Search Results (916)

Mixed pixels remain a central obstacle to reliable endmember extraction from hyperspectral imagery. We present AMEE–PPI, a hybrid method that embeds the Pure Pixel Index (PPI) within morphological structuring elements and propagates spectral purity via dilation/erosion, thereby coupling spatial context with spectral cues while avoiding a user-fixed number of projections. On GaoFen-5 (GF-5) AHSI data from a geologically complex outcrop region, we benchmark AMEE–PPI against four widely used algorithms—PPI, OSP, VCA, and AMEE. The pipeline uses HySime for noise estimation and signal-subspace inference to set the endmember count prior to extraction and applies morphological elements spanning 3 × 3 to 15 × 15 to balance spatial support with local heterogeneity. Quantitatively, AMEE–PPI achieves the lowest spectral angle distance (SAD) for all outcrop types—purple–red: 0.135; yellow–brown: 0.316; gray: 0.191—surpassing the competing methods. It also attains the lowest spectral information divergence (SID)—purple–red: 0.028; yellow–brown: 0.184; gray: 0.055—confirming superior similarity to field reference spectra across materials. Visually, AMEE–PPI avoids the vegetation endmember leakage observed with several baselines on purple–red and gray outcrops, yielding cleaner, more representative endmembers. These results indicate that integrating spatial morphology with spectral purity improves robustness to illumination, mixing, and local variability in GF-5 imagery, with direct benefits for downstream unmixing, classification, and geological interpretation. Full article

Crop diseases cause significant losses in agricultural production; early capture and identification of disease spores enable disease monitoring and prevention. This study experimentally optimized the preparation of Au@Ag NRS (Gold core@Silver shell Nanorods) sol as a Surface-Enhanced Raman Scattering (SERS) enhancement reagent via a modified seed-mediated growth method. Using an existing microfluidic chip developed by the research group, disease spores were separated and enriched, followed by combining Au@Ag NRS with Crop Disease Spores through electrostatic adsorption. Raman spectroscopy was employed to collect SERS fingerprint spectra of Crop Disease Spores. The spectra underwent baseline correction using Adaptive Least Squares (ALS) and standardization via Standard Normal Variate (SNV). Dimensionality reduction preprocessing was performed using Principal Component Analysis (PCA) and Successive Projections Algorithm combined with Competitive Adaptive Reweighted Sampling (SCARS). Classification was then executed using Support Vector Machine (SVM) and Multilayer Perceptron (MLP). The SCARS-MLP model achieved the highest accuracy at 97.92% on the test set, while SCARS-SVM, PCA-SVM, and SCARS-MLP models attained test set accuracy of 95.83%, 95.24%, and 96.55%, respectively. Thus, the proposed Au@Ag NRS-based SERS technology can be applied to detect airborne disease spores, establishing an early and precise method for Crop Disease detection. Full article

Microplastics pose a growing environmental concern, necessitating accurate and scalable methods for their detection and classification. This study presents a novel deep learning framework that integrates a transformer-based architecture with domain adaptation techniques to classify microplastics using reflectance micro-FTIR spectroscopy. A key challenge addressed in this work is the domain shift between laboratory-prepared reference spectra and environmentally sourced spectra, which can significantly degrade model performance. To overcome this, three domain-adaptation strategies—Domain Adversarial Neural Networks (DANN), Deep Subdomain-Adaptation Networks (DSAN), and Deep CORAL—were evaluated for their ability to enhance cross-domain generalization. Experimental results show that while DANN was unstable, DSAN and Deep CORAL improved target domain accuracy. Deep CORAL achieved 99% accuracy on the source and 94% on the target, offering balanced performance. DSAN reached 95% on the target but reduced source accuracy. Overall, statistical alignment methods outperformed adversarial approaches in transformer-based spectral adaptation. The proposed model was integrated into a reflectance micro-FTIR workflow, accurately identifying PE and PP microplastics from unlabelled spectra. Predictions closely matched expert-validated results, demonstrating practical applicability. This first use of a domain-adaptive transformer in microplastics spectroscopy sets a benchmark for high-throughput, cross-domain analysis. Future work will extend to more polymers and enhance model efficiency for field use. Full article

