Search Results (1,578)

Artificial intelligence-based image processing is critical for sensor fusion and image transformation in mobility systems. Advanced driver assistance functions such as forward monitoring and digital side mirrors are essential for driving safety. Degradation due to raindrops, fog, and high-dynamic range (HDR) imbalance caused by lighting changes impairs visibility and reduces object recognition and distance estimation accuracy. This paper proposes a diffusion framework to enhance visibility under multi-degradation conditions. The denoising diffusion probabilistic model (DDPM) offers more stable training and high-resolution restoration than the generative adversarial networks. The DDPM relies on large-scale paired datasets, which are difficult to obtain in raindrop scenarios. This framework applies the Palette diffusion model, comprising data augmentation and raindrop-removal modules. The data augmentation module generates raindrop image masks and learns inpainting-based raindrop synthesis. Synthetic masks simulate raindrop patterns and HDR imbalance scenarios. The raindrop-removal module reconfigures the Palette architecture for image-to-image translation, incorporating the augmented synthetic dataset for raindrop removal learning. Loss functions and normalization strategies improve restoration stability and removal performance. During inference, the framework operates with a single conditional input, and an efficient sampling strategy is introduced to significantly accelerate the process. In post-processing, tone adjustment and chroma compensation enhance visual consistency. The proposed method preserves fine structural details and outperforms existing approaches in visual quality, improving the robustness of vision systems under adverse conditions. Full article

Driving wheel slippage in agricultural tractors is influenced by soil moisture, density, and penetration resistance. These surface variations reflect post-tillage composition, enabling dynamic mapping via Remotely Piloted Aircraft (RPAs). This study evaluated ballast recommendations based on soil surface data and slippage percentages, correlating added wheel weights at different speeds for a tractor-reversible plow system. Six 94.5 m² quadrants were analyzed for slippage monitored by RPA (Mavic3M-RTK) pre- and post-agricultural operation overflights and soil sampling (moisture, density, penetration resistance). A 2 × 2 factorial scheme (F-test) assessed soil-surface attribute correlations and slippage under varying ballasts (52.5–57.5 kg/hp) and speeds. Results showed slippage ranged from 4.06% (52.5 kg/hp, fourth reduced gear) to 11.32% (57.5 kg/hp, same gear), with liquid ballast and gear selection significantly impacting performance in friable clayey soil. Digital Elevation Model (DEM) and spectral indices derived from RPA imagery, including Normalized Difference Red Edge (NDRE), Normalized Difference Water Index (NDWI), Bare Soil Index (BSI), Green–Red Vegetation Index (GRVI), Visible Atmospherically Resistant Index (VARI), and Slope, proved effective. The approach reduced tractor slippage from 11.32% (heavy ballast, 4th gear) to 4.06% (moderate ballast, 4th gear), showing clear improvement in traction performance. The integration of indices and slope metrics supported ballast adjustment strategies, particularly for secondary plowing operations, contributing to improved traction performance and overall operational efficiency. Full article

The rapid development of deep learning (DL) has demonstrated its modular contributions to artificial intelligence (AI) techniques, such as large language models (LLMs). DL variants have proliferated across domains such as feature extraction, normalization, lightweight architecture design, and module integration, yielding substantial advancements in these subfields. However, the absence of a unified review framework to contextualize DL’s modular evolutions within AI development complicates efforts to pinpoint future research directions. Existing review papers often focus on narrow technical aspects or lack systemic analysis of modular relationships, leaving gaps in our understanding how these innovations collectively drive AI progress. This work bridges this gap by providing a roadmap for researchers to navigate DL’s modular innovations, with a focus on balancing scalability and sustainability amid evolving AI paradigms. To address this, we systematically analyze extensive literature from databases including Web of Science, Scopus, arXiv, ACM Digital Library, IEEE Xplore, SpringerLink, Elsevier, etc., with the aim of (1) summarizing and updating recent developments in DL algorithms, with performance benchmarks on standard dataset; (2) identifying innovation trends in DL from a modular viewpoint; and (3) evaluating how these modular innovations contribute to broader advances in artificial intelligence, with particular attention to scalability and sustainability amid shifting AI paradigms. Full article

