Search Results (10,339)

Identifying the molecular origins of thermal and electric-field response can guide polyimide insulation design. We combined an 18-film aromatic imide oligomer ensemble, reactive molecular dynamics, and signed finite-field r²SCAN-3c calculations to separate packing, chain-end chemistry, and directional electronic response. Independently initialized films exhibited a fivefold variation in probe-accessible volume, with distinct connected void geometries at similar accessible fractions. Across sixteen 5 ps thermal trajectories, terminal imide N–H sites accounted for 20 of 21 persistent reference-bond losses; the nonterminal inventory contained 14 new C–O connectivities and one C–N loss. In six matched film pairs, nonterminal records increased from 2 to 5 at 1600 K and from 12 to 19 at 1800 K between 5 and 10 ps, revealing the growing contribution of nonterminal rearrangement. Trajectories extending to 20 ps also resolved the onset of increased mass escape. At ±2 V nm⁻¹, the isolated ether-linked bis-imide fragment showed 0.342–0.841 eV frontier-gap narrowing across six directions, compared with at most 0.049 eV for the imide fragment, despite overlapping size-normalized induced responses. Together, these results identify chain termination and segment orientation as candidate design variables and provide quantitative descriptors for morphology-matched thermal comparisons and orientation-resolved electronic screening. Full article

Human activity recognition (HAR) on wearable and mobile devices requires models that can operate with limited computational resources while remaining reliable under changing real-world sensing conditions. However, most TinyML-based HAR systems follow the same computational path for every input, regardless of whether an activity pattern is familiar or affected by changes in the user, sensor placement, or sensing configuration. This study presents a shift-aware adaptive TinyML framework that adjusts the inference effort according to the prediction confidence and feature-space distribution shift. A lightweight multi-exit architecture allows confident samples to be classified at early stages, while uncertain or shifted samples are routed to deeper exits. A source-derived feature-space shift score complements confidence-based routing by identifying distribution changes that may not be reflected by predictive uncertainty alone. Persistent severe shifts are additionally considered through a guarded label-free test-time adaptation mechanism that does not require new activity annotations or full model retraining. The framework was evaluated on three publicly available HAR benchmark datasets, REALDISP, WISDM, and PAMAP2, covering complementary sensor-displacement, sensing-domain, unseen-user, and body-location shifts. Across the three benchmarks, explicit shift-aware routing produced a stronger adaptive computational response to changing sensing conditions than confidence-only routing. In REALDISP, the computational cost remained nearly unchanged from IDEAL to SELF conditions with confidence-only routing, whereas shift-aware routing increased the normalized computation by 0.0187 (p = 0.0253). In WISDM and PAMAP2, the increase in computation under the distribution shift was approximately 2.88 and 2.85 times greater with shift-aware routing than with confidence-only routing, respectively. The effect on recognition performance, however, depended on the shift scenario. Shift-aware routing significantly improved the target-domain Macro-F1 in the WISDM WATCH-to-PHONE scenario (ΔMacro-F1 = +0.0096, p = 0.0023), whereas no significant differences were observed for REALDISP SELF or PAMAP2 ANKLE. These findings demonstrate that explicit distribution-shift information can help resource-constrained HAR models allocate additional computation when sensing conditions change, while also showing that deeper inference alone may not overcome severe domain mismatch. The proposed framework provides a practical basis for adaptive inference in TinyML-based wearable and edge sensing systems operating under variable real-world conditions. Full article