Trichophyton indotineae is an emerging dermatophyte species responsible for recalcitrant and terbinafine-resistant dermatophytosis, raising concerns over diagnostic accuracy and treatment efficacy. This study aimed to improve the identification and resistance profiling of T. indotineae by integrating molecular methods with machine learning-assisted analysis of MALDI-TOF mass spectra. A total of 56 clinical isolates within the Trichophyton mentagrophytes complex were analyzed using ITS and ERG1 gene sequencing, antifungal susceptibility testing, and MALDI-TOF MS profiling. Terbinafine resistance was detected in 23 isolates and correlated with specific ERG1 mutations, including F397L, L393S, F415C, and A448T. While conventional MALDI-TOF MS failed to reliably distinguish T. indotineae from closely related species, unsupervised statistical methods (PCA and hierarchical clustering) revealed distinct spectral groupings. Supervised machine learning algorithms, particularly PLS-DA and SVM, achieved 100% balanced accuracy in species classification using 10-fold cross-validation. Biomarker analysis identified discriminatory spectral peaks for both T. indotineae and T. mentagrophytes (3417.29 m/z and 3423.53 m/z). These results demonstrate that combining MALDI-TOF MS with multivariate analysis and machine learning improves diagnostic resolution and may offer a practical alternative to sequencing in resource-limited settings. This approach could enhance the routine detection of terbinafine-resistant T. indotineae and support more targeted antifungal therapy. Full article

Objectives: This study presents a versatile, AI-guided workflow for the targeted isolation and characterization of prenylated flavonoids from Paulownia tomentosa (Thunb.) Steud. (Paulowniaceae). Methods: The approach integrates established extraction and chromatography-based fractionation protocols with LC-UV-HRMS/MS analysis and supervised machine-learning (ML) custom-trained classification models, which predict prenylated flavonoids from LC-HRMS/MS spectra based on the recently developed Python package AnnoMe (v1.0). Results: The workflow effectively reduced the chemical complexity of plant extracts and enabled efficient prioritization of fractions and compounds for targeted isolation. From the pre-fractionated plant extracts, 2687 features were detected, 42 were identified using reference standards, and 214 were annotated via spectra library matching (public and in-house). Furthermore, ML-trained classifiers predicted 1805 MS/MS spectra as derived from prenylated flavonoids. LC-UV-HRMS/MS data of the most abundant presumed prenyl-flavonoid candidates were manually inspected for coelution and annotated to provide dereplication. Based on this, one putative prenylated (C5) dihydroflavonol (1) and four geranylated (C10) flavanones (2–5) were selected and successfully isolated. Structural elucidation employed UV spectroscopy, HRMS, and 1D as well as 2D NMR spectroscopy. Compounds 1 and 5 were isolated from a natural source for the first time and were named 6-prenyl-4′-O-methyltaxifolin and 3′,4′-O-dimethylpaulodiplacone A, respectively. Conclusions: This study highlights the combination of machine learning with analytical techniques to streamline natural product discovery via MS/MS and AI-guided pre-selection, efficient prioritization, and characterization of prenylated flavonoids, paving the way for a broader application in metabolomics and further exploration of prenylated constituents across diverse plant species. Full article

Electrochemical impedance spectroscopy (EIS) is a technique used to analyze the kinetics and interfacial processes of electrochemical systems. The selection of an appropriate equivalent circuit model for EIS interpretation was traditionally reliant on expert experience, rendering the process subjective and prone to error. To address these limitations, an automated framework for both model selection and parameter estimation was proposed. The methodology was structured such that initial model screening was performed by a global heuristic search algorithm, adaptive optimization was guided by an integrated XGBoost-based error feedback mechanism, and precise parameter estimation was achieved using a Differential Evolution–Levenberg–Marquardt (DE-LM) algorithm. When evaluated on a purpose-built dataset comprising 4.8 × 10⁵ spectra across diverse circuit and biofilm scenarios, a model classification accuracy of 96.32% was achieved, and a 72.3% reduction in parameter estimation error was recorded. The practical utility of the method was validated through the quantitative analysis of bovine serum albumin–Clenbuterol hydrochloride (BSA-CLB), wherein an accuracy of 95.2% was demonstrated and a strong linear correlation with target concentration (R² = 0.999) was found. Through this approach, the limitations of traditional black-box models were mitigated by resolving the physical meaning of parameters. Consequently, the automated and quantitative monitoring of processes such as biofilm formation was facilitated, enabling the efficient evaluation of antimicrobial drugs or anti-fouling coatings. Full article

