a single institution. Procedures included robot-assisted partial nephrectomy (RAPN, n = 15) and robot-assisted pyeloplasty (n = 15), performed by a single experienced surgeon. A standardized retroperitoneal access and reproducible trocar configuration were adopted. A three-arm configuration with predefined docking and tilt angles was used. Feasibility was defined as completion of the procedure without conversion to open or transperitoneal surgery. Perioperative data were prospectively collected and reported as median and interquartile range (IQR). Results: All procedures were completed without conversion or additional trocar placement. In the RAPN cohort, median docking time was 5 min (IQR 5–5), operative time was 180 min (IQR 172–200), and estimated blood loss was 200 mL (IQR 150–250). Three patients (20%) experienced Clavien–Dindo grade II complications; no major complications occurred. Positive surgical margins were observed in 1 patient (6%), with no radiological recurrence at 6-month follow-up. In the pyeloplasty cohort, operative time was 150 min (IQR 147–180) and estimated blood loss was 70 mL (IQR 50–100). One patient (7%) developed a grade II complication. Renal function remained stable in both groups. Conclusions: Retroperitoneal renal surgery using the Hugo™ RAS system is feasible, safe, and reproducible when supported by a standardized docking and port configuration. Full article

This document outlines the quality standards, experimental configurations, and precise workflows utilized to execute a metallic powder bed fusion process for creating aerospace test pieces. Using AlSi10Mg alloy, validation specimens were produced on a TRUMPF TruPrint 2000 machine according to a dedicated commissioning strategy. Metallographic evaluations targeted grain dimensions, morphology, boundaries, and phase compositions like precipitates or carbides. Findings showed scattered precipitates averaging 4 µm across the cross-sections. Subsequent X-ray microtomography (XCT) assessed porosity levels plus defect geometry and placement, offering comprehensive data on specific inclusions alongside broader statistics. XCT detected internal pore sizes ranging from 0.02 to 0.9 mm and rare clusters of high-density particles with a diameter of approximately 0.03 mm. Tensile evaluations of two batches showed a 72 GPa longitudinal elasticity modulus, 240 MPa offset yield strength, 350 MPa ultimate tensile strength, and 7% failure elongation. These positive results finalized a successful commissioning phase, enabling the subsequent design of lightweight aerospace mechanical assemblies. Full article

Yuxian paper cutting is a distinctive regional form within UNESCO-recognized Chinese paper cutting. Its opera figure images remain difficult to organize at scale because interpretation and cataloguing rely heavily on manual expertise, while automated tools for preliminary annotation and retrieval remain limited. This study formulates a domain-specific object detection task and constructs a curated dataset of 926 original images containing 976 annotated instances across three operational visual categories: armor-clad, robe-wearing, and female-role figures. A YOLOv8n-based detector is adapted by integrating the Spatial and Channel Synergistic Attention (SCSA) module and the Inner-IoU regression strategy. On the test set, the resulting configuration achieves 95.8% precision, 92.9% recall, 98.8% mAP@0.5, and 89.6% mAP@0.5:0.95, exceeding the YOLOv8n baseline by 3.0, 3.6, 2.9, and 2.9 percentage points, respectively. It also achieves a forward-pass throughput of 92.3 FPS under the reported GPU configuration. Among the evaluated configurations, the combined SCSA placement with Inner-IoU achieved the highest overall detection performance under the current dataset and experimental protocol. The results support its use for preliminary visual-category annotation, retrieval, and organization of Yuxian paper-cut opera figure resources. However, the findings are based on a small domain-specific dataset and require further validation across alternative data partitions and broader heritage collections. Full article

