The International Pediatric Stroke Study: Insight into Childhood Stroke from a Developmental Perspective.

AB-654. Analysis of clinical and electrophysiologic features of contactin 1 antibody-mediated autoimmune nodopathy

• This work presents a novel idea in the field of forest fire detection and addresses the critical limitations of existing bias mitigation approaches. • The proposed approach is able to handle the complex interaction of environmental factors and adapts quickly to quickly changing forest fire scenarios. • The proposed approach uses the complex relationships seen between meteorological variables, generative adversarial networks and data fusion to mitigate bias. • The proposed approach addresses comprehensive bias mitigation through the analysis of both high-level and low-level image features, which in turn significantly improve the specificity and accuracy in forest fire detection. Imbalanced data sets exacerbate recognition biases in forest fire prediction models, as disproportionate representation of class instances leads to skewed results. Existing work on bias mitigation has limited ability to generalize and extract features specific to forest fires. Internet of Things (IoT)-based sensor networks can provide real-time, granular data on environmental factors such as temperature, humidity, and soil moisture, helping to capture the dynamic nature of forest conditions and alleviate data imbalance. To address these challenges, this work introduces a novel hybrid approach that explores complex probabilistic relationships among environmental factors, incorporating IoT-driven data, and using a generative adversarial network (GAN) to synthetically augment minority classes. The proposed model is validated on publicly available datasets, and the performance is reported on evaluation metrics such as accuracy, precision, recall, F1-score, computational efficiency and training cost. The results show that the proposed hybrid model is able to achieve a significant improvement over the exiting methods achieving classification accuracy of 95.08%, a precision of 93.03%, a recall of 92.80%, and an F1-score of 92.91%.

Multiple sclerosis (MS) is a chronic inflammatory disorder of the central nervous system, where timely and accurate diagnosis is essential for effective management. Optical coherence tomography (OCT) enables non-invasive evaluation of retinal changes that may serve as biomarkers for MS. Unlike other ophthalmologic diseases, raw cross-sectional OCT images in MS show subtle alterations often indistinguishable from healthy controls (HCs). Consequently, retinal layer thickness and boundary-derived surface features offer greater discriminatory power. We investigated three categories of artificial intelligence (AI) models: (1) feature extraction with auto-encoder (AE) and shallow networks, (2) custom-designed deep networks, and (3) fine-tuned pre-trained networks. Retinal layer thickness and surface maps derived from OCT were analyzed to determine the most informative features, with channel-wise combination and mosaicing applied for feature integration. Model interpretability was assessed using occlusion sensitivity and Gradient-weighted Class Activation Mapping (Grad-CAM) visualizations. The dataset included 38 HC and 78 MS eyes obtained from independent public and local sources. Patient-wise partitioning was implemented to prevent data leakage. The proposed deep network using channel-wise combined thickness maps of retinal nerve fiber layer (RNFL), ganglion cell and inner plexiform layer (GCIPL), and inner nuclear layer (INL) layers achieved balanced accuracy of 97.3% (SD=4.16; 95% CI: 92.3-100%), specificity of 97.3% (SD=5.59; 95% CI: 92.6-100%), sensitivity of 97.4% (SD=3.54; 95% CI: 92.6-100%), g-mean of 97.3% (SD=4.18; 95% CI: 92.24-100%), F1-score of 98.0% (SD=3.86; 95% CI: 92.6-100%), and an AUC of 0.96 (SD=0.08; 95% CI: 0.95-1.00). Notably, the high performance observed in internal cross-validation was achieved when public and local datasets were combined. However, performance decreased substantially in cross-dataset evaluations, where models were trained on one dataset and tested on the other, indicating limited external generalizability, particularly when trained on public data and applied to local clinical data. AI-based analysis of OCT-derived retinal layer features enables accurate and interpretable classification of MS, supporting its potential as a valuable clinical biomarker.