Traffic accidents caused by black ice have become a serious public safety concern due to their high fatality rates and the limitations of conventional detection systems under low visibility. Millimeter-wave (mmWave) radar, capable of operating reliably in adverse weather and lighting conditions, offers a promising alternative for road surface monitoring. In this study, six representative road surface conditions—dry, wet, thin-ice, ice, snow, and sludge—were experimentally implemented on asphalt and concrete specimens using a temperature and humidity-controlled chamber. mmWave radar data were repeatedly collected to analyze the temporal variations in reflected signals. The acquired signals were transformed into range-based spectra using Range-Fast Fourier Transform (Range-FFT) and converted into statistical features and graphical representations. These features were used to train and evaluate classification models, including Extreme Gradient Boost (XGBoost), Light Gradient-Boosting Machine (LightGBM), Convolutional Neural Networks (CNN), and Vision Transformer (ViT). While machine learning models performed well under dry and wet conditions, their accuracy declined in hazardous states. Both CNN and ViT demonstrated superior performance across all conditions, with CNN showing consistent stability and ViT exhibiting competitive accuracy with enhanced global pattern-recognition capabilities. Comprehensive robustness evaluation under various noise and blur conditions revealed distinct characteristics of each model architecture. This study demonstrates the feasibility of mmWave radar for reliable road surface condition recognition and suggests potential for improvement through multimodal sensor fusion and time-series analysis. Full article

New technologies and equipment for medicine analysis and diagnostics have always been critical in clinical medication and pharmaceutical production. Especially in the field of traditional Chinese medicine (TCM) where the chemical composition is not fully clear, cross-device analysis and identification using the same technology can sometimes even lead to misjudgments. Akebia species, capable of inducing heat clearing, diuresis, and anti-inflammatory effects, show great potential in clinical applications. However, the three commonly used species differ in pharmacological effects and therefore should not be used interchangeably. We proposed a method combining LIBS with random forest for species identification and established a modeling and verification scheme across device platforms. Spectra of three Akebia species were collected using two LIBS systems equipped with spectrometers of different resolutions. The data acquired from the low-resolution spectrometer were used for model training, while the data from the high-resolution spectrometers were used for testing. A spectral correction and feature selection (SCFS) method was proposed, in which spectral data were first corrected using a standard lamp, followed by feature selection via analysis of variance (ANOVA) to determine the optimal number of discriminative features. The highest classification accuracy of 80.61% was achieved when 28 features were used. Finally, a post-processing (PP) strategy was applied, where abnormal spectra in the test set were removed using density-based spatial clustering of applications with noise (DBSCAN), resulting in a final classification accuracy of 85.50%. These results demonstrate that the proposed “SCFS-PP” framework effectively enhances the reliability of cross-instrument data utilization and expands the applicability of LIBS in the field of TCM. Full article

Bolted connections are extensively utilized in aerospace, civil, and mechanical systems for structural assembly. However, inevitable structural vibrations can induce bolt loosening, leading to preload reduction and potential structural failure. Early-stage preload degradation, particularly during initial loosening, is often undetectable by conventional monitoring methods due to limited sensitivity and poor noise resilience. To address these limitations, this study proposes an intelligent bolt preload monitoring framework that combines electromechanical impedance (EMI) signal analysis with a parallel deep learning architecture. A multiphysics-coupled model of flange joint connections is developed to reveal the nonlinear relationships between preload degradation and changes in EMI conductance spectra, specifically resonance peak shifts and amplitude attenuation. Based on this insight, a parallel convolutional neural network (P-CNN) is designed, employing dual branches with 1 × 3 and 1 × 7 convolutional kernels to extract local and global spectral features, respectively. The architecture integrates dilated convolution to expand frequency–domain receptive fields and an enhanced SENet-based channel attention mechanism to adaptively highlight informative frequency bands. Experimental validation on a flange-bolt platform demonstrates that the proposed P-CNN achieves 99.86% classification accuracy, outperforming traditional CNNs by 20.65%. Moreover, the model maintains over 95% accuracy with only 25% of the original training samples, confirming its robustness and data efficiency. The results demonstrate the feasibility and scalability of the proposed approach for real-time, small-sample, and noise-resilient structural health monitoring of bolted connections. Full article

