Search Results (1,463)

Accurate precipitation forecasting is paramount for water security and disaster mitigation, yet it remains formidable due to atmospheric stochasticity and the inherent class imbalance in rainfall datasets. This study proposes an integrated “architecture-feature-augmentation” framework to circumvent these limitations. Through a systematic evaluation of CNN-LSTM and Transformer architectures, we delineate distinct performance profiles: The Transformer model, when coupled with feature engineering and physics-informed augmentation, yields a peak F1-score of 0.1429, marking the optimal configuration for harmonizing precision and recall. Conversely, CNN-LSTM demonstrates superior robustness in extreme event detection, consistently maintaining high recall rates (up to 0.90) across diverse scenarios. We identify feature engineering as a critical performance modulator, substantially bolstering CNN-LSTM’s baseline metrics while enabling the Transformer to realize its maximum predictive capacity. Although synthetic oversampling techniques—such as SMOTE and GAN—effectively extend the detection range for heavy precipitation, physics-informed augmentation provides the most consistent performance gains, particularly in multi-class contexts. We conclude that the Transformer, augmented by physical constraints, is the optimal candidate for high-precision requirements, whereas CNN-LSTM, integrated with synthetic augmentation, offers a more sensitive alternative for early warning systems prioritizing recall. These findings provide empirical guidance for advancing extreme weather preparedness and strategic water resource management. Full article

Background/Objectives: Lung cancer is the leading cause of cancer-related mortality worldwide, and early detection using low-dose computed tomography (LDCT) substantially improves survival outcomes. However, variations in CT acquisition and reconstruction parameters including Hounsfield Unit (HU) calibration, reconstruction kernels, slice thickness, radiation dose, and scanner vendor introduce significant intensity and contrast variability that undermine the robustness and generalizability of deep-learning (DL) systems. Methods: This systematic review followed PRISMA 2020 guidelines and searched PubMed, Scopus, IEEE Xplore, Web of Science, ACM Digital Library, and Google Scholar for studies published between 2020 and 2025. A total of 100 eligible studies were included. The review evaluated preprocessing and harmonization strategies aimed at mitigating CT intensity variability, including perceptual contrast enhancement, HU-preserving normalization, physics-informed harmonization, and DL-based reconstruction. Results: Perceptual methods such as contrast-limited adaptive histogram equalization (CLAHE) enhanced nodule conspicuity and reported sensitivity improvements ranging from 10 to 15% but frequently distorted HU values and reduced radiomic reproducibility. HU-preserving approaches including HU clipping, ComBat harmonization, kernel matching, and physics-informed denoising were the most effective, reducing cross-scanner performance degradation, specifically in terms of AUC or Dice score loss, to below 8% in several studies while maintaining quantitative integrity. Transformer and hybrid CNN–Transformer architectures demonstrated superior robustness to acquisition variability, with observed AUC values ranging from $0.90$ to $0.92$ compared with $0.85$–$0.88$ for conventional CNN models. Conclusions: The evidence indicates that standardized HU-faithful preprocessing pipelines, harmonization-aware modeling, and multi-center external validation are essential for developing clinically reliable and vendor-agnostic AI systems for lung-cancer screening. However, the synthesis of results is constrained by the heterogeneous reporting of acquisition parameters across primary studies. Full article