Introduction: Mediastinal tumors in children are rare and heterogeneous, and their surgical management is challenging due to limited working space and proximity to critical structures. While thoracoscopy has expanded, technical limitations persist in complex mediastinal dissections. Robot-assisted surgery may address these limitations, but its role in pediatric mediastinal tumor resection is not well defined. This systematic review aims to synthesize the available evidence on robot-assisted resection of mediastinal tumors in children, focusing on patient selection, technical feasibility, perioperative outcomes, and available oncologic results. Methods: A systematic review was conducted in accordance with PRISMA 2020 guidelines and registered in PROSPERO (CRD420251245437). PubMed/MEDLINE, Web of Science, Scopus, and the Cochrane Library were searched up to 1 December 2025. Eligible studies included pediatric patients (0–18 years) undergoing robot-assisted resection of mediastinal tumors. Study selection, data extraction, and risk of bias assessment were performed independently by two reviewers. Due to heterogeneity and lack of comparative data, a qualitative narrative synthesis was undertaken. Results: Fifteen studies were included, comprising retrospective case series and case reports, including 215 children, 149 (69.3%) of whom were derived from a single retrospective study. The most frequent indications were localized neurogenic tumors, predominantly in the posterior mediastinum. Robot-assisted surgery was consistently performed with curative intent. The da Vinci platform was used in all cases, with reproducible port-placement strategies adapted to tumor location and patient size. Conversion to open surgery occurred in 11 (5.1%) cases, mainly due to bleeding or limited exposure, with no emergency conversions reported. Reported perioperative complications were uncommon, although complication reporting was heterogeneous across studies. No perioperative deaths were reported. Two disease-related deaths occurred during follow-up in patients with recurrent Ewing sarcoma. Long-term oncologic outcomes could not be adequately assessed because of heterogeneous reporting and limited follow-up. Conclusions: Robot-assisted resection of mediastinal tumors in children appears feasible in carefully selected cases treated at experienced centers. Current evidence mainly supports its use for localized tumors with favorable anatomy, particularly in the posterior mediastinum, while long-term oncologic adequacy remains insufficiently defined. Full article

The course of altering “wind flow energy” into mechanical energy by means of two counter-rotating wind helms is considered in this article. Mechanical energy is then transformed into electrical energy in a generator where the carcase and the inductance switch are in reverse order. This counter-rotation principle increases the relative speed between the magnetic field and the armature, thereby improving the overall efficiency of energy conversion. The key purpose of this investigation is to advance an effective power source system for independent customers grounded on a specifically planned wind power enterprise with an improved wind energy use constant. The emphasis of the paper is to fix the best strategy parameters of the wind turbine by performing mathematical modelling of the electricity generation process based on the counter-rotation principle and the developed physical prototype. In order to evaluate the key performance parameters for efficient power supply to autonomous consumers, agriculture experimental studies were conducted on the prototype. The key project and techno-economic dials comprised the optimisation of wind power enterprise data, wind helm extents, their comparative placement, and producer control production as purposes of the predictable wind rapidity. The practical implementation of the proposed design was demonstrated by patents granted for a wind generator with variable rotor torque (Patent of the Republic of Kazakhstan No. 36903; Eurasian Patent No. 047230). The results demonstrated that the developed wind turbine achieved higher energy efficiency compared to the conventional single-rotor systems. Thus, the proposed design is suitable for applications in alternative energy and decentralised power supply for remote agrarian customers. Full article

Learning-based multi-view 3D reconstruction pipelines, such as transformer-based approaches, enable the accurate reconstruction of 3D scenes from multiple images, but their deployment across the edge–cloud continuum is challenging due to high computational demands and large intermediate data transfers. Effective pipeline scheduling in the continuum must therefore balance latency constraints with the cost of cloud resource usage. In this work, we address cost-aware scheduling under latency constraints for a multi-stage 3D reconstruction pipeline consisting of depth estimation, transformer-based multi-view fusion, and point cloud merging with export to a rendering-ready representation. We implement a service-oriented pipeline where each stage can be executed either on edge or cloud nodes, and we experimentally characterize its performance on representative hardware platforms. The results show a strong imbalance between the computational time and communication latency across platforms, mainly due to large intermediate data. Based on these insights, we propose an online scheduler that dynamically selects stage placements to minimize the cloud cost while satisfying latency constraints. The scheduler incorporates a top-K edge selection mechanism that reduces the decision complexity by jointly considering the network conditions and node utilization. Simulation results parameterized with real-system measurements show that the proposed approach effectively reduces cloud usage while meeting latency constraints, outperforming the baseline strategies based on single-node pipeline execution. Full article