The Early Cretaceous post-collisional magmatic evolution of the central Lhasa Terrane remains hotly debated, and the Qula syenogranite provides new constraints on crust–mantle interaction. Whole-rock major and trace element analyses, zircon LA-ICP-MS U–Pb dating, and Sr–Nd–Pb–Hf isotopic analyses were carried out on seven samples from this newly recognized pluton to decipher its petrogenesis and tectonic setting. This pluton exhibits typical geochemical signatures of a weakly peraluminous, ferroan A₂-type granite, including elevated SiO₂ and total alkalis, low Al₂O₃ and CaO contents, high FeO^T/(FeO^T + MgO) (0.86–0.91) and 10,000 × Ga/Al (2.9–3.1), prominent negative Eu anomalies, chondrite-normalized rare earth element patterns with a seagull shape, enrichment in Rb, Th, U, Zr, and Hf, and depletion in Ba, Sr, P, Nb, and Ti. Zircon U–Pb dating yields a weighted mean crystallization age of 105.6 ± 0.46 Ma (MSWD = 1.16, n = 24), corresponding to the late Early Cretaceous. Positive zircon εHf(t) (+4.4 to +5.8), moderately negative whole-rock εNd(t) values (−5.3 to−3.4), and radiogenic Pb isotopic compositions together indicate that the parental magma most likely originated from the partial melting of juvenile lower crust, with contributions from both ancient basement and mantle-derived melts. Slab break-off of the southward-subducted Bangong–Nujiang lithosphere and consequent asthenospheric upwelling following the Qiangtang–Lhasa collision provide the tectonic scenario most consistent with these data, and the magma subsequently underwent extensive fractional crystallization dominated by K-feldspar, plagioclase, and biotite to form the Qula A₂-type syenogranite. Full article

Background: Late-onset psychosis (onset after age 60) in a patient with no prior psychiatric history is uncommon and frequently signals an underlying neurocognitive or structural brain disorder rather than a primary psychiatric illness. Erotomania, the delusional belief that an inaccessible or higher-status person is in love with the patient, has rarely been described together with cognitive impairment and ventriculomegaly. We describe a case in which these three features occurred together and set out the competing explanations for their co-occurrence. Methods: We report a single retrospective case, drawing on inpatient psychiatric and neurology consultation records, cognitive screening with the Montreal Cognitive Assessment (MoCA), head computed tomography, brain magnetic resonance imaging, and one-month outpatient follow-up. Written informed consent for publication was obtained from the patient during the hospitalization, and his capacity for that specific decision was assessed by the treating team. Results: A 75-year-old man with no psychiatric history developed subacute, fixed erotomanic delusions involving a public figure, with an illogical and circumstantial thought process, intermittent disorientation, and severely impaired judgment. Acting on these beliefs, he sold his home and transferred approximately $200,000 to individuals impersonating the celebrity online. Neuroimaging showed ventriculomegaly disproportionate to cortical sulcal prominence together with chronic small-vessel ischemic change. Neurology characterized the picture as borderline or early normal pressure hydrocephalus versus hydrocephalus ex vacuo with a superimposed neurocognitive disorder. The recommended diagnostic large-volume lumbar puncture was not performed, and no quantitative morphometric measurements were obtained, so normal pressure hydrocephalus remained unconfirmed. MoCA performance worsened from 23/30 at admission to 17/30 one month later, although both assessments were obtained during active psychosis and antipsychotic treatment. He was treated with low-dose antipsychotic medication and discharged with outpatient psychiatric and neurosurgical follow-up, and delusions and cognitive impairment persisted at short-term follow-up. Conclusions: New-onset erotomanic delusions in an older adult should prompt evaluation for an underlying neurocognitive or neurological disorder, including structural imaging for potentially reversible causes. This single case cannot establish the direction of the relationship among the erotomanic syndrome, the cognitive impairment, and the ventriculomegaly. Because impaired judgment in this setting can produce catastrophic financial harm, clinicians should pair neuropsychiatric evaluation with formal decisional capacity assessment. Full article

Greywater can be a valuable source of energy, which is often wasted when discharged into the sewer system. The use of drain water heat recovery (DWHR) exchangers in buildings allows for the recovery of this energy. With this in mind, a new horizontal shower heat exchanger solution was developed, and a prototype was built. This device was tested in the laboratory under conditions similar to those of normal shower use. The obtained results allowed the determination of the temperature-based effectiveness for a wide range of variables. The experimental results also formed the basis for a Life Cycle Cost analysis of various domestic hot water system variants and for determining the environmental impact of implementing the tested exchanger. Depending on the system configuration, the DWHR temperature-based effectiveness ranged from 24% to 29% (Configuration 1), from 29% to 40% (Configuration 2), and from 38% to 57% in Configuration 3. The tests also confirmed that the new exchanger solution is financially viable. For two-person households with average water consumption per bath, the payback period does not exceed 7 years, and for a four-person family, it is reduced to approximately 3 years. A key aspect of the research was the demonstration of a significant reduction in operational CO₂ emissions resulting from the implementation of this DWHR exchanger in the building. This study highlights the importance of selecting the optimal configuration for connecting the exchanger to the sanitary system and confirms the need for continuous development and research into wastewater heat recovery technologies. Full article