The article presents a comprehensive determination and analysis of the dynamic accuracy of the AC traction network–pantograph interface using an equivalent lumped-parameter RLC model derived from a distributed-parameter representation of the traction line. The study investigates the system’s response to representative excitation signals: step, sinusoidal, and multi-harmonic, where the root mean square value of the voltage error at the network–pantograph interface is adopted as the main performance indicator. A novel contribution of this work lies in determining the upper bound on the dynamic error (UBDE) for input signals constrained by realistic physical limitations: initially by magnitude and duration, and subsequently extended with an additional rate of change constraint. In the first case, an iterative optimization procedure is applied to determine the constrained excitation and its corresponding error, while in the extended case, the problem of maximizing the dynamic error energy is solved numerically using a genetic algorithm. In both formulations, the objective is to identify extreme, physically admissible excitation waveforms that represent the most unfavorable dynamic scenarios for voltage reproduction within the traction network–pantograph RLC interface. The results obtained in this study are of both theoretical and practical significance. They allow the identification of frequency ranges and resonance conditions that intensify dynamic errors, support the design of compensation and filtering strategies, and enable the assessment of the system robustness to fast disturbances and supply voltage distortions. From a theoretical point of view, the article introduces a unified methodology for the determination and evaluation of dynamic errors and their worst-case upper estimates under realistic signal constraints, providing a foundation for future research on control design, optimization, and voltage quality requirements in AC traction power systems. Full article

Background and Objectives: Radiomics and artificial intelligence (AI) offer emerging quantitative tools for enhancing the diagnostic evaluation of testicular cancer. Conventional imaging—ultrasound (US), magnetic resonance imaging (MRI), and computed tomography (CT)—remains central to management but has limited ability to characterize tumor subtypes, detect occult nodal disease, or differentiate necrosis, teratoma, and viable tumor in post-chemotherapy residual masses. This systematic review summarizes current advances in radiomics and AI for both primary tumors and retroperitoneal disease. Materials and Methods: A systematic search of PubMed, Scopus, and Web of Science identified studies applying radiomics or AI to testicular tumors, retroperitoneal lymph nodes and post-chemotherapy residual masses. Eligible studies included quantitative imaging analyses performed on ultrasound, MRI, and CT, with optional integration of clinical or molecular biomarkers. Eighteen studies met inclusion criteria and were evaluated with respect to methodological design, diagnostic performance, and translational readiness. Results: Across modalities, radiomics demonstrated encouraging discriminatory capacity, with accuracies of 74–82% for ultrasound, 80.7–97.9% for MRI, and 71.7–85.3% for CT. CT-based radiomics for post-chemotherapy residual masses showed moderate ability to distinguish necrosis/fibrosis, teratoma, and viable germ-cell tumor, though heterogeneous methodologies and limited external validation constrained generalizability. The strongest performance was observed in multimodal approaches: integrating radiomics with clinical variables or circulating microRNAs improved accuracy by up to 12% and 15%, respectively, mirroring gains reported in other oncologic radiomics applications. Persisting variability in segmentation practices, acquisition protocols, feature extraction, and machine-learning methods highlights ongoing barriers to reproducibility. Conclusions: Radiomics and AI-enhanced frameworks represent promising adjuncts for improving the noninvasive evaluation of testicular cancer, particularly when combined with clinical or molecular biomarkers. Future progress will depend on standardized imaging protocols, harmonized radiomics pipelines, and multicenter prospective validation. With continued methodological refinement and clinical integration, radiomics may support more precise risk stratification and reduce unnecessary interventions in testicular cancer. Full article

The development of lithium-ion batteries for electric vehicles and other applications requires numerous complex and time-consuming research efforts. Numerical modeling can significantly reduce both the scope and duration of laboratory testing by enabling rapid prediction of cell behavior under various operating conditions. In this study, it is demonstrated that the parameters of the Newman–Tiedemann–Gu–Kim (NTGK) battery model can be determined using only extreme discharge current values, omitting intermediate currents. This approach increases the average voltage error by 0.23% but reduces the average temperature error by 0.22%. Additionally, the use of limited experimental data leads to extrapolation errors at an 8 A discharge current from 1.20% to 0.65% for voltage and from 7.04% to 5.78% for temperature. Furthermore, the proposed model enables accurate prediction of the state of charge (SoC) and battery temperature evolution without additional measurements under realistic driving conditions, such as the Worldwide Harmonized Light-Duty Vehicle Test Cycle (WLTC). Full article

