Search Results (1,584)

Trust in machine-learning models is critical for deployment by users, especially for high-risk tasks such as healthcare. Model trust involves much more than performance metrics such as accuracy, precision, or recall. It includes user readiness to allow a model to make decisions. Model trust is a multifaceted concept commonly associated with the stability of model predictions under variations in training data, noise, algorithmic parameters, and model explanations. This paper extends existing model trust concepts by introducing a novel Model Sureness measure. Some alternatively purposed Model Sureness measures have been proposed. Here, Model Sureness quantitatively measures the model accuracy stability under training data variations. For any model, this is carried out by combining the proposed Bidirectional Active Processing and Visual Knowledge Discovery. The proposed Bidirectional Active Processing method iteratively retrains a model on varied training data until a user-defined stopping criterion is met; in this work, this criterion is set to a 95% accuracy when the model is evaluated on the test data. This process further finds a minimal sufficient training dataset required for a model to satisfy this criterion. Accordingly, the proposed Model Sureness measure is defined as the ratio of the number of unnecessary cases to all cases in the training data along with variations of these ratios. Higher ratios indicate a greater Model Sureness under this measure, while trust in a model is ultimately a human decision based on multiple measures. Case studies conducted on three benchmark datasets from biology, medicine, and handwritten digit recognition demonstrate a well-preserved model accuracy with Model Sureness scores that reflect the capabilities of the evaluated models. Specifically, unnecessary case removal ranged from 20% to 80%, with an average reduction of approximately 50% of the training data. Full article

Microbiological analysis is crucial for identifying species, assessing infections, and diagnosing infectious diseases, thereby supporting both research studies and medical diagnosis. In response to these needs, accurate and efficient identification of bacterial colonies is essential. Conventionally, this process is performed through manual counting and visual inspection of colonies on agar plates. However, this approach is prone to several limitations arising from human error and external factors such as lighting conditions, surface reflections, and image resolution. To overcome these limitations, an automated bacterial colony counting and classification system was developed by integrating a custom-designed imaging device with advanced deep learning models. The imaging device incorporates controlled illumination, matte-coated surfaces, and a high-resolution camera to minimize reflections and external noise, thereby ensuring consistent and reliable image acquisition. Image-processing algorithms implemented in MATLAB were employed to detect bacterial colonies, remove background artifacts, and generate cropped colony images for subsequent classification. A dataset comprising nine bacterial species was compiled and systematically evaluated using five deep learning architectures: ResNet-18, ResNet-50, Inception V3, GoogLeNet, and the state-of-the-art EfficientNet-B0. Experimental results demonstrated high colony-counting accuracy, with a mean accuracy of 90.79% ± 5.25% compared to manual counting. The coefficient of determination (R² = 0.9083) indicated a strong correlation between automated and manual counting results. For colony classification, EfficientNet-B0 achieved the best performance, with an accuracy of 99.78% and a macro-F1 score of 0.99, demonstrating strong capability in distinguishing morphologically distinct colonies such as Serratia marcescens. Compared with previous studies, this research provides a time-efficient and scalable solution that balances high accuracy with computational efficiency. Overall, the findings highlight the potential of combining optimized imaging systems with modern lightweight deep learning models to advance microbiological diagnostics and improve routine laboratory workflows. Full article

The increasing deployment of Internet of Things (IoT) sensors in precision agriculture has created critical challenges related to wireless communication range, energy efficiency, and data transmission latency, particularly in large-scale rural operations. This systematic survey, conducted following the PRISMA 2020 guidelines, investigates how drones, acting as mobile data collectors and communication gateways, can enhance the performance of agricultural wireless sensor networks (WSNs). The literature search was carried out in the Scopus and IEEE Xplore databases, considering peer-reviewed studies published in English between 2014 and 2025. After duplicate removal, 985 unique articles were screened based on predefined inclusion and exclusion criteria related to relevance, agricultural application, and communication technologies. Following full-text evaluation, 64 studies were included in this review. The survey analyzes how drones can be efficiently integrated with WSNs to improve data collection, addressing technical and operational challenges such as energy constraints, communication range limitations, propagation losses, and data latency. It further examines the primary applications of drone-based data acquisition supporting efficiency and sustainability in agriculture, identifies the most relevant wireless communication protocols and Technologies and discusses their trade-offs and suitability. Finally, it considers how drone-assisted data collection contributes to improved prediction models and real-time analytics in digital agriculture. The findings reveal persistent challenges in energy management, coverage optimization, and system scalability, but also highlight opportunities for hybrid architectures and the use of intelligent reflecting surfaces (IRSs) to improve connectivity. This work provides a structured overview of current research and future directions in drone-assisted agricultural communication systems. Full article