Contactless palmprint verification must generalize to unseen identities while remaining robust to variations in hand pose, scale, illumination, camera distance, and background. This study presents a cosine-consistent deep metric-learning framework combining landmark-first Hybrid region-of-interest (ROI) extraction, contrast-limited adaptive histogram equalization (CLAHE) enhancement, an ImageNet-initialized ResNet18 encoder, trainable generalized-mean (GeM) pooling, a compact projection head, batch-normalization neck (BNNeck), batch-hard triplet learning, and auxiliary identity supervision. The principal contribution lies not in introducing a new backbone or individual architectural component, but in the mathematically consistent integration of metric learning and verification with a rigorously controlled open-set evaluation protocol. The triplet objective operates on L2-normalized pre-batch-normalization (pre-BN) descriptors, whereas enrollment and cosine-based verification use L2-normalized post-batch-normalization (post-BN) descriptors. For unit-normalized vectors, squared Euclidean distance is shown to be a strictly decreasing affine function of cosine similarity, establishing exact ranking equivalence within each embedding space. A training-distribution lower bound relates the triplet margin and violation probability to the expected pre-BN score difference, without extending this result to post-BN test scores. Conditional sub-Gaussian false acceptance rate (FAR) and false rejection rate (FRR) bounds and Hoeffding-based plug-in threshold evaluations are also developed, with empirical score statistics explicitly distinguished from finite-sample risk certificates. Experiments on Tongji Session 1 and BMPD use one fixed participant-disjoint split, validation-only threshold selection, five model-training seeds, controlled one-factor ablations, and participant-level bootstrap confidence intervals for principal verification measures. The results provide the clearest empirical support for auxiliary identity supervision under the more variable BMPD acquisition conditions. The effect of ROI normalization is acquisition-dependent, while the effects of CLAHE, pooling, BNNeck, mining strategy, embedding normalization, and triplet margin are generally smaller or more variable across seeds and datasets. Normalized Euclidean and cosine matching produce identical verification results when applied to the same normalized post-BN descriptors. Full article

The precise determination of the coefficient of friction is of fundamental importance in many engineering applications, especially in cases where different material pairs, surface-treated systems, polymers or coated surfaces come into contact with each other. Classical tables of friction data can often only be used as approximate values, since friction behavior is considerably influenced by surface pressure, sliding velocity, contact geometry and the condition of the surfaces. In this article, we present the principle, structural design and application possibilities of a new friction-measuring device that can be adapted to a universal testing machine (UTM). Our device provides a symmetric, two-sided flat-on-flat contact configuration, a closed loading system and a two-stage self-aligning mechanism, with adjustable normal load. The relevant rotational axes are positioned in the contact planes to reduce parasitic moments and the coupling between the tensile measuring force and the applied normal load, thereby reducing a potential source of measurement variability. With the device, both static and dynamic friction coefficients can be determined (and the transition ranges can be explored in detail), at different nominal contact pressures (p, up to 50 MPa) and sliding velocities (v, mm/min). The demonstrative measurement results show that the system is suitable for the separate examination of the characteristic stages of the friction process, such as the initial force build-up, the static friction maximum and the steady-state dynamic range. Full article