This study investigates rim seal flow in axial turbine configurations through a combined experimental–numerical approach, with the objective of identifying sealing-flow conditions that minimize ingestion while limiting aerodynamic losses. Experimental measurements from the University of BATH are used to validate computational methodology, ensuring consistency with established sealing-effectiveness trends. The work places particular emphasis on the influence of computational domain selection and interface treatment, which is shown to strongly affect the prediction of ingestion mechanisms. A key contribution of this study is the systematic assessment of multiple domain configurations, demonstrating that a frozen rotor MRF formulation provides the most reliable steady-state representation of pressure-driven ingress, whereas stationary and non-interface domains tend to overpredict sealing effectiveness. A simplified thin-seal model is also evaluated and found to offer an efficient alternative for global performance predictions. Furthermore, a statistical orifice-based model is introduced to estimate minimum sealing flow for different rim seal geometries, providing a practical engineering tool for purge-flow scaling. The effects of pre-swirl injection are examined and shown to substantially reduce rotor wall shear and moment coefficient, contributing to lower windage losses without significantly modifying sealing characteristics. Unsteady flow features are explored using a harmonic balance method, revealing Kelvin–Helmholtz-type instabilities that drive large-scale structures within the rim seal cavity, particularly near design-speed operation. Finally, results highlight a clear trade-off between sealing-flow rate and turbine isentropic efficiency, underlining the importance of optimized purge-flow management. Full article

Leptospirosis remains a significant public health and economic burden in Southeast Asia, particularly in low- and middle-income countries where environmental, occupational, and socioeconomic factors contribute to its endemicity. Transmission is driven by close interactions between humans and infected animal reservoirs, alongside climatic conditions such as heavy rainfall and flooding. The region’s high but variable seroprevalence reflects inconsistencies in diagnostic methodologies and surveillance systems, complicating disease burden estimation. Major gaps persist in diagnostic capabilities, with current tools often unsuitable for resource-limited settings, leading to underdiagnosis and delayed treatment. Environmental modeling and spatial epidemiology are underutilized due to limited interdisciplinary data integration and predictive capacity. Addressing these challenges requires a One Health approach that integrates human, animal, and environmental health sectors. Key policy recommendations include harmonized surveillance, standardized and validated diagnostics, expanded vaccination programs, improved animal husbandry, and targeted public education. Urban infrastructure improvements and early warning systems are also critical, particularly in disaster-prone areas. Strengthened governance, cross-sectoral collaboration, and investment in research and innovation are essential for sustainable leptospirosis control. Implementing these measures will enhance preparedness, reduce disease transmission, and contribute to improved public health outcomes in all sectors across the region. Full article

Microalgae are sustainable sources of bioactive compounds with broad hepato-protective potential. This review synthesizes evidence for five major classes—carotenoids such as astaxanthin and fucoxanthin, polysaccharides such as paramylon and fucoidan, phycobiliproteins such as phycocyanin, omega-3 fatty acids, and phenolic extracts—linking their actions to key liver injury mechanisms. Preclinically, these compounds enhance antioxidant defenses, improve mitochondrial function, suppress inflammatory signaling, regulate lipid metabolism, modulate the gut–liver axis, and inhibit hepatic stellate cell activation, thereby attenuating fibrosis. Consistent benefits are observed in models of non-alcoholic and alcoholic fatty liver disease, drug-induced injury, ischemia–reperfusion, and fibrosis, with marked improvements in liver enzymes, oxidative stress, inflammation, steatosis, and collagen deposition. Emerging evidence also highlights their roles in regulating endoplasmic reticulum stress and ferroptosis. Despite their promise, translational challenges include compositional variability, a lack of standardized quality control, limited safety data, and few rigorous human trials. To address these challenges, we propose a framework integrating multi-omics and AI-assisted strain selection with specification-driven quality control and formulation-aware designs—such as lipid carriers for carotenoids or rational combinations like fucoxanthin with low-molecular-weight fucoidan. Future priorities include composition-defined randomized controlled trials in non-alcoholic fatty liver disease, alcoholic liver disease, and drug-induced liver injury; harmonized material specifications; and multi-constituent interventions that synergistically target oxidative, inflammatory, metabolic, and fibrotic pathways. Full article

