Search Results (7,142)

Fractures as critical discontinuous structural planes in rock masses, directly govern their stability and serve as the core controlling factor in rock mechanics engineering. Existing deep learning models for fracture extraction face persistent challenges, including imbalanced integration of deep and shallow features, limited suppression of background noise, inadequate multi-scale feature representation, and large parameter sizes—making it difficult to strike a balance between detection accuracy and deployment efficiency. Focusing on the Wanshanshan quarry in Yunnan, this study first constructs a high-precision digital model using close-range photogrammetry and 3D real-scene reconstruction. A lightweight yet high-accuracy intelligent detection method, termed MSEM-Deeplabv3+, is then proposed for rock mass fracture extraction. The model adopts lightweight MobileNetV2 as the backbone network, incorporating inverted residual modules and depthwise separable convolutions, resulting in a parameter size of only 6.02 MB and FLOPs of 30.170 G—substantially reducing computational overhead. Furthermore, the proposed MAGF (Multi-Scale Attention Gated Fusion) and SCSA (Spatial-Channel Synergistic Attention) modules are integrated to enhance the representation of fracture details and semantic consistency while effectively suppressing multi-source and multi-scale background interference. Experimental results demonstrate that the proposed model achieves an mPA of 89.69%, mIoU of 83.71%, F1-Score of 90.41%, and Kappa coefficient of 80.81%, outperforming the classic Deeplabv3+ model by 5.81%, 6.18%, 4.53%, and 9.2%, respectively. It also significantly surpasses benchmark models such as U-Net and HRNet. The method accurately captures fine and continuous fracture details, preserves the spatial distribution of long-range continuous fractures, and maintains robust performance on the CFD cross-scene dataset, showcasing strong adaptability and generalization capability. This approach effectively mitigates the risks associated with manual high-altitude inspections and provides a lightweight, high-precision, non-contact intelligent solution for fracture detection in high-steep rock slopes. Full article

The Gulf of Taranto (Ionian Sea) is a key transitional sector between the Southern Apennines collisional belt and the Calabrian Arc system, where the expression of Pleistocene–Holocene deformation in the shallow stratigraphic record remains debated. This study focuses on the Taranto Canyon area, the main morphologic feature of the northeastern Gulf of Taranto slope. We integrate high-resolution multibeam bathymetry (10 m grid) with Sparker seismic profiles to (i) define the shallow seismo-stratigraphic framework and (ii) document spatial relationships between shallow discontinuities, morphostructural lineaments, and submarine channel network organization. A simplified tie to the Livia 001 well constrains the subdivision of the shallow succession into four seismic units: the late Pleistocene–Holocene unit (PtH), the Santerno Formation (SNT), the Calcarenite di Gravina (GRA), and the Cupello Limestones (CPL). The PtH interval shows the strongest lateral variability and includes widespread acoustically disturbed bodies and recurrent sub-vertical fluid escape acoustic anomalies. Steep discontinuities producing reflector terminations, minor vertical separation, and localized bending affect PtH and, locally, SNT, with normal fault geometries prevailing where resolvable. Bathymetric mapping reveals multiple lineament families and preferred channel orientations that persist across higher Strahler orders, supporting a structurally conditioned template that guides seafloor morphology, sediment routing, and canyon–slope evolution in the northeastern Gulf of Taranto. Full article

Satellite observations of methane are frequently compromised by extensive data gaps caused by cloud cover and aerosol contamination, limiting their utility for continuous regional monitoring. To reconstruct these spatiotemporal discontinuities, this study developed the Stacked Ensemble Reconstruction Framework for Methane (SERF-XCH₄). By integrating Sentinel-5P TROPOMI retrievals with 25 multi-source environmental covariates, we generated a spatiotemporally continuous, high-resolution (0.1°) monthly dataset (SERF-XCH₄-IM) for Inner Mongolia spanning 2019 to 2023. Comprehensive validation demonstrates that the framework achieves exceptional predictive fidelity with a Coefficient of Determination ($R^2$) of 0.93 and a Root Mean Square Error (RMSE) of 7.89 ppb, significantly surpassing the performance of individual base learners and traditional interpolation methods. Furthermore, spatial block cross-validation confirmed robust generalization capabilities ($R^2=0.90$) in data-void regions. To unravel the “black box” of the model, SHapley Additive exPlanations (SHAP) analysis was employed, revealing that temporal factors (contributing 63.9%), air temperature, and elevation are the dominant drivers governing XCH₄ variability. Spatiotemporal analysis further identified the Hulunbuir region as a significant growth “hotspot” with an annual increase rate exceeding 18.5 ppb/yr, a trend primarily driven by intensified emissions during the autumn and winter seasons. Consequently, this framework establishes a high-precision, interpretable paradigm for regional methane monitoring and geo-information reconstruction. Full article