Forest fires significantly disturb the global ecosystem, profoundly impacting carbon cycling, biodiversity, and socio-economic sustainability. Understanding the seasonal patterns of forest fire occurrence is crucial for developing effective prevention and control strategies that support sustainable forest management and disaster risk reduction. However, research on the seasonal dynamics of forest fires in the tropical forests of China’s Hainan Island remains scarce, and the underlying driving mechanisms, particularly in relation to the distinct dry and rainy seasons, are still insufficiently understood. In this study, we analyzed the seasonal patterns and driving factors of forest fires on Hainan Island, a tropical region, from 2000 to 2022, and compared the applicability of emerging machine learning algorithms and traditional Logistic Regression algorithms in constructing tropical forest fire prediction models for the dry and rainy seasons. The results show that tropical forest fires in Hainan Island are primarily concentrated from January to May, peaking in April, with a lagged effect relative to the dry-season months (November to April of the following year). Average temperature and average rainfall were identified as common important predictors of fire occurrence probability in both seasons, with the Normalized Difference Vegetation Index additionally included in only the dry season. The predictive accuracy of the Random Forest model consistently surpassed that of the traditional Logistic Regression model in both seasons. Additionally, our models enable the generation of accurate forest fire probability maps, highlighting risk areas in Hainan Island during both seasons and providing new insights for wildfire prevention strategies and sustainable forest management in tropical regions. Full article

Background/Objectives: Thoracic re-irradiation is increasingly considered for recurrent, persistent, or new lung cancers after prior radiation therapy. The central clinical-technical challenge is relating prior and prospective doses across changes in anatomy, breathing state, fractionation, and imaging. Methods: This clinical-technical Perspective synthesizes consensus guidance, technical and the methodological literature, and representative clinical evidence to discuss rigid and deformable registration, dose mapping, biological normalization with EQD2/BED, uncertainty, quality assurance, and practical decision-making in lung re-irradiation. Results: Cumulative dose is an estimate with inherent uncertainty whose clinical value depends on the clinical question, anatomic stability, tissue correspondence, dose-transfer method, biological assumptions, and the consequence of error. Published cumulative dose constraints can provide useful orientation, but they are not universal pass/fail thresholds and should be interpreted together with their original endpoints, assumptions, and accumulation method. Conclusions: A focused, risk-adapted workflow can support more defensible use of cumulative dose in lung re-irradiation. Practical considerations in this article are intended to complement formal consensus guidance, institutional protocols, or patient-specific multidisciplinary judgment. Full article

Background and Clinical Significance: Infective endocarditis following transcatheter aortic valve implantation (TAVI) is a severe complication that may be associated with prosthetic valve dysfunction, local infectious extension, and systemic embolization. Visceral mycotic aneurysms are uncommon, particularly those involving the hepatic arterial circulation. Case Presentation: We report the case of an 80-year-old woman who developed Streptococcus gallolyticus TAVI-associated prosthetic valve endocarditis (TAVI-PVE) four months after valve implantation. Despite prolonged antimicrobial therapy and subsequent negative blood cultures, the clinical course was complicated by persistent prosthetic and periprosthetic echocardiographic abnormalities, cerebral and splenic embolic events, and new-onset abdominal pain. Contrast-enhanced computed tomography revealed two hepatic arterial aneurysmal lesions with surrounding inflammatory changes, consistent with mycotic aneurysms. After initially declining surgery, the patient underwent redo surgical aortic valve replacement; cultures of the surgically excised specimens were negative. Abdominal pain resolved and inflammatory markers normalized following surgery and continued antimicrobial therapy. Follow-up contrast-enhanced CT demonstrated marked interval enlargement of the hepatic arterial aneurysmal lesion to 28 mm × 28 mm × 26 mm despite clinical improvement and normalization of inflammatory markers. Conclusions: This case highlights that microbiological, clinical, and biological improvement does not necessarily indicate resolution of local cardiac or extracardiac structural complications and emphasizes the importance of repeated imaging, guideline-based antimicrobial therapy, timely source control, and multidisciplinary vascular management in complicated TAVI-PVE. Full article