Digital Twin technology is transforming how buildings are designed, operated, and optimized, serving as a key enabler of smarter, more energy-efficient, and sustainable built environments. By creating dynamic, data-driven virtual replicas of physical assets, Digital Twins support continuous monitoring, predictive maintenance, and performance optimization across a building’s lifecycle. This paper provides a structured review of current developments and future trends in Digital Twin applications within the building sector, particularly highlighting their contribution to decarbonization, operational efficiency, and performance enhancement. The analysis identifies major challenges, including data accessibility, interoperability among heterogeneous systems, scalability limitations, and cybersecurity concerns. It emphasizes the need for standardized protocols and open data frameworks to ensure seamless integration across Building Management Systems (BMSs), Building Information Models (BIMs), and sensor networks. The paper also discusses policy and regulatory aspects, noting how harmonized standards and targeted incentives can accelerate adoption, particularly in retrofit and renovation projects. Emerging directions include Artificial Intelligence integration for autonomous optimization, alignment with circular economy principles, and coupling with smart grid infrastructures. Overall, realizing the full potential of Digital Twins requires coordinated collaboration among researchers, industry, and policymakers to enhance building performance and advance global decarbonization and urban resilience goals. Full article

Food loss and waste (FLW) is a chronic problem across food systems worldwide, with meat being one of the most resource-intensive and perishable categories. The perishable character of meat, combined with complex cold chain requirements and consumer behavior, makes the sector particularly sensitive to inefficiencies and loss across all stages from production to consumption. This review synthesizes the latest advancements in new preservation technologies and supply chain efficiency strategies to minimize meat wastage and also outlines current challenges and future directions. New preservation technologies, such as high-pressure processing, cold plasma, pulsed electric fields, and modified atmosphere packaging, have substantial potential to extend shelf life while preserving nutritional and sensory quality. Active and intelligent packaging, bio-preservatives, and nanomaterials act as complementary solutions to enhance safety and quality control. At the same time, blockchain, IoT sensors, AI, and predictive analytics-driven digitalization of the supply chain are opening new opportunities in traceability, demand forecasting, and cold chain management. Nevertheless, regulatory uncertainty, high capital investment requirements, heterogeneity among meat types, and consumer hesitancy towards novel technologies remain significant barriers. Furthermore, the scalability of advanced solutions is limited in emerging nations due to digital inequalities. Convergent approaches that combine technical innovation with policy harmonization, stakeholder capacity building, and consumer education are essential to address these challenges. System-level strategies based on circular economy principles can further reduce meat loss and waste, while enabling by-product valorization and improving climate resilience. By integrating preservation innovations and digital tools within the framework of UN Sustainable Development Goal 12.3, the meat sector can make meaningful progress towards sustainable food systems, improved food safety, and enhanced environmental outcomes. Full article

Background/Objectives: Contrast-enhanced harmonic endoscopic ultrasonography (CH-EUS) is a promising tool for differentiating gastric subepithelial tumors (SETs). However, most published studies have mainly included gastrointestinal stromal tumors (GIST) and leiomyomas in the gastrointestinal tract, not limited to gastric SETs. This study evaluated the diagnostic performance of CH-EUS in gastric SETs encountered in clinical practice. Methods: We retrospectively analyzed 68 patients who underwent CH-EUS for gastric SETs between March 2021 and July 2025 at our institution. Gastric SETs were classified into benign (n = 27: ectopic pancreas, leiomyoma, schwannoma, glomus tumor, plexiform fibromyxoma, desmoid tumor, solitary fibrous tumor, and abscess) and GIST groups (n = 41). CH-EUS features, including arterial enhancement, irregular vessels, and diffuse enhancement, were assessed. Histopathological confirmation was obtained through EUS-guided fine-needle biopsy or endoscopic/surgical resection. Results: The GIST group showed significantly higher rates of arterial enhancement (95.1% vs. 74.1%, p = 0.024), irregular vessels (51.2% vs. 22.2%, p = 0.017), and diffuse enhancement (87.8% vs. 66.7%, p = 0.035) than the benign SETs. The diagnostic performance of arterial enhancement showed a sensitivity of 95.1% and specificity of 25.9%, while irregular vessels demonstrated a sensitivity of 51.2% and specificity of 77.8%, and diffuse enhancement showed a sensitivity of 87.8% and specificity of 33.3%. When combining ≥2 CH-EUS features, the sensitivity and specificity were 92.7% and 33.3%, respectively, with an overall accuracy of 69.1%. The presence of all three features yielded a specificity of 81.5% but a lower sensitivity (46.3%). Conclusions: CH-EUS exhibited a high sensitivity but low specificity in differentiating GISTs from various benign gastric SETs when using a combination of at least two CE-EUS features, including arterial enhancement, irregular vessels, and diffuse enhancement. Full article