This study proposes a 300 W class Gallium Nitride (GaN) Solid-State Power Amplifier (SSPA)-based microwave plasma generator system for implementing next-generation light sources with high brightness and color rendering at 2.45 GHz. To overcome the lifetime limitations and control constraints of conventional magnetron systems, the proposed system introduces custom packaging technology utilizing GaN-on-SiC Bare-dies fabricated via the Win-semiconductor’s NP25 process. This approach minimizes parasitic components and significantly reduces thermal resistance compared to standard packages, ensuring reliability during high-power operation. A stable RF output of 300 W was achieved through two-stage power combining. For the plasma source, an Ar-InBr-Hg gas mixture was employed to optimize optical characteristics. This gas mixture is commonly used in electrodeless plasma lamps due to its high luminous efficacy and stable discharge characteristics. To analyze the rapid impedance discontinuity during gas ignition, numerical analysis based on the Drude model was performed, theoretically identifying the complex permittivity transition of the medium and the resulting resonant frequency up-shift mechanism. To mitigate system instability during this transition, an adaptive frequency tracking and feedback control loop based on real-time VSWR monitoring was implemented. Experimental results demonstrate precise tracking within a 100 MHz frequency variable range, achieving a system efficiency of over 53% and maintaining a VSWR below 1.15:1. These results validate the practical feasibility of GaN SSPA technology in electrodeless lighting and industrial plasma applications utilizing high-power RF energy. Full article

Background: The optimal sequencing of regorafenib and trifluridine/tipiracil (FTD/TPI) in refractory metastatic colorectal cancer (mCRC) remains uncertain, particularly in Asian populations. Methods: We retrospectively analyzed 110 patients with mCRC who sequentially received both agents between 2011 and 2025. Patients were categorized into regorafenib followed by FTD/TPI (Rego → FTD/TPI, n = 88) and FTD/TPI followed by regorafenib (FTD/TPI → Rego, n = 22). Co-primary endpoints were time to treatment discontinuation (TTD) and overall survival (OS). Propensity score-based weighting methods, including stabilized inverse probability of treatment weighting (primary analysis), were used to adjust for baseline imbalances. Multivariable Cox regression was performed as a sensitivity analysis. Results: No statistically significant differences were observed between treatment sequences. In the primary analysis, the hazard ratio (HR) for TTD was 1.01 (95% CI 0.71–1.43), and for OS was 1.19 (95% CI 0.67–2.12), with FTD/TPI → Rego as reference. Median TTD was 6.8 versus 8.9 months, and median OS was 14.6 versus 20.2 months, respectively. Conclusions: Clinical outcomes were comparable regardless of treatment order, supporting individualized sequencing decisions in refractory mCRC. Full article