This paper presents the study of using the Wavelet Transform (WT) for Particle Induced X-ray Emission (PIXE) analysis. Wavelet transform is a mathematical method for the decomposition of the signal into time and frequency components. In a PIXE spectrum, the background is at low frequency, while the signal consists in the medium frequencies. Therefore, decomposition into different frequencies can be used to estimate the background. In this paper, the method of the Dual-Tree Complex Wavelet Transform (DTCWT) for background removal and also common data conditioning operations, such as peak finding, are investigated for their usage to Cr-doped UO 2 spectrum analysis. Employed algorithms enable the extraction and analysis of the weak features that are hidden within high-level backgrounds. Additionally, the PIXEK code is presented - a new tool using wavelet transform dedicated to PIXE analysis.

Gear surface defect detection with hierarchical multi-scale feature analysis and dynamic attention enhancement network

Identification and quantification of approval desire in social networking service posts and analysis of their linguistic features.

Improving rainfall-triggered landslide susceptibility mapping through source-area boundary sampling and multi- dimensional feature analysis

UAV-based deep learning for biodiversity monitoring: Advances, applications, and future directions

Functional slow feature analysis based in-situ monitoring of crystallization process via near-infrared spectrum

Fine extraction and morphological feature analysis of tidal creeks in the Yellow River Delta based on Gaofen-2 imagery

A resource-friendly indoor localization approach using spatial feature analysis of WiFi signal fingerprints

Correlative analysis between ocular surface features and carotid plaque : A multimodal machine learning framework.

Multi-branch convolutional neural network and intracranial EEG high-frequency oscillations predict post-surgical seizure outcomes.

Poincaré feature-based classification of electroencephalography signals for multiple sclerosis diagnosis.

WAD-YOLO (Wavelet-Based Adaptive Defect Detection with Multi-Resolution Feature Analysis) is an intelligent system for automatic steel surface defect detection. The project combines wavelet transform techniques with the YOLO deep learning model for accurate and real-time inspection. Wavelet decomposition extracts multi-resolution features to highlight defects of different sizes and textures. Adaptive enhancement improves image clarity and reduces noise effects. The processed features are fed into the YOLO network for defect localization and classification. The system detects various defects such as cracks, scratches, pits, and rolled-in scale. Multi-resolution analysis ensures better detection of both small and large surface defects. The model improves detection accuracy compared to traditional single-scale methods. It supports real-time industrial inspection on production lines. Overall, WAD-YOLO enhances quality control efficiency in steel manufacturing industries.

The article is devoted to the development of a conceptual design for Thomson scattering diagnostics (TSD) for the KTM tokamak, intended for conducting experimental studies, measuring the dynamics of temperature profiles and the electron concentration of high-temperature plasma. The work shows an extensive literature review of TSD implemented on similar installations. Schemes for plasma probing and collection of scattered laser radiation of TSD are presented. Based on an analysis of the structural features of the KTM tokamak and existing technical constraints, the choice of a tangential diagnostic layout in the equatorial plane has been justified as the most reliable and practically feasible option. Key technical requirements for the Thomson scattering diagnostic system have been defined: an electron temperature measurement range of 50 to 3100 eV, an electron density range of 1·10 19 to 2.5·10 20 m −3 , and coverage of at least 10 spatial observation points. Various diagnostic configurations, including vertical viewing schemes, were analyzed, and their implementation on KTM was shown to be infeasible due to a number of engineering and physical limitations. The proposed configuration ensures alignment stability, protection against parasitic radiation, and high spatial resolution. This work represents a well-substantiated design of a Thomson scattering diagnostic system, the implementation of which will significantly enhance the diagnostic capabilities of the KTM tokamak for plasma and material research under controlled thermonuclear fusion conditions.