Bladder cancer (BC) represents a major global health issue with high recurrence and significant mortality rates in cases of advanced disease. Currently, the development of molecular profiling, liquid biopsy technologies, and artificial intelligence (AI) software has resulted in unprecedented opportunities to improve diagnosis, prognostic assessment, and treatment selection. Recent multicenter studies have identified emerging metabolomic, proteomic, and genomic biomarkers with high sensitivity and specificity that may help replace or complement invasive approaches. AI-driven models that combine multi-omics datasets with radiomics and clinical parameters have demonstrated improved accuracy for predicting both therapeutic response and long-term outcomes, compared to standard approaches for risk stratification. Additionally, the incremental clinical usefulness of liquid biopsy platforms has been demonstrated for the monitoring of non-muscle-invasive bladder cancer and minimal disease detection. As these innovations converge, they herald the advent of a new era of personalized management of urothelial carcinoma; however, broad-based clinical implementation will require large-scale validation, standardization, regulatory harmonization, and economic analyses. Background: Bladder cancer continues to be a global health problem, particularly in the advanced disease setting where treatment options are limited, and mortality remains high. The exciting advances in precision medicine, including breakthrough molecular profiling techniques, liquid biopsy, and opportunities to apply AI to interpret these molecular data, hold unprecedented promise in improving the accuracy of diagnosis, prognostic stratification, and therapeutic decision-making. Full article

The rapid commercialization of next-generation photovoltaic (PV) technologies, particularly perovskite, thin-film roll-to-roll (R2R) architectures, and tandem devices, requires robust assessment of environmental performance at the level of industrial manufacturing processes. Environmental impacts can no longer be evaluated solely at the device or module level. Although many life-cycle assessment (LCA) studies compare silicon, cadmium telluride (CdTe), copper indium gallium selenide (CIGS), and perovskite technologies, most rely on aggregated indicators and database-level inventories. Few studies systematically compile and harmonize process-level life-cycle inventories (LCIs) for the manufacturing steps that differentiate emerging industrial routes, such as solution coating, R2R processing, atomic layer deposition, low-temperature annealing, and advanced encapsulation–metallization strategies. In addition, inconsistencies in functional units, system boundaries, electricity-mix assumptions, and scale-up modeling continue to limit meaningful cross-study comparison. To address these gaps, this review (i) compiles and critically analyzes process-resolved LCIs for innovative PV manufacturing routes across laboratory, pilot, and industrial scales; (ii) quantifies sensitivity to scale-up, yield, throughput, and electricity carbon intensity; and (iii) proposes standardized methodological rules and open-access LCI templates to improve reproducibility, comparability, and integration with techno-economic and prospective LCA models. The review also synthesizes current evidence on recycling, circularity, and critical-material management. It highlights that end-of-life (EoL) benefits for emerging PV technologies are highly conditional and remain less mature than for crystalline-silicon systems. By shifting the analytical focus from technology class to manufacturing process and life-cycle configuration, this work provides a harmonized evidence base to support scalable, circular, and low-carbon industrial pathways for next-generation PV technologies. Full article