The variability of B-type stars offers valuable insights into the interiors of stars and the processes that drive pulsation and rotation in massive stars. In this study, we present the classification of the variability of 197 B-type stars observed in various Kepler/K2 campaigns, including 73 newly classified stars from Campaigns 13–18. For these stars, we derived atmospheric and evolutionary parameters using space-based photometry and ground-based spectroscopy. We obtained spectroscopic data for 34 targets with high-resolution instruments at OPD/LNA, which were supplemented by archival LAMOST spectra. After correcting for instrumental systematics, we analyzed the light curves using Fourier transforms and wavelet decomposition to identify both periodic and stochastic signals. The identified variability types included SPB stars, $\beta$ Cephei/SPB hybrids, fast-rotating pulsators, stochastic low-frequency variables, eclipsing binaries, and rotational variables. We also revised classifications of misidentified stars using Gaia astrometry, confirming the main-sequence nature of objects once considered subdwarfs. Our results indicate that hot-star variability exists along a continuum shaped by mass, rotation, and internal mixing rather than distinct instability domains. This study enhances our understanding of B-type star variability and supports future asteroseismic modeling with missions like PLATO. Full article

Food fraud in meat products presents serious economic and public health challenges, underscoring the need for rapid and reliable detection methods. This study investigates the potential of Raman spectroscopy combined with machine learning to accurately discriminate between pure and mixed minced meat preparations. We evaluated three classification algorithms: Support Vector Machines (SVMs), Artificial Neural Networks (ANNs), and Random Forests (RFs). Raman spectra were collected from 19 distinct samples consisting of different ratios of pork, beef, and lamb minced meat. Our findings suggest that homogenization markedly enhances spectral consistency and classification accuracy. In the pure meat samples case, all three models (SVM, ANN, and RF) achieved notable increases in classification accuracies (from 0.50–0.70 to above 0.85), a dramatic improvement over unhomogenized samples. In more complex homogenized mixtures, SVM delivered the highest performance, achieving an accuracy of up to 0.88 for 50:50 mixtures and 0.86 for multi-ratio samples, often outperforming both ANN and RF. While the underlying interpretation of the classification models remains complex, the findings consistently underscore the critical role of homogenization on model performance. This work demonstrates the robust potential of the Raman spectroscopy-coupled machine learning approach for the rapid and accurate identification of minced meat species. Full article

Hydrocarbon fluids have a widespread presence in modern society due to their role in the global energy and fuel supply. The ability to distinguish between hydrocarbon fluids from different manufacturing processes is essential in industrial and government settings. Currently, performing such analyses is expensive and time-consuming, as standard practice involves sending samples to a laboratory for gas chromatography-mass spectrometry (GC-MS) analysis. The inherent limitations of traditional separation techniques often make them unsuitable for the demands of real-time process monitoring and control. This work proposes the use of handheld Raman spectroscopy for rapid classification of petroleum- and synthetic-based hydrocarbon fluids. A total of 600 Raman spectra were collected from six different hydraulic fluids and analyzed. Preliminary visual observations revealed reproducible spectral differences between various types of hydraulic fluids. Principal component analysis (PCA) and linear discriminant analysis (LDA) were used to investigate the data further. The findings indicate that handheld Raman spectrometers are capable of detecting chemical features of hydrocarbon fluids, supporting the classification of their formulations. Full article