Objectives: To describe the efficacy and safety of tocilizumab (TCZ) and rituximab (RTX) in real-world patients with systemic sclerosis (SSc), across different clinical phenotypes and lines of therapy, evaluating both global clinical outcomes and lung function. Methods: SSc patients treated with TCZ (n = 27) or RTX (n = 23) were retrospectively followed for 12–24 months. Clinical measures, including modified Rodnan Skin Score (mRSS), C-reactive protein (CRP), and revised EUSTAR activity index 2017 (RAI), as well as spirometric parameters, were recorded at baseline and 6, 12, and 24 months. Statistical methods for repeated measures were applied to investigate outcome trends. Given the baseline differences, between-group comparisons were considered exploratory. Results: RTX was used earlier in disease course, while TCZ was mainly used as a rescue therapy. In both groups, mRSS, CRP levels and RAI significantly decreased over time. RTX-treated patients showed a greater absolute mRSS improvement, in line with higher baseline skin scores. No treatment discontinuations due to adverse events occurred in either group; one death and one discontinuation due to inefficacy were observed in the TCZ group. Among SSc-ILD patients, FVC% showed a modest decline in both groups, while DLCO% remained overall stable, and only a few patients met the OMERACT criteria for functional progression. Conclusions: In this real-world, single-center cohort of SSc patients, both agents were associated with a positive impact on disease activity, with a low rate of lung progression and with favorable safety profiles. Owing to substantial baseline imbalances and confounding by indication, between-group comparisons do not allow firm conclusions on comparative effectiveness. Overall, these data support the use of both agents in different clinical scenarios. Full article

Concerns about weight gain are commonly cited with combined oral contraceptive pill (COCP) use, yet it remains unclear whether perceived weight gain as a reason for discontinuation differs by body mass index (BMI). We analyzed data from the 2017–2019 National Survey of Family Growth (NSFG), including 3709 non-pregnant women aged 20–49 years who had ever used COCPs and had BMI calculated from self-reported height and weight. Trained NSFG staff interviewed participants on reasons for discontinuation and coded them into predefined categories, including weight gain. Discontinuation was examined by BMI category (underweight, normal weight, overweight, obesity) using survey-weighted logistic regression adjusted for demographic and socioeconomic covariates. Overall, 35.2% (95% CI 32.3–38.1%) of women reported discontinuing COCPs due to dissatisfaction, with 20.2% (95% CI 18.1–22.3%) citing side effects. Weight gain was reported by 7.0% (95% CI 5.6–8.4%) of ever-users, with higher prevalence among women with overweight (8.4%) and obesity (7.7%) compared with normal-weight women (5.5%). In adjusted analyses, women with overweight (aOR 1.76, p = 0.048) and obesity (aOR 1.68, p = 0.033) had higher odds of COCP discontinuation due to self-reported weight gain. These findings highlight the importance of addressing weight-related concerns during contraceptive counseling, particularly for women with higher BMI. Full article

Synthetic Aperture Radar (SAR) enables all-weather maritime surveillance, yet ship-oriented bounding box (OBB) detection remains challenging in complex scenes. Strong sea clutter and dense harbor scatterers often mask the slender characteristics of ships as well as the weak responses of small ships. Meanwhile, the periodicity of angle parameterization introduces regression discontinuities, and near-symmetric, bright-scatterer-dominated signatures further cause heading ambiguity, undermining the stability of orientation prediction. Moreover, in most detectors, multi-scale feature fusion and angle estimation lack explicit coordination, and rotated-box localization performance is often jointly affected by feature degradation and unstable orientation prediction. To this end, we propose a unified framework that simultaneously strengthens multi-scale representations and stabilizes orientation modeling. Specifically, we design a Gaussian-Guided Stage-Aware Deformable Feature Pyramid Network (GSDFPN) and a Coarse-to-Fine Unit-Circle Resolver (CF-UCR). GSDFPN enhances multi-scale fusion with two plug-in components: (i) a Gaussian-guided High-level Semantic Refinement Module (GHSRM) that suppresses clutter-dominated semantics while strengthening ship-responsive cues, and (ii) a Stage-aware Deformable Fusion Module (SDFM) for low-level features, which disentangles channels into a geometry-preserving spatial stream and a clutter-resistant semantic stream, and couples them via deformable interaction with bidirectional cross-stream gating to better capture the inherent slender characteristics of ships and localize small ships. For orientation, CF-UCR decomposes angle prediction into direction-cluster classification and intra-cluster residual regression on the unit circle, effectively mitigating periodicity-induced discontinuities and stabilizing rotated-box estimation. On SSDD+ and RSDD, our method achieves AP/AP${}_{50}$/AP${}_{75}$ of 0.5390/0.9345/0.4529 and 0.4895/0.9210/0.4712, respectively, while reaching AP${}_{s75}$/AP${}_{m75}$/AP${}_{l75}$ of 0.5614/0.8300/0.8392 and 0.4986/0.8163/0.8934, evidencing strong rotated-box localization across target scales in complex maritime scenes. Full article