Aflatoxin M₁ (AFM₁), a hydroxylated metabolite of aflatoxin B₁ (AFB₁), is a potent hepatotoxic and carcinogenic compound frequently detected in milk and dairy products. Its thermal stability and resistance to processing make it a persistent public health concern, especially in regions prone to fungal contamination of animal feed. This review integrates bibliometric mapping (2015–2025) with toxicological and mitigation perspectives to provide a comprehensive understanding of AFM₁. The bibliometric analysis reveals a sharp global rise in research output over the last decade, with Iran, China, and Brazil emerging as leading contributors and Food Control identified as the most prolific journal. Five research clusters were distinguished: feed contamination pathways, analytical detection, toxicological risk, regulatory frameworks, and mitigation strategies. Toxicological evidence highlights AFM₁’s mutagenic and hepatocarcinogenic effects, intensified by co-exposure to other mycotoxins or hepatitis B infection. Although regulatory limits range from 0.025 µg/kg in infant formula (EU) to 0.5 µg/kg in milk (FDA), non-compliance remains prevalent in developing regions. Current mitigation approaches—adsorbents (bentonite, zeolite), oxidation (ozone, hydrogen peroxide), and biological detoxification via lactic acid bacteria and yeasts—show promise but require optimization for industrial application. Persistent challenges include climatic variability, inadequate feed monitoring, and heterogeneous regulations. This review emphasizes the need for harmonized surveillance, improved analytical capacity, and sustainable intervention strategies to ensure dairy safety and protect consumer health. Full article

Early-onset colorectal cancer (EOCRC) continues to rise, with the steepest increases observed among Hispanic/Latino (H/L) populations, underscoring the urgency of identifying ancestry- and treatment-specific biomarkers. The JAK-STAT signaling axis plays a central role in colorectal tumor biology, yet its relevance under FOLFOX-based chemotherapy in EOCRC remains poorly defined. In this study, we evaluated 2515 colorectal cancer (CRC) cases (266 H/L; 2249 non-Hispanic White [NHW]), stratifying analyses by ancestry, age of onset, and FOLFOX exposure. Statistical comparisons were performed using Fisher’s exact and chi-square tests, and survival patterns were assessed via Kaplan–Meier analysis. To extend conventional analytics, we deployed AI-HOPE and AI-HOPE-JAK-STAT, conversational artificial intelligence platforms capable of harmonizing genomic, clinical, demographic, and treatment variables through natural language queries, to accelerate multi-parameter biomarker exploration. JAK-STAT pathway alterations showed marked variation by ancestry and treatment context. Among H/L EOCRC cases, alterations were significantly enriched in patients who did not receive FOLFOX compared with those who did (21.2% vs. 4.1%; p = 0.003). A similar pattern emerged in late-onset CRC (LOCRC) NHW patients, where alterations were more frequent without FOLFOX exposure (13.3% vs. 7.5%; p = 0.0002). Notably, JAK-STAT alterations were significantly more common in untreated H/L EOCRC compared with untreated NHW EOCRC (21.2% vs. 9.9%; p = 0.002). Survival analyses revealed that JAK-STAT pathway alterations conferred improved overall survival across several NHW strata, including EOCRC treated with FOLFOX (p = 0.0008), EOCRC not treated with FOLFOX (p = 0.07), and LOCRC not treated with FOLFOX (p = 0.01). These findings suggest that JAK-STAT alterations may function as ancestry- and treatment-dependent prognostic markers in EOCRC, particularly among disproportionately affected H/L patients. However, prognostic interpretation in H/L subgroups is limited by small mutation-positive sample sizes, reflecting historical underrepresentation and highlighting the need for larger ancestry-balanced studies. The integration of AI-enabled platforms streamlined analyses and reveals the potential of artificial intelligence to accelerate discovery and advance precision medicine for populations historically underrepresented in cancer genomics research. Full article