Multidimensional force sensors are key devices capable of simultaneously perceiving and analyzing force in multiple directions (normally triaxial forces). They are designed to provide intelligent systems with skin-like precision in environmental interaction, offering high sensitivity, spatial resolution, decoupling capability, and environmental adaptability. However, the inherent complexity of tactile information coupling, combined with stringent demands for miniaturization, robustness, and low cost in practical applications, makes high-performance and reliable multidimensional sensing and decoupling a major challenge. This drives ongoing innovation in sensor structural design and sensing mechanisms. Various structural strategies have demonstrated significant advantages in improving sensor performance, simplifying decoupling algorithms, and enhancing adaptability—attributes that are essential in scenarios requiring fine physical interactions. From this perspective, this article reviews recent advances in multidimensional force sensing technology, with a focus on the operating principles and performance characteristics of sensors with different structural designs. It also highlights emerging trends toward multimodal sensing and the growing integration with system architectures and artificial intelligence, which together enable higher-level intelligence. These developments support a wide range of applications, including intelligent robotic manipulation, natural human–computer interaction, wearable health monitoring, and precision automation in agriculture and industry. Finally, the article discusses remaining challenges and future opportunities in the development of multidimensional force sensors. Full article

Unrecognized mechanical faults and emergency sounds in vehicles can compromise safety, particularly for individuals with hearing impairments and in sound-insulated or autonomous driving environments. As intelligent transportation systems (ITSs) evolve, there is a growing need for inclusive, non-intrusive, and real-time diagnostic solutions that enhance situational awareness and accessibility. This study introduces an interpretable, sound-based machine learning framework to detect vehicle faults and emergency sound events using acoustic signals as a scalable diagnostic source. Three purpose-built datasets were developed: one for vehicular fault detection, another for emergency and environmental sounds, and a third integrating both to reflect real-world ITS acoustic scenarios. Audio data were preprocessed through normalization, resampling, and segmentation and transformed into numerical vectors using Mel-Frequency Cepstral Coefficients (MFCCs), Mel spectrograms, and Chroma features. To ensure performance and interpretability, feature selection was conducted using SHAP (explainability), Boruta (relevance), and ANOVA (statistical significance). A two-phase experimental workflow was implemented: Phase 1 evaluated 15 classical models, identifying ensemble classifiers and multi-layer perceptrons (MLPs) as top performers; Phase 2 applied advanced feature selection to refine model accuracy and transparency. Ensemble models such as Extra Trees, LightGBM, and XGBoost achieved over 91% accuracy and AUC scores exceeding 0.99. SHAP provided model transparency without performance loss, while ANOVA achieved high accuracy with fewer features. The proposed framework enhances accessibility by translating auditory alarms into visual/haptic alerts for hearing-impaired drivers and can be integrated into smart city ITS platforms via roadside monitoring systems. Full article

Reactive oxygen species (ROS) function as critical signaling molecules in cancer biology, promoting proliferation, angiogenesis, and metastasis at controlled levels while inducing lethal damage when exceeding the cell’s buffering capacity. To survive under this state of chronic oxidative stress, cancer cells become dependent on a hyperactive antioxidant shield, primarily orchestrated by the Nrf2, glutathione (GSH), and thioredoxin (Trx) systems. These defenses maintain redox homeostasis and sustain oncogenic signaling, notably through the oxidative inactivation of tumor-suppressor phosphatases, such as PTEN, which drives the PI3K/AKT/mTOR pathway. Targeting this addiction to a rewired redox state has emerged as a compelling therapeutic strategy. Pro-oxidant therapies aim to overwhelm cellular defenses, with agents like high-dose vitamin C and arsenic trioxide (ATO) showing significant tumor-selective toxicity. Inhibiting the master regulator Nrf2 with compounds such as Brusatol or ML385 disrupts the core antioxidant response. Disruption of the GSH system by inhibiting cysteine uptake with sulfasalazine or erastin potently induces ferroptosis, a non-apoptotic cell death driven by lipid peroxidation. Furthermore, the thioredoxin system is targeted by the repurposed drug auranofin, which irreversibly inhibits thioredoxin reductase (TrxR). Extensive preclinical data and ongoing clinical trials support the concept that this reliance on redox adaptation is a cancer-selective vulnerability. Moreover, novel therapeutic strategies, including the expanding field of redox-active metal complexes, such as manganese porphyrins, which strategically leverage the differential redox state of normal versus cancer cells through both pro-oxidant and indirect Nrf2-mediated antioxidative mechanisms (triggered by Keap1 oxidation), with several agents currently in advanced clinical trials, have also been discussed. Essentially, pharmacologically tipping the redox balance beyond the threshold of tolerance offers a rational and powerful approach to eliminate malignant cells, defining a novel frontier for targeted cancer therapy. Full article