To address the range-angle coupling issue of frequency diverse array (FDA), a beamforming method based on subarray partitioning is proposed, with a focus on analyzing uniform continuous and nonuniform discontinuous subarray structures. Based on the transmit–receive signal model established to solve the time-varying issue of FDA, two subarray partitioning methods under the same array aperture are investigated. In the case of uniform continuous subarray structure, when different linear frequency offsets (FOs) are applied to each subarray, the mainlobe width in range dimension is 4.35 km, and the peak sidelobe level (PSLL) is −7.25 dB. When nonlinear FOs are applied, the mainlobe width is reduced to 2.76 km, and the PSLL is decreased to −9.64 dB. Furthermore, by adopting a nonuniform discontinuous subarray structure combined with nonlinear FOs, the mainlobe width is further narrowed to 1.29 km, and the PSLL is reduced to −11.75 dB. The simulation results demonstrate that under the same conditions, the nonuniform discontinuous subarray structure significantly improves range resolution and effectively suppresses sidelobe. Based on above results, a joint optimization combining the bat algorithm (BA) and K-means++ clustering algorithm is proposed to optimize the subarray structure and element amplitudes simultaneously. Finally, the mainlobe width of the optimized FDA is 1.18 km and the PSLL is −12.32 dB. Simulation results confirm the effectiveness and potential of the proposed method in enhancing range resolution and achieving a focused beampattern. Full article

Introduction: Acute gallstone pancreatitis is a potentially severe disease associated with morbidity and mortality. Cholecystectomy is recommended to prevent recurrence. During pregnancy, surgical management is challenging, and in the post-partum period small gallstones may spontaneously disappear. Ursodeoxycholic acid (UDCA) is safe during the last 6 months of pregnancy and effective in dissolving small gallstones, although recurrence after discontinuation is common in the general population. The optimal strategy to prevent recurrent acute pancreatitis during and after pregnancy remains unclear. Methods: Between 2002 and 2017 we prospectively treated women with acute pancreatitis related to small gallstones (≤1 cm in diameter) during the last six months of pregnancy or within the first post-partum year who declined surgery. Patients received UDCA until stone dissolution. A patent cystic duct was confirmed by ultrasonography; after delivery, a non-contrast CT scan was performed to exclude calcified stones. Patients were followed for at least 6 years or until recurrence, with serial clinical and ultrasonographic examinations. Results: UDCA was associated with complete dissolution in 13/14 women within a mean ± SD of 7.77 + 3.1 months. One patient experienced gallstone recurrence 75 months after treatment discontinuation. Two patients developed recurrent pancreatitis (at 1 and 88 months respectively). Twelve women remained free of recurrence over a mean ± SD follow-up of 79.5 + 9.4 months. Discussion: This is an observational study in which we document that UDCA may facilitate the spontaneous dissolution of small gallstones after delivery and can be considered a bridge strategy during pregnancy when surgery is not feasible. However, this study cannot determine the additional benefit of UDCA over the spontaneous disappearance of stones observed after delivery because we had no control group. Cholecystectomy remains the standard of care post-partum. Medical therapy should be reserved for women who refuse surgery and it requires close ultrasonographic surveillance. The main strength of this study is the prospective long-term follow-up of a consecutive cohort with a rare condition. Limitations include the small sample size, missing control group and single-center design. Full article

Computed tomography (CT) enables non-destructive insight into internal log structure, yet fully automated interpretation of CT images remains limited by inconsistent annotations, boundary ambiguity, and insufficient spatial context in 2D slice-based analysis. These challenges restrict the industrial deployment of deep learning for wood quality assessment. This study applies artificial intelligence (AI) and deep learning to the automated analysis of computed tomography (CT) scans of wood logs for detecting internal qualitative features and segmenting bark. Using convolutional neural networks (CNNs), trained models accurately distinguish healthy and damaged regions and segment bark, including discontinuous parts. We introduce a novel pseudo-spatial representation by merging consecutive slices into red–green–blue (RGB) format, which improves prediction accuracy and model robustness across logs. To enhance interpretability, Gradient-weighted Class Activation Mapping (Grad-CAM) highlights regions contributing most to defect detection, particularly knots. Comprehensive evaluation using Sørensen–Dice similarity coefficients and confusion matrices confirms the effectiveness of the proposed approach under industrial conditions. These findings demonstrate that AI-driven CT image analysis can address key limitations of current log-grading workflows and enable more reliable, objective, and scalable quality assessment for timber-dependent economies. Full article