This study proposes a circular economy strategy to recover phenolic compounds by valorizing shrimp shell waste into a chitosan biosorbent (CH-B). Its adsorption efficiency was evaluated compared to commercial activated carbon (AC) and synthetic XAD-4 resin. Kinetic analysis revealed that while both pseudo-first-order (PFO) and pseudo-second-order (PSO) models exhibited high correlations (R²$\ge$ 0.96), both CH-B and XAD-4 resin were best described by the PFO model. This aligns with diffusion-controlled processes consistent with the porous and physical nature of these adsorbents. In contrast, AC followed the PSO model. Isotherm modeling indicated that CH-B and AC fit the Temkin model, reflecting heterogeneous surfaces, whereas XAD-4 followed the Langmuir model (monolayer adsorption). Notably, CH-B exhibited a maximum adsorption capacity (q_m) of 229.2 mg/g, significantly outperforming XAD-4 (104.8 mg/g) and AC (90.2 mg/g). Thermodynamic and kinetic modeling confirmed that the adsorption mechanism was governed by a combination of electrostatic interactions, π–π stacking, and hydrogen bonding between the hydroxyl/amine groups of chitosan and phenolic compounds. Optimization using Box–Behnken design for CH-B showed optimal acidic pH and moderate temperature but non-significant effect of CH-B dose in the experimental domain. Optimisation results showed unexpected high removal efficiency at low CH-B dosages. A tentative explanation may be adsorbent aggre-gation, which needs to be confirmed by further experimental evidence. Full article

Land seismic data are always contaminated by surface waves, which demonstrate strong energy, low velocity, and long vibrations. Such noises often mask deep effective reflections, seriously reducing the data’s signal-to-noise ratio while limiting the imaging accuracy of complex deep structures and the efficiency of hydrocarbon reservoir identification. To address this critical technical bottleneck, this paper proposes a surface wave joint reconstruction method based on stationary phase analysis, combining the cross-correlation seismic interferometry method with the convolutional seismic interferometry method. This approach integrates cross-correlation and convolutional seismic interferometry techniques to achieve coordinated reconstruction of surface waves in both shot and receiver domains while introducing adaptive weight factors to optimize the reconstruction process and reduce interference from erroneous data. As a purely data-driven framework, this method does not rely on underground medium velocity models, achieving efficient noise reduction by adaptively removing reconstructed surface waves through multi-channel matched filtering. Application validation with field seismic data from the piedmont regions of western China demonstrates that this method effectively suppresses high-energy surface waves, significantly restores effective signals, improves the signal-to-noise ratio of seismic data, and greatly enhances the clarity of coherent events in stacked profiles. This study provides a reliable technical approach for noise reduction in seismic data under complex near-surface conditions, particularly suitable for hydrocarbon exploration in regions with developed scattering sources such as mountainous areas in western China. It holds significant practical application value and broad dissemination potential for advancing deep hydrocarbon resource exploration and improving the quality of complex structural imaging. Full article

Hexavalent chromium (Cr(VI)) is a hazardous pollutant frequently found in industrial wastewater. Constructed wetlands (CWs) provide an alternative for Cr(VI) removal, but their effective removal is essentially governed by the extent of Cr accumulation in plants. This study evaluated the effects of pyrrhotite addition on a Cr-hyperaccumulator Leersia hexandra Swartz (L. hexandra) in CW microcosms with different substrates (pyrrhotite and gravel) and influent Cr(VI) concentrations (2 and 10 mg·L⁻¹). All microcosms achieved substantial Cr(VI) removal, while pyrrhotite significantly facilitated the removal of NO₃⁻-N, COD, and TP. Pyrrhotite alleviated Cr-induced oxidative stress and thus promoted photosynthesis in L. hexandra, reflected by 27.32–39.09% lower malondialdehyde levels, 1.67–8.37% higher total chlorophyll contents, and 17.36–39.61% higher net photosynthetic rates. Consequently, maximum aboveground Cr standing stock reached 164.50 mg·m⁻² in the P10 group, where L. hexandra contributed 6.63% to the total Cr removal. Microbial analysis showed reduced Cr-stress responses in pyrrhotite groups. Structural equation modeling indicated that pyrrhotite and its dissolution products promote Cr standing stock of L. hexandra through establishing in/ex planta defensive mechanisms. These findings provide new perspectives on phytoremediation coupled with CWs for the treatment of Cr(VI)-containing wastewater. Full article