Due to limitations in large-scale toxicity assessment, the actual biological toxicity of wastewater effluents remains insufficiently characterized. The nematode Caenorhabditis elegans is a well-established model organism for evaluating whole effluent toxicity (WET). However, its standardized methods (e.g., ISO 10872) rely on time-consuming manual quantification, hindering large-scale toxicity assessment for decision-making in wastewater risk management. Herein, a model-driven high-throughput assay was developed that integrates spatiotemporal analysis of the nematode behavioral features with machine learning, reducing WET testing time by ∼77% compared to standard methods and enabling a comprehensive risk assessment of nationwide wastewater treatment plants (WWTPs) across China. The results showed that WWTP effluents consistently showed high toxicity (toxicity unit [TU] = 0.47-2.16), even when meeting permissible discharge limits for chemical indicators, substantially exceeding the toxicity of the corresponding receiving waters (p < 0.05, ANOVA). Interestingly, WWTP treatment capacity emerged as the predominant driver of WET variation, underscoring the need to prioritize large-size WWTPs in flexible wastewater risk control strategies. These findings expose a significant gap between wastewater risk management needs and current control practices, as WWTP effluents showed substantially higher toxicity than their receiving waters, advocating for the scale-prioritized toxicity-driven discharge standards to secure more safe and efficient water sustainability management in China.

Clinically actionable biomarkers that accurately reflect the health status of the beta cell are needed to improve risk stratification and optimise the timing of interventions in type 1 diabetes. We hypothesised that inflammatory stress elicits a reproducible microRNA (miRNA) program in human islets and islet-derived extracellular vesicles (EVs) that can be detected in plasma EVs to stratify diabetes risk, while also providing insight into molecular pathways linked to beta cell dysfunction. Human islets were exposed to IL-1β+IFN-γ, and small RNA-seq was performed on islets and islet-derived EVs. Differentially expressed miRNAs were validated in islets, using RT-PCR, in plasma-derived EVs from individuals with autoantibody positivity (AAb+) or recent-onset type 1 diabetes and matched control individuals using ultrasensitive, label-free localised surface plasmon resonance (LSPR) biosensors, and in pancreatic sections from organ donors using in situ hybridisation and spatial feature analysis. Finally, beta cell-targeted in vivo inhibition of miR-155 was tested in the NOD mouse model. Inflammatory cytokine exposure altered a restricted subset of miRNAs, identifying 20 differentially expressed miRNAs in islets and 14 in islet-derived EVs. Only two miRNAs, miR-155-5p and miR-146a-5p, were concordantly upregulated in both compartments. Machine learning prioritised an EV miRNA panel for translational validation, and custom LSPR biosensors enabled quantification of these miRNAs in plasma EVs. This plasma EV miRNA signature, consisting of miR-155-5p, miR-146a-5p, miR-30c-1-3p, miR-802 and miR-124-3p, differentiated individuals with AAb+ and those with recent-onset type 1 diabetes from control individuals with good sensitivity and specificity. In pancreatic tissue, miR-155 abundance and beta cell spatial/subcellular distribution were altered in donors with AAb+ and type 1 diabetes compared with non-diabetic control individuals. Functionally, beta cell-targeted inhibition of miR-155 improved glucose tolerance and reduced insulitis in prediabetic NOD mice. Using an organ-based model system of inflammatory stress, we validated a signature of EV-associated miRNAs capable of stratifying type 1 diabetes risk. Furthermore, we provided new mechanistic and imaging insights into miRNA expression patterns in pancreatic sections from human organ donors with type 1 diabetes or AAb+, and we used a preclinical model of type 1 diabetes to demonstrate the potential therapeutic efficacy of targeting these miRNAs. The data from small RNA sequencig of human islets and islet-derived EVs have been deposited in the GEO database (accession no. GSE160391).

Y chromosome short tandem repeat (Y-STR) is an important genetic marker in forensic practices. Length-based Y-STR genotyping method has been used to screen paternal lineages and successfully solved many serious cases. However, it is hard to discriminate paternal lineages with similar or identical length-based Y-STR genotypes. Next-generation sequencing-based Y-STR genotyping method could be used to solve the problem, and already be applied to criminal scene investigations. Nevertheless, previously studied data were inadequate and scattered, and sequence features of Y-STR loci were insufficiently summarized, which hindered the deep forensic application of Y-STR loci. Here, we review the sequence features of repeat and flanking regions on 41 widely used forensic Y-STR loci based on public literature data. Furthermore, we identify haplogroup-associated Y chromosome single nucleotide polymorphisms within these regions and explore potential applications of sequence-based polymorphisms. This review is expected to serve as a valuable reference for paternal lineage discrimination and paternal biogeographic ancestry inference using Y-STR sequence features.