Tau protein, a neuron-enriched microtubule-associated protein encoded by the MAPT gene, plays pivotal roles in microtubule stabilisation, axonal transport, and synaptic plasticity. Aberrant post-translational modifications (PTMs), hyperphosphorylation, acetylation, ubiquitination, oxidative stress and neuroinflammation disrupt tau’s normal functions, drive its mislocalization, and promote aggregation into neurofibrillary tangles, a hallmark of Alzheimer’s disease (AD) and related tauopathies. Over the past two decades, tau-targeted therapies have advanced into clinical development, yet most have failed to demonstrate efficacy in human trials. This review synthesises mechanistic insights into tau biology and pathology, highlighting phosphorylation and acetylation pathways, aggregation-prone motifs, and immune-mediated propagation. We analyse the current therapeutic landscape, including kinase and phosphatase modulators, O-GlcNAcase inhibitors, aggregation blockers, immunotherapies, and microtubule-stabilising agents, while examining representative clinical programs and the reasons underlying their limited success. By combining mechanistic understanding with clinical experience, this review outlines emerging opportunities for rational treatment development, aiming to inform future tau-targeted strategies for AD and other tauopathies. Full article

This study examined the spatiotemporal dynamics and driving mechanisms of ecosystem service value (ESV) in Wuhan from 1985 to 2020. Using multi-temporal land-use data, remotely sensed vegetation indices, and socioeconomic statistics, we estimated the ESV with an improved equivalent-factor method and analyzed its drivers using a Geodetector and geographically weighted regression (GWR). Over the 35-year period, total ESV for Wuhan showed a mildly declining trend, decreasing from CNY 37.464 billion in 1985 to CNY 36.439 billion in 2020. Waterbodies contributed the largest share of ESV, followed by croplands and forests. In the urban core, ESV declined significantly, with low-value zones expanding outward from the city center. Spatial autocorrelation analysis revealed significant “high–high” and “low–low” clustering. Geodetector results indicated slope, elevation, and normalized difference vegetation index (NDVI) as the primary natural drivers, with human footprint, gross domestic product (GDP), and population density acting as important socioeconomic auxiliaries. Interactions between natural and socioeconomic factors substantially increased the explanatory power. Furthermore, GWR revealed pronounced spatial heterogeneity in the sign and magnitude of the factor effects across the study area, underscoring the complexity of ESV drivers. These findings provide quantitative evidence to support spatially differentiated ecological planning and conservation strategies during urbanization in Wuhan and the broader mid-Yangtze region. Full article

Vehicle-to-everything (V2X) communication, a current key area of research, has a large impact on traffic safety, traffic efficiency, autonomous driving technology development, and intelligent transport. In order to achieve the low-latency performance and high transmission efficiency required for V2X communication, channel estimation for transmission channels is particularly important. In this regard, this paper proposes an improved general orthogonal match pursuit (GOMP) channel estimation algorithm based on the base extension model for an orthogonal time frequency space (OTFS) system. Firstly, the channel matrix is decomposed using the basis expansion model. Then, the strong sparsity of the basis function is exploited for channel estimation using the GOMP algorithm, while the ordinal difference restriction method and the weak selectivity principle are introduced to improve the system. The obtained improved GOMP algorithm not only shows a greater improvement in terms of normalized mean square error (NMSE) and bit error rate (BER) performance but also greatly reduces computational complexity, enabling it to better satisfy the needs of V2X communication. Full article

Fibrous sheath interacting proteins 1 and 2 (FSIP1 and FSIP2) are evolutionarily conserved testis-specific antigens, exclusively expressed in germ cells of adult human tissues, where they play essential roles in spermatogenesis and testicular development. Aberrant re-expression of FSIP1 and FSIP2, however, has been frequently reported in multiple malignancies, driving oncogenic processes including uncontrolled proliferation, invasion, migration, and metastasis, and correlating with unfavorable clinical outcomes. Their restricted expression in normal tissues, together with their consistent association with poor prognosis across cancer types, highlights their potential as diagnostic biomarkers, therapeutic targets, and prognostic indicators. This review summarizes the structural features and biological functions of the FSIP family, emphasizes recent advances in elucidating their regulatory roles in tumor-associated signaling pathways, and outlines the major challenges and future perspectives in this emerging field. Full article