In this paper, composite structures were fabricated by incorporating silver nanowires (AgNWs) with various metal oxides via the sol–gel method. This approach enhanced the electrical performance of AgNW-based transparent electrodes while simultaneously improving their stability under damp heat conditions and modifying the local medium environment surrounding the AgNW meshes. The randomly distributed AgNW meshes fabricated via drop-coating were treated with plasma to remove surface organic residues and reduce the inter-nanowire contact resistance. Subsequently, a zinc oxide (ZnO) coating was applied to further decrease the sheet resistance (R_sheet) value. The pristine AgNW mesh exhibits an R_sheet of 17.4 ohm/sq and an optical transmittance of 93.06% at a wavelength of 550 nm. After treatment, the composite structure achieves a reduced R_sheet of 8.7 ohm/sq while maintaining a high optical transmittance of 92.20%. The use of AgNW meshes as window electrodes enhances electron injection efficiency and facilitates the coupling mechanism between localized surface plasmon resonances and excitons. Compared with conventional ITO transparent electrodes, the incorporation of the AgNW mesh leads to a 17-fold enhancement in ZnO emission intensity under identical injection current conditions. Moreover, the unique scattering characteristics of the AgNW and metal oxide composite structure effectively reduce photon reflection at the device interface, thereby broadening the angular distribution of emitted light in electroluminescent devices. Full article

We introduce a new method for preparing a percolation system by employing an inverse percolation model. Unlike standard percolation, where the site occupancy is uniform, the new model imposes a distance-dependent probability of site removal, where sites closer to the lattice center have a higher probability of being removed and are more prone to damage as compared to those at the periphery of the system. The variation in this removal probability is a function of the distance $\left(d\right)$ from the central point. Thus, the central point plays a key role. This is reflected in our effort to model the role of a tumor cell and its surroundings (the tumor microenvironment). The tumor causes a decrease in the concentration of key elements, such as ${\mathrm{O}}_2$ (resulting in hypoxia) and Ca, in the region close to it, which in turn is an impediment to the efficiency of radiotherapy and chemotherapy. This decrease is the largest in sites adjacent to the tumor and smaller away from the tumor. Such change in the concentrations of these elements is vital in the mechanism of cancer therapies. Starting from a fully occupied lattice, we introduce a distance-dependent removal probability $q\left(d\right)$. The value of $q\left(d\right)$ is 1 at and next to the tumor (center) and decreases linearly away from it to a limiting value $q_p$, which is the value of q at the lattice boundaries. We investigate the system properties as a function of $q_p$ and observe a significant decrease in the critical percolation threshold $p_c$ as $q_p$ decreases, falling from the standard value of $p_c=0.5927$ to approximately $p_c=0.20$. Furthermore, we demonstrate that the size of the spanning cluster and the total number of clusters exhibit a strong dependence on $q_p$ as well. Full article

Background and Objectives: Recent advances in dentistry include microbiological and metabolomic analyses, which have the potential to improve the understanding of oral microbiome–host imbalances during orthodontic treatment. Fixed appliances, functional devices and, more recently, clear aligners have been associated with several oral health conditions, including enamel demineralization, dental caries, gingivitis, periodontitis and root and bone resorption. In this context, metabolomic approaches may enable the identification of metabolites in biological samples that could potentially serve as biomarkers and reflect functional biological changes within the oral ecosystem. Investigating orthodontic appliances and associated metabolomic alterations may therefore contribute to advancing current knowledge in orthodontics. This systematic review aimed to describe the available evidence on oral metabolomic changes during orthodontic treatment. Materials and Methods: A systematic literature search was conducted in PubMed, Web of Science, Scopus and the Cochrane Library. A total of 1632 records were identified. After duplicate removal and screening, 18 full-text articles were assessed for eligibility. Of these, 15 studies were excluded, and three studies met the inclusion criteria. Risk of bias was assessed using the ROBINS-I and RoB 2 tools, and the GRADE approach was applied to evaluate the certainty of evidence. The review protocol was registered in PROSPERO (CRD420251141544). Results: Three studies met the inclusion criteria. Overall, the available evidence was limited and heterogeneous. The included studies suggested potential differences in oral microbiome composition and metabolomic profiles between patients treated with fixed appliances and those treated with clear aligners. Reported metabolomic findings were exploratory and involved amino acid-related, immune-associated, and acidic metabolic pathways. Limitations: Only three studies were included, all conducted in a single country. The small sample size and methodological heterogeneity limit the generalizability of the findings. In addition, potential confounding variables highlight the need for further standardized longitudinal studies. Full article