Citrus diseases severely impact fruit yield and quality. To facilitate in-field, non-destructive, and rapid detection of citrus leaf physiological and pathological conditions, this study proposes a classification method for citrus leaf physiological and pathological statuses that integrates visible/near-infrared multispectral technology with deep learning. First, a handheld spectrometer was employed to acquire spectral images of five sample categories—Healthy, Huanglongbing, Yellow Vein Disease, Magnesium Deficiency and Manganese Deficiency. Mean spectral data were extracted from regions of interest within the 350–2500 nm wavelength range, and various preprocessing techniques were evaluated. The Standard Normal Variate (SNV) transformation, which demonstrated optimal performance, was selected for data preprocessing. Next, we innovatively introduced an adaptive spectral positional encoding mechanism into the Transformer framework. A lightweight, learnable network dynamically optimizes positional biases, yielding the ASTransformer (Adaptive Spectral Transformer) model, which more effectively captures complex dependencies among spectral features and identifies critical wavelength bands, thereby significantly enhancing the model’s adaptive representation of discriminative bands. Finally, the preprocessed spectra were fed into three deep learning architectures (1D-CNN, 1D-ResNet, and ASTransformer) for comparative evaluation. The results indicate that ASTransformer achieves the best classification performance: an overall accuracy of 97.7%, underscoring its excellent global classification capability; a Macro Average of 97.5%, reflecting balanced performance across categories; a Weighted Average of 97.8%, indicating superior performance in classes with larger sample sizes; an average precision of 97.5%, demonstrating high predictive accuracy; an average recall of 97.7%, showing effective detection of most affected samples; and an average F1-score of 97.6%, confirming a well-balanced trade-off between precision and recall. Furthermore, interpretability analysis via Integrated Gradients quantitatively assesses the contribution of each wavelength to the classification decisions. These findings validate the feasibility of combining a handheld spectrometer with the ASTransformer model for effective citrus leaf physiological and pathological detection, enabling efficient classification and feature visualization, and offer a valuable reference for disease detection of physiological and pathological conditions in other fruit crops. Full article

Cardiovascular disease (CVD) is a term used for disorders affecting the heart. Globally, it is the most common cause of death. The main purpose of this study was the rapid detection of CVD, which is essential for effective cure and inhibition. Early detection may lower the risk of myocardial infarction (MI) and reduce the death rate in CVD patients. Laser-induced breakdown spectroscopy (LIBS) is a non-invasive and less sample preparation technique for early detection of CVD. LIBS technique investigated the variation in intensities of different biochemical elements such as Calcium (Ca), Nitrogen (N), Sodium (Na), Carbon (C) and CN-band in the spectra of healthy and CVD patients. Machine learning algorithms applied to LIBS spectral data for the determination of validation accuracy and classification between CVD and healthy individuals. Several models achieved a perfect 100% highest accuracy, which showed the exceptional precision in the given configuration. The Narrow Neural Network achieved 100% accuracy on both the validation and test datasets in a short duration of 10.008 s. This preliminary research of LIBS combined with machine learning may provide a complementary method over existing analytical techniques for early detection of CVD. Full article

Focusing on the characteristic tea resource Zijuan tea, this study addresses the difficulty of grading on production lines and the complexity of quality evaluation. On the basis of the fusion of near-infrared (NIR) spectroscopy and visual features, a novel method is proposed for classifying different tenderness levels and quantitatively assessing key anthocyanin components in Zijuan tea fresh leaves. First, NIR spectra and visual feature data were collected, and anthocyanin components were quantitatively analyzed using UHPLC-Q-Exactive/MS. Then, four preprocessing techniques and three wavelength selection methods were applied to both individual and fused datasets. Tenderness classification models were developed using Particle Swarm Optimization–Support Vector Machine (PSO-SVM), Random Forest (RF), and Convolutional Neural Networks (CNNs). Additionally, prediction models for key anthocyanin content were established using linear Partial Least Squares Regression (PLSR), nonlinear Support Vector Regression (SVR) and RF. The results revealed significant differences in NIR spectral characteristics across different tenderness levels. Model combinations such as TEX + Medfilt + RF and NIR + Medfilt + CNN achieved 100% accuracy in both training and testing sets, demonstrating robust classification performance. The optimal models for predicting key anthocyanin contents also exhibited excellent predictive accuracy, enabling the rapid and nondestructive detection of six major anthocyanin components. This study provides a reliable and efficient method for intelligent tenderness classification and the rapid, nondestructive detection of key anthocyanin compounds in Zijuan tea, holding promising potential for quality control and raw material grading in the specialty tea industry. Full article