Background: Calvarial venous malformations (VMs) are rare and genetically understudied. While somatic TEK receptor tyrosine kinase (TEK) mutations drive sporadic VMs, their role in scalp–calvarial VMs is unknown. We report the first pediatric case of a calvarial VM with a pathogenic somatic TEK mutation and its molecular implications. Methods: A 16-year-old female with a symptomatic parietal scalp VM underwent neurosurgical resection. Exome sequencing was performed on both lesional and blood DNA. Single-cell RNA sequencing (scRNA-seq) data from normal brain vasculature were analyzed for TEK expression and pathway enrichment. Results: A novel somatic TEK L914F mutation (chr9:27212760-C-T [GRCh38]), absent in germline DNA and population databases, was identified and predicted to be deleterious (CADD: 24). scRNA-seq data analysis revealed TEK enrichment in endothelial cells, particularly in fetal and arterial subtypes, and implicated angiogenesis and PI3K/Rho signaling as potential downstream phenotypic and molecular consequences. Conclusions: This first pediatric scalp VM with a somatic TEK L914F mutation expands the phenotypes associated with TEK-related vascular anomalies. These findings emphasize the role of somatic TEK mutation in diverse VMs and support genetic testing in sporadic cases. Further studies are needed to define therapeutic targets. Full article

A fast-paced lifestyle contributes to heightened emotional stress, driving the demand for milder and safer cosmetic ingredients that can counteract stress-induced skin damage—a focus of cutting-edge research in the field. Aim: The aim was to elucidate the role and mechanistic basis of tea (Camellia sinensis) seed meal saponin (Sap) in regulating stress-induced sebum overproduction and inflammatory responses. Methods: The composition and chemical structure of Sap were analyzed using UV-vis absorption spectroscopy, Fourier-transform infrared spectroscopy (FT-IR), and ultra-high-performance liquid chromatography–mass spectrometry (UHPLC-MS). In vitro models of cortisone-induced excessive lipid accumulation and the tumor necrosis factor-alpha (TNF-α)-stimulated inflammatory models were established on sebaceous gland cells (SZ95) and normal human epidermal keratinocytes (NHEKs), respectively. Cortisol and inflammatory cytokine secretion levels in cells were detected using ELISA. Additionally, the signaling pathways were revealed by Western blot (WB) and real-time quantitative polymerase chain reaction (RT-PCR). Results: Five saponins were identified in the Sap extract, all belonging to the oleanolic-acid-type pentacyclic triterpenes. Sap treatment significantly attenuated cortisone-induced cortisol secretion and lipid accumulation in SZ95 sebocytes. Mechanistically, Sap inhibited the 11β-HSD1/SREBP-1 pathway, which mediates its sebosuppressive effects, while concurrently down-regulating the mRNA expression of key downstream transcription factors and enzymes, including SREBP-1, FAS, and ACC. Additionally, Sap treatment significantly attenuated TNF-α-stimulated cortisol secretion and inflammatory cytokine (IL-1β, IL-6, and IL-8) production in NHEK cells through the inhibition of the 11β-HSD1/TLR2/NF-κB signaling pathway. Conclusion: Sap demonstrated dual inhibitory effects, suppressing both emotional-stress-induced sebum overproduction and inflammatory cytokines secretion. Full article

Background/Objectives: Glutamatergic neurotransmission is essential for the normal functioning of the retina. Photoreceptor to bipolar and bipolar to ganglion cell signaling is mediated by L-glutamate, which is stored in and released from vesicular glutamate transporter 1 (VGLUT1) containing synaptic vesicles. VGLUT1 is expressed postnatally, P2 onwards, and is required for the glutamatergic retinal wave observed between P10 and P12 in the developing mouse retina. P9–P13 postnatal age is critical for retinal development as VGLUT1 expressing ribbon synapses activate in the outer and inner plexiform layers, and rod/cone mediated visual signaling commences in that period. Although it has been hypothesized that glutamatergic extrinsic signaling drives cell cycle exit and initiates cellular differentiation in the developing retina, it is not clear whether intracellular, synaptic, or extrasynaptic vesicular glutamate release contributes to this process. Recent studies have attempted to decipher VGLUT’s role in retinal development. Here, we investigate the potential effect of genetic loss of VGLUT1 on early postnatal histogenesis and development of retinal neural circuitry. Methods: We employed immunohistochemistry and electrophysiology to ascertain the density of glutamatergic, cholinergic, and dopaminergic cells, spontaneous retinal activity, and light responses in VGLUT1 null retina, and contrasted them with wildtype (WT) and melanopsin null retina. Results: We have demonstrated here that VGLUT1 null retina shows signs of age dependent retinal degeneration, similar to other transgenic mice models with dysfunctional photoreceptor to bipolar cell synapses. The loss of VGLUT1 specifically alters glutamatergic cell density and morphological maturation of retinal ganglion cells. Moreover, VGLUT2 expression is lost in the majority of VGLUT2 cones in the absence of VGLUT1 coexpression, except when VGLUT2 coexpresses transiently with VGLUT3 in these cones, or when VGLUT1 null mice are dark reared. Conclusions: We present the first evidence that synaptic or extrasynaptic postnatal glutamate release from VGLUT1 containing vesicles impacts histogenesis of glutamatergic cells, pruning of retinal ganglion cell dendrites and VGLUT2 expression in cones. Full article