Forest canopy near-infrared reflectance and mass-based canopy nitrogen concentration (canopy %N) have been shown to be positively correlated. While the mechanisms underpinning this relationship remain unresolved, the broad range of wavelengths involved points to structural properties that influence scattering and covary with %N. Despite this, efforts that have focused on commonly measured structural properties such as leaf area index (LAI) have failed to identify a causal mechanism. Here, we sought to understand how lidar-derived canopy traits related to additional properties of foliar arrangement and structural complexity modulate the effects of leaf spectra and leaf area index (LAI) on canopy reflectance. We developed a leaf layer spectra model to explore how canopy reflectance would change if complex foliage arrangements were removed, compressing the canopy into optically dense, uniform stacked layers while maintaining the same leaf area index. Model results showed that LAI-weighted leaf reflectance saturates at a leaf area index of approximately two for needleleaf species and four for broadleaf species. When upscaled to estimate plot-level canopy reflectance in the absence of structural complexity (NIRr_LAI), results showed a strong positive relationship with canopy %N (r² = 0.86), despite a negative relationship for individual leaves or “big-leaf” canopies with an LAI of one (NIRrL, r² = 0.78). This result implies that the relationship between canopy near-infrared reflectance and canopy %N results from the integrated effects of canopy complexity acting on differences in leaf-level optical properties. We introduced an index of relative reflectance (IRr) that shows that the relative contribution of structural complexity to canopy near-infrared reflectance (NIRr_C) is related to canopy %N (r² = 0.55), with a three-fold reduction from potential canopy near-infrared reflectance observed in stands with low %N compared to a two-fold reduction in stands with high %N. These findings support the hypothesis that the correlation between canopy %N and canopy reflectance is the result of interactions between leaf traits and canopy structural complexity. Full article

Background: Parasitic skin-related conditions represent a frequent and evolving challenge in human dermatology, as they often mimic other dermatoses, and are increasingly complicated by therapeutic resistance. With this study, we aimed to provide a practical, clinician-oriented overview of our experience, contextualising it within the current literature. Materials and Methods: We conducted a retrospective, single-centre observational study, reporting a case series of 88 patients diagnosed with parasitic or arthropod-related skin infestations at the San Raffaele Hospital Dermatology Unit (Milan) between 2019 and 2024, and integrated a concise narrative review of contemporary evidence on diagnosis, non-invasive imaging and management. For each case, we documented clinical presentation, dermoscopic or reflectance confocal microscopy (RCM) findings, and treatment response. Non-invasive tools (dermoscopy, videodermoscopy, RCM) were used when appropriate. Results: The spectrum of conditions included flea bites, bed bug bites, cutaneous larva migrans, subcutaneous dirofilariasis, Dermanyssus gallinae dermatitis, pediculosis, tick bites (including Lyme disease), myiasis, scabies, and cutaneous leishmaniasis. One case of eosinophilic dermatosis of haematologic malignancy was also considered due to its possible association with arthropod bites. Non-invasive imaging was critical in confirming suspected infestations, particularly in ambiguous cases or when invasive testing was not feasible. Several cases highlighted suspected therapeutic resistance: a paediatric pediculosis and three adult scabies cases required systemic therapy after standard regimens failed, raising concerns over putative resistance to permethrin and pyrethroids. In dirofilariasis, the persistence of filarial elements visualised by RCM justified the extension of antiparasitic therapy despite prior surgical removal. Conclusions: Our findings underline that accurate diagnosis, early intervention, and tailored treatment remain essential for the effective management of cutaneous infestations. The observed vast spectrum of isolated parasites reflects broader health and ecological dynamics, including zoonotic transmission, international mobility, and changing environmental conditions. At the same time, diagnostic delays, inappropriate treatments, and neglected parasitic diseases continue to pose significant risks. To address these challenges, clinicians should remain alert to atypical presentations, and consider a multidisciplinary approach including the consultation with parasitologists and veterinarians, as well as the incorporation of high-resolution imaging and alternative therapeutic strategies into their routine practice. Full article

Background/Objectives: Hepatocellular carcinoma (HCC) constitutes a leading cause of mortality worldwide. Liver transplantation (LT) and liver resection (LR) represent the main curative-intent treatment modalities for early-stage HCC. LT can offer the advantage of both removing the HCC and alleviating the potential underlying liver disease, yet its application is limited by organ scarcity, waitlist dropout, and eligibility criteria. Hence, LR remains widely used due to greater accessibility but is associated with high recurrence rates. Salvage LT is a treatment option for patients with HCC recurrence post-LR, but up to 40% of patients develop non-transplantable recurrence (NTR), defined as recurrence beyond transplant criteria, which precludes LT and is associated in poor outcomes. Methods: The present review aims to summarize the current state of evidence on the comparison of LT and LR, the management of recurrent HCC, and the risk factors associated with NTR. Results: Clinical and histopathologic factors consistently associated with NTR across studies include larger tumor size, multiple tumors, elevated alpha-fetoprotein levels, underlying liver fibrosis or cirrhosis, microvascular invasion, and satellite nodules—features that reflect aggressive tumor biology and impaired hepatic reserve. Conclusions: Improved preoperative risk stratification and identification of patients at high risk for NTR is essential to inform optimal treatment selection. Full article