Background: Among individuals with type 2 diabetes (T2D), lifestyle improvements can restore glycemic control, yet few studies have examined deprescribing in settings where it was necessitated by improvements in health. This study aimed to (1) identify instances of medication deprescribing among adults with T2D in a primary care setting where patients had access to lifestyle medicine (LM), (2) document lifestyle changes among deprescribed patients, (3) assess changes in body mass index (BMI), glucose, and hemoglobin A1c (HbA1c) following deprescribing, and (4) assess the safety of deprescribing in the context of LM-informed care by identifying adverse events. Methods: A retrospective review of electronic health records (EHR) was conducted among 650 adults with a diagnosis of T2D per ICD-10 code at two primary care practices. To be included in the study, individuals had to be seen at least two times during the study period, from 2014 to 2023. Using a previously developed deprescribing framework, records were reviewed to identify deprescribing events. Among patients who were identified as deprescribed, BMI, glucose, and HbA1c, were extracted from the EHR, and age-, sex-, and time-adjusted differences in least squares means were calculated. Mentions of lifestyle change in provider notes in the EHR were also extracted pre- vs. post-deprescribing. Results: Forty-one deprescribing events were confirmed, totaling 6.3% of the study population. The most common medication changes included metformin dose reduction 34%, metformin discontinuation 19.5%, and insulin dose reduction 19.5%. Among patients with follow-up data, mean BMI decreased by 2.25 kg/m², p = 0.0003. Mean decreases of 25% in glucose and 13% in HbA1c were also observed, p < 0.0003 and p < 0.0013, respectively. Lifestyle modifications were specifically cited in 51% of records among deprescribed patients, most frequently related to diet and exercise. No serious adverse events were identified in patients who were deprescribed. Conclusions: In a primary care setting where patients had access to lifestyle medicine, a subset of adults with T2D experienced meaningful health improvements and were able to reduce glucose-lowering medications without any serious adverse events noted in the EHR. Full article

Background: Tirzepatide, a dual GIP/GLP-1 receptor agonist, demonstrates substantial glycemic and weight benefits versus GLP-1 receptor agonists in indirect comparisons, but direct comparative safety evidence versus dulaglutide remains limited. We evaluated comparative safety (primary outcome: overall adverse events) and efficacy. Methods: Following PRISMA 2020 (prospectively registered: PROSPERO CRD420251276594), we searched MEDLINE, Embase, Scopus, and CENTRAL (inception–31 December 2025) for randomized controlled trials (≥26 weeks) comparing once-weekly tirzepatide with dulaglutide in adults with type 2 diabetes. Three trials (N = 13,590 participants) were included. Dichotomous outcomes were pooled using random-effects models (risk ratios [RRs], 95% confidence intervals [CIs]). GRADE assessed certainty of evidence. Results: Overall adverse event incidence did not differ significantly (RR 1.04 [0.98–1.10]; I² = 36%; moderate-certainty evidence). Discontinuation due to adverse events was consistently higher with tirzepatide (RR 1.32 [1.20–1.45]; I² = 0%; high-certainty evidence), representing a 32% increased risk across all populations. Categorical HbA1c target achievement was analyzed in two trials; the third trial reported HbA1c as a continuous outcome only. At the primary threshold (HbA1c < 7.0%), tirzepatide was consistently superior with no heterogeneity (RR 1.48 [1.33–1.64]; I² = 0%; p < 0.00001). Across all thresholds combined, heterogeneity was extreme (I² = 92%), limiting confidence in any pooled summary estimate; the greatest instability occurred at the strictest threshold (HbA1c < 5.7%; I² = 98%; p = 0.40). Tirzepatide showed greater HbA1c target attainment in treatment-naive patients receiving dulaglutide 0.75 mg, whereas the glycemic advantage was smaller in patients with established cardiovascular disease receiving dulaglutide 1.5 mg. Categorical weight-loss outcomes were analyzed in two trials; tirzepatide was associated with greater weight-loss threshold achievement (RR 8.80 [4.04–19.17]; very low-certainty evidence), although interpretation is limited by substantial heterogeneity and restricted generalizability. Serious adverse events were not significantly different (RR 0.82 [0.47–1.43]; I² = 42%). Conclusions: Overall adverse events were similar between treatments, but tirzepatide consistently increased discontinuation risk, indicating a clinically important tolerability-persistence trade-off. Glycemic efficacy was highly population-dependent: benefits were consistent at the primary HbA1c target (<7.0%; I² = 0%) in early-stage disease, whereas the advantage was smaller in long-standing disease with established cardiovascular disease. Tirzepatide may be favored when glycemic or weight efficacy is prioritized in earlier-stage disease, provided tolerability is proactively managed. Dulaglutide remains appropriate when persistence is threatened by tolerability concerns or cardiovascular risk reduction is the primary goal. Full article