On 2 December 2020, an earthquake with a magnitude of M_w 4.5 occurred near the city of Thiva (Greece). The aftershock sequence, triggered by ruptures on or near the Kallithea fault, continued until January 2021. Seven months later, new seismic activity began a few kilometers west of the initial events, with the swarm displaying a general trend of spatiotemporal migration toward the east–southeast until the middle of 2022. In order to understand the physical and statistical pattern of the swarm, the seismicity was relocated using HypoDD, and the magnitude of completeness was determined using the frequency–magnitude distribution. In order to define the existence of spatiotemporal seismicity clusters in an objective way, the DBSCAN clustering algorithm was applied to the 2020–2022 Thiva earthquake sequence. The extracted clusters permit the analysis of the spatiotemporal scaling properties of the main clusters using the Non-Extensive Statistical Physics (NESP) approach, providing detailed insights into the nature of the long-term correlation of the seismic swarm. The statistical pattern observed aligns with a Q-exponential distribution, with q_D values ranging from 0.7 to 0.8 and q_T values from 1.44 to 1.50. Furthermore, the frequency–magnitude distributions were analyzed using the fragment–asperity model proposed within the NESP framework, providing the non-additive entropic parameter (q_M). The results suggest that the statistical characteristics of earthquake clusters can be effectively interpreted using NESP, highlighting the complexity and non-additive nature of the spatiotemporal evolution of seismicity. In addition, the analysis of the properties of the seismicity clusters extracted using the DBSCAN algorithm permits the suggestion of possible physical mechanisms that drive the evolution of the two main and larger clusters. For the cluster that activated first and is located in the west–northwest part, an afterslip mechanism activated after the 2 September 2021, M 4.0 events seems to predominately control its evolution, while for the second activated cluster located in the east–southeast part, a normal diffusion mechanism is proposed to describe its migration pattern. Concluding, we can state that in the present work the application of the DBSCAN algorithm to recognize the existence of any possible spatiotemporal clustering of seismicity could be helping to provide detailed insight into the statistical and physical patterns in earthquake swarms. Full article

Background: Gastrointestinal (GI) cancers are common and pose a major public health issue. An inflammatory microenvironment drives their development and progression. The α7nAChR receptor, known to suppress autoimmune and inflammatory bowel diseases, is also linked to colorectal cancer. It enhances anti-inflammatory activity, influences tumor growth, metastasis, and treatment response, and is associated with tobacco use. NLRP3, a key inflammatory mediator, connects immunity and cancer. The α7nAChR receptor modulates tumorigenesis and therapy response by suppressing inflammatory pathways, while also regulating NLRP3 inflammasome activation through inhibition of mitochondrial DNA release. This study examines α7nAChR and NLRP3 expression in gastric and colorectal cancers, colitis, and normal tissues to clarify pathogenic mechanisms and identify therapeutic targets. Methods: Tissue samples of gastric tumor (S-Tm) (n = 10), colorectal tumor (C-Tm) (n = 10), colitis (UC) (n = 10), healthy stomach (S-C) (n = 10) and healthy colorectal tissue (C-C) (n = 10) taken during routine endoscopy protocols were homogenized. The α7nAChR and NLRP3 levels were examined using the ELISA method, and groups were compared. Results: We identified statistically significant differences in α7nAChR levels between the S-C and S-Tm (p < 0.05), C-C and C-Tm (p < 0.05), and S-C and C-Tm (p < 0.001) groups. The NRLP3 levels also differed significantly between the UC and C-Tm (p < 0.05), the S-C and C-Tm (p < 0.01), and the C-C and C-Tm groups (p < 0.01). Conclusions: Paradoxically, given the inflammatory regulatory role and oncogenic effects of α7nAChR, the relationship between α7nAChR and NLRP3 has become an important target for both oncological and inflammatory therapeutic approaches, particularly in inflammation-related GI cancers. Full article