In building measurement using terrestrial laser scanners (TLSs), acquired 3D point clouds (3DPCs) often contain significant reflection artifacts caused by reflective glass surfaces. Such reflection artifacts significantly degrade the performance of downstream applications. This study proposes a novel strategy, called GRASS, to remove these reflection artifacts. Specifically, candidate glass points are identified based on multi-echo returns caused by glass components. These potential glass regions are then refined through planar segmentation using geometric constraints. Then, we trace laser beam trajectories to identify the reflection affected zones based on the estimated glass planes and scanner positions. Finally, reflection artifacts are identified using dual criteria: (1) Reflection symmetry between artifacts and their source entities across glass components. (2) Geometric similarity through a 3D deep neural network. We evaluate the effectiveness of the proposed solution across a variety of 3DPC datasets and demonstrate that the method can reliably estimate multiple glass regions and accurately identify virtual points. Furthermore, both qualitative and quantitative evaluations confirm that GRASS outperforms existing methods in removing reflection artifacts by a significant margin. Full article

The dimensions and the reduction capacity of Fe₃O₄ nanoparticles are considered to be the key parameters in achieving the successful, efficient removal of hexavalent chromium, aiming for drinking water purification. This research study focuses on the optimization of reaction parameters during the oxidative precipitation of FeSO₄ carried out in a microwave-heated plug-flow reactor, to realize the preparation of Fe₃O₄ nanoparticles with an increased reduction potential as reflected in the Fe²⁺/Fe³⁺ ratio by approximating the ideal value of 0.5. In particular, the coupling of synthesis with features that allow for control of the oxidation extent, and include the addition of a reducing agent, an increase in ageing temperature, and inhibition of aggregation, were tested as potential approaches to tune the reducing potential and overcome reported Cr(VI) capture efficiencies provided by Fe₃O₄ nanoparticles. The evaluation results showed that adding a reductant after nanoparticle formation inhibits spontaneoussurface oxidation, bringing an improvement in the Cr(VI) uptake capacity for a residual concentration equal to the new EU regulation limit, by around 40%, reaching a value of 2.15 mg/g. However, working at an ageing temperature of around 100 °C resulted in an even better performance with an uptake increase of 120% and a capacity value of 3.45 mg/g. Finally, adding nanoparticles in the form of a dispersion instead of a dried powder provides an extra 10% improvement as a consequence of limited aggregation. Full article

Background: Repetitive peripheral magnetic stimulation (rPMS) delivers magnetic pulses to peripheral nerves and muscles, producing afferent input that can modulate corticospinal excitability (CSE). While the effects of rPMS on upper-limb muscles have been explored, its short-term effects on lower-limb CSE remain less understood. This study aimed to investigate the short-term effects of rPMS on CSE in the tibialis anterior (TA) muscle among healthy individuals. Methods: Twenty participants completed a repeated- measure, pre-post study. rPMS was applied to the non-dominant TA muscle at 10% above motor threshold for 15 min. CSE was assessed using transcranial magnetic stimulation (TMS), with measurements of motor evoked potential (MEP) amplitude, latency, and duration recorded at baseline, immediately after, 30 min, and 60 min post-stimulation. All analyses were conducted on clean datasets following removal of artifact-related outliers. Results: MEP amplitude showed a significant main effect of Side (p = 0.005), with greater values on the stimulated compared to the non-stimulated side. No significant main effects were found for Time (p = 0.351) or for the Side × Time interaction (p = 0.900). Descriptively, the largest increase in amplitude on the stimulated side was observed at 30 min post-stimulation (12% above baseline). MEP latency and duration showed no significant main or interaction effects. Conclusions: In conclusion, a single rPMS session applied to the TA produced a modest, side-specific increase in CSE lasting up to 60 min, as reflected in MEP amplitude. However, the absence of a significant time effect and perhaps non-optimized stimulation parameters limit the interpretation of sustained neuromodulatory effects. Future studies should examine optimal stimulation parameters and explore underlying mechanisms using measures such as the cortical silent period and interhemispheric inhibition. Full article