Background: Genetic causes of growth failure should be suspected in patients born small for gestational age (SGA) who fail to show postnatal catch-up growth, present with severe short stature (SS), and exhibit a poor or absent response to growth hormone (rhGH) therapy. Mutations in the insulin-like growth factor 1 receptor (IGF1R) gene are associated with impaired growth, intrauterine growth restriction (IUGR), low birth weight and/or length, and postnatal SS. Case Description: A 9-year-old boy, born SGA for birth length, was evaluated for severe SS. Common causes of SS were excluded. At 9 years and 7 months of age, his height was 112.6 cm (−3.99 SDS), weight 18 kg (−3.79 SDS), and BMI 14.2 kg/m² (−1.8 SDS); pubertal development was Tanner stage 1. The target height was 158 cm (−2.62 SDS). Bone age was delayed by approximately one year compared with chronological age. Serum IGF-1 levels were within the upper-normal range for age. GH therapy (0.035 mg/kg/day) was initiated due to the lack of catch-up growth in an SGA subject. After three years of treatment, the height gain was only 0.5 SDS. IGF-1 levels showed a transient treatment-related increase, followed by persistent normalization during ongoing therapy. Next-generation sequencing (NGS) analysis identified novel heterozygous paternal nonsense variant in the IGF1R gene: c.3498C>G (p.Tyr1166Ter). At 12 years of age, impaired fasting glucose and reduced glucose tolerance were detected; consequently, it was decided to discontinue rhGH therapy, also in light of the IGF1R mutation and the lack of height recovery. Conclusions: This case underlines the critical role of genetic testing in the evaluation of patients born SGA. The coexistence of SGA status and an IGF1R gene mutation may provide a clear explanation for both the poor response to rhGH therapy and the increased risk of alterations in glucose metabolism. An extensive narrative review of the literature on growth outcomes and glucose metabolism abnormalities during GH treatment in SGA patients carrying IGF1R variants was also performed. Full article

As critical infrastructures, long, large-diameter shield tunnels require a reliable long-term seismic safety assessment. This study investigates the coupled effects of spatially non-uniform ground motions and bolt corrosion on the seismic response of long, large-diameter shield tunnels. Three types of non-uniform seismic waves (coherent, traveling, and coherent traveling) consistent with the design spectrum were generated, and a 3D discontinuous finite element model was used to analyze tunnel response under various corrosion locations and rates. Results show that non-uniform excitation significantly amplifies the seismic response in the order coherent traveling waves > traveling waves > coherent waves > uniform waves, with coherent traveling waves amplifying key seismic responses by up to 10 times compared with uniform excitation. Corrosion at circumferential joints causes more substantial responses than at longitudinal joints, and combined longitudinal corrosion shows a greater effect than single-joint corrosion. Under coherent traveling waves with high corrosion rates, the seismic response exhibits a distinct “top stronger than bottom” distribution. This study concludes that accurate seismic performance evaluation requires consideration of both variability in spatial ground motion and bolt corrosion, thereby supporting the seismic design and long-term maintenance of such structures. Full article