Rule-based inspection of web server logs cannot detect attacks it has not already been told to look for, and most anomaly-detection studies validate their methods on a single dataset, which risks overstating how well a method generalises to new traffic. This paper presents a hybrid framework that fuses a content-based autoencoder, which scores individual HTTP requests, with a session-based variational LSTM autoencoder, which scores per-client request sequences. The framework is evaluated across eight datasets spanning three log formats, including the public CSIC 2010 HTTP dataset, under both within-dataset and cross-dataset threshold-transfer protocols. Correcting a session-construction fault that had mixed different clients’ requests into the same window raises the session branch’s AUC on CSIC 2010 from 0.63 to 0.98, evaluated on a held-out split disjoint from the requests the corrected model was trained on; the branch depends on genuine multi-request sessions and produces no score where the traffic does not contain them. Fusion under raw score averaging is not uniformly beneficial: on CSIC 2010, weighting fully toward the session branch outperforms every blended raw-average weight tested, though normalising each branch’s score against its own training-score distribution before averaging largely closes this gap. A threshold calibrated once on normal CSIC 2010 traffic and transferred unchanged to three attack-heavy datasets raises fused F1 six- to seven-fold over a threshold tuned separately on each dataset, because those datasets contain too much attack traffic for self-calibration to represent normal behaviour. The framework also achieves a lower false positive rate than a labelled Random Forest baseline, at comparable or better AUC than unsupervised baselines, and runs fast enough for near-real-time use. Taken together, these results show that session structure and threshold calibration can matter as much as model choice. Full article

Resonant DC–DC converter design requires family selection, feasible operating regions and component targets to be evaluated consistently. ResoSym-RD links topology-aware normalized screening to physical-unit synthesis based on the fundamental-harmonic approximation (FHA), rounded component targets and independently tested timing schedules. Series resonant (SRC), parallel-loaded resonant (PLRC), inductor–capacitor–capacitor (LCC) and inductor–inductor–capacitor (LLC) families retain their native reduced models. Nested search domains expose boundary-dependent coverage and trade-offs rather than a universal topology ranking. For a 360–420 V/48 V, 1 kW LLC case, the retained 20 µH/88 nF/106 µH targets have a minimum nominal normalized margin of 1.441. A prespecified rounded FHA comparator has a lower nominal margin and a higher ranking proxy, but a higher fixed-timing sampled pass fraction. With matched timing perturbations, an independent 20,000-realization test gives 99.925% charge-only feasibility for the locked schedule versus 98.860% for fixed 250 ns timing. A new common parasitic switched-circuit assessment compares both tanks over all nine corners. At equal 48 V output, the ResoSym-RD tank has 1.42% lower uniformly averaged modeled dissipation, but not at every corner. The uncorrected FHA schedule produces output errors up to 5.61%, and the ResoSym-RD peak-current error reaches 21.06%, establishing the need for time-domain closure. Executed commutation traces, step refinement and a separate numerical integrator support the declared generic circuit results. The study is an auditable pre-synthesis methodology, not a claim of measured efficiency, field reliability, electromagnetic interference (EMI) compliance or hardware qualification. Full article

Time-series anomaly detection holds value across various research fields and application domains, serving purposes such as fault diagnosis, identification of unexpected system intrusions, or signaling the onset of new disease outbreaks. To identify anomalous timepoints in a practical manner, unsupervised learning methods have been introduced. These methods leverage normality’s feature representations to reconstruct data, determining anomalies through the calculation of anomaly scores based on reconstruction errors. Still, recent approaches assume the use of a normal dataset for model training, implicitly necessitating true anomaly labels. Furthermore, the absence of guidelines for determining anomaly thresholds hinders the easy application of current detection methods in diverse research fields. In this paper, we propose VRGAN, a deep generative anomaly detection framework designed for unsupervised multivariate time-series analysis. VRGAN stands out by eliminating the need for assumptions or true label information. Through the training of variational recurrent neural network and GAN modules on a training dataset masked for anomaly candidates, VRGAN reconstructs the dataset based on learned normal data distribution and calculates anomaly scores to identify anomalies. Evaluation on seven benchmark datasets demonstrates VRGAN’s performance improvement compared with recent unsupervised anomaly detection methods. The proposed model is further tested on a time-series metagenomic dataset, showcasing its applicability in fully unsupervised settings for wastewater-based surveillance to monitor and track patterns of antibiotic resistance genes that are prevalent among bacteria carried by a given human community. Full article

This paper presents BSTwin, a computational workflow and application for the automatic generation of multi-level-of-detail building models and semantically enriched building stock records in urban areas of Slovakia. Its outputs are positioned as the building layer of a city information model (CIM) that can support future urban digital twin applications, not as an operational digital twin. BSTwin couples a lightweight, rule-based roof reconstruction engine for airborne LiDAR point clouds with a benchmark dataset mined from 13 open geodata sources: national registers (INSPIRE HVD Buildings, 2021 CENSUS, INFOREG-EC), European and global datasets (European building stock (EUBUCCO v0.2), Global Human Settlement Layer (GHSL R2023A), Copernicus Urban Atlas, GlobalBuildingAtlas (GBA), OpenBuildingMap (OBM), Microsoft Global Building Footprints) and open-source maps (OpenStreetMap, Overture Maps). Rather than proposing new learned segmentation models, the approach integrates published geometric methods (normal-seeded RANSAC with Hough parameter-space merging, ring-based ground estimation, 3DBAG reference heights, mSTEP azimuth rectification and recursive footprint decomposition) into a deterministic pipeline that runs on an ordinary computer without GIS/BIM software. The term GeoAI is used in its knowledge-driven sense, since no machine-learned model is involved. Each reconstruction and integrated attribute is verified through a quality-control block. On synthetic ground truth, the engine classified all five tested roof archetypes correctly with a point-to-plane RMSE of about 5 cm; on real national LiDAR covering 1468 buildings in the Staré Mesto district of Košice, it reconstructed 99.9% of the buildings, 96.2% of them without a quality flag, in a median of 0.20 s per building. The LiDAR ridge heights agreed with the INSPIRE register heights with a median difference of +0.48 m (65% within 1 m). For the Košice Self-governing Region, the bundled EUBUCCO v0.2 extract covers 488,363 buildings, 81.7% of which have heights of governmental origin. The findings indicate the potential of the workflow for building stock monitoring, renovation planning and as a data foundation for future urban digital twins. Full article

Background/Objectives: Drug repurposing represents a promising strategy for determining new anticancer drugs through exploration of new therapeutic applications of clinically used drugs. Alfuzosin and doxazosin are α-adrenergic blockers clinically used for the treatment of benign prostatic hyperplasia (BPH). Growing evidence suggests that α-blockers may have anticancer properties. The purpose of this study was to investigate the potential of alfuzosin and doxazosin as repurposed anticancer medicines and to study their potential mechanisms of action, specifically their effects on apoptosis, cell cycle progression, enzyme panel, and cancer cell proliferation. Methods: The cytotoxicity of alfuzosin and doxazosin was investigated against multiple cancer cell lines using the MTT assay, with sunitinib used as a positive control. The enzyme inhibition effects of both drugs against potential target proteins were investigated, followed by cytometric analysis to determine the effects of the drugs on cell cycle progression and apoptosis in MCF-7 breast cancer cells. Molecular docking studies were also conducted to investigate possible interactions with the active site of EGFR. Results: Both alfuzosin and doxazosin demonstrated promising antiproliferative activity against cancer cells, with alfuzosin showing comparatively lower toxicity toward normal cells. The drugs also revealed potent inhibitory activity against the studied target proteins, pointing to their potential effect as multi-target inhibitors. Cell cycle analysis revealed that both drugs inhibited cell cycle progression at the G2/M phase in MCF-7 cells. Following treatment with alfuzosin and doxazosin, both drugs significantly induced apoptosis, with total apoptotic populations reaching 39.48% and 35.61%, respectively, compared with only 2.77% in untreated MCF-7 cells. Molecular docking studies further indicated the ability to interact with key amino acids within the EGFR active site. Conclusions: The present findings demonstrate the potential repurposing of alfuzosin and doxazosin as anticancer agents and offer a good background for further mechanistic and in vivo investigations. Full article