Search Results (541)

Alzheimer’s Disease (AD) is an advanced brain illness that affects millions of individuals across the world. It causes gradual damage to the brain cells, leading to memory loss and cognitive dysfunction. Although Magnetic Resonance Imaging (MRI) is widely used in AD diagnosis, the existing studies rely solely on the visual representations, leaving alternative features unexplored. The objective of this study is to explore whether MRI sonification can provide complementary diagnostic information when combined with conventional image-based methods. In this study, we propose a novel dual-stream multimodal framework that integrates 2D MRI slices with their corresponding audio representations. MRI images are transformed into audio signals using a multi-scale, multi-orientation Gabor filtering, followed by a Hilbert space-filling curve to preserve spatial locality. The image and sound modalities are processed using a lightweight CNN and YAMNet, respectively, then fused via logistic regression. The experimental results of the multimodal achieved the highest accuracy in distinguishing AD from Cognitively Normal (CN) subjects at 98.2%, 94% for AD vs. Mild Cognitive Impairment (MCI), and 93.2% for MCI vs. CN. This work provides a new perspective and highlights the potential of audio transformation of imaging data for feature extraction and classification. Full article

Heart rate variability (HRV) comprises several components driven by various internal processes, the least understood of which is the ultra-low frequency (ULF) one. Recently published research has shown that the HRV frequency distribution in this range is bimodal. The main aims of this work were to verify this finding, to determine the basic characteristics of these two components and to analyze their potential physiological couplings. For this purpose, two components within the conventional ULF band (below 4 mHz) were extracted from HRVs of 25 patients with apnea using adaptive variational mode decomposition (AVMD) and continuous wavelet transform (CWT), and then analyzed with the Hilbert transform (HT), Savitzky–Golay filter, and empirical distributions of instantaneous amplitudes and frequencies. These studies have demonstrated the existence of both components in HRVs of all subjects and apnea groups: extremely low frequencies (ELFs) in the range of 0.01–0.4 mHz and narrowed ultra-low frequencies (nULFs) in the range of 0.1–4 mHz. The independence of both components is also shown. Concluding, heart rate variability is separately regulated by circadian rhythms (ELF bound) and ultradian fluctuations (nULF bound), which can be assessed by decomposing HRV, and the obtained components may be helpful to better understand the underlying homeostatic mechanisms, as well as in the long-term monitoring of patients. Full article

Exploring the design of beneficial nonlinear restoring force structures has become a highly popular topic due to their extensive applications in energy harvesting, actuation, energy absorption, robotics, etc. However, the current literature lacks a systematic review and classification that addresses the design, modeling, and parameter identification of nonlinear restoring forces. Thus, the present paper provides a thorough examination of the latest advancements in the design of nonlinear restoring forces, as well as modeling and parameter identification in contemporary beneficial nonlinear designs. The seven design methodologies, namely magnetic coupling, oblique spring linkages, static or dynamic preloading, metamaterials, bio-inspired, MEMS (Micro-Electromechanical Systems) manufacturing, and dry friction applied approaches, are classified. The polynomial, hysteretic, and piecewise linear models are summarized for nonlinear restoring force characterization. The system parameter identification methods covering restoring force surface, Hilbert transform, time-frequency analysis, nonlinear subspace identification, unscented Kalman filter, optimization algorithms, physics-informed neural networks, and data-driven sparse regression are reviewed. Moreover, possible enhancement strategies for nonlinear system identification of nonlinear restoring forces are presented. Finally, broader implications and future directions for the design, characterization, and identification of nonlinear restoring forces are discussed. Full article

This article introduces an iterative algorithm that is created by integrating the $S^{*}$-iteration process with the inertial forward-backward algorithm. The algorithm is designed to solve optimization problems formulated as variational inclusions in a real Hilbert space. We establish the weak convergence of the algorithm under conventional assumptions. One of the applications of the algorithm is to solve the extreme learning machine, which can be transformed into the variational inclusion problem. Different algorithms, with all parameters set to be identical, are employed to solve the stroke classification problem in order to evaluate the algorithm’s performance. The results indicate that our algorithm converges faster than others and achieves a precision of 93.90%, a recall of 100%, and an F1-score of 96.58%. Full article

Partial discharge pattern recognition is a crucial task for assessing the insulation condition of Gas-Insulated Switchgear (GIS). However, the on-site environment presents challenges such as strong electromagnetic interference, leading to acquired signals with a low signal-to-noise ratio (SNR). Furthermore, traditional pattern recognition methods based on statistical parameters suffer from redundant and inefficient features that compromise classification accuracy, while existing artificial-intelligence-based classification methods lack the ability to quantify the uncertainty in defect classification. To address these issues, this paper proposes a novel GIS partial discharge pattern recognition method based on time–frequency energy grayscale maps and an improved variational Bayesian autoencoder. Firstly, a denoising-based approximate message passing algorithm is employed to sample and denoise the discharge signals, which enhances the SNR while simultaneously reducing the number of sampling points. Subsequently, a two-dimensional time–instantaneous frequency energy grayscale map of the discharge signal is constructed based on the Hilbert–Huang Transform and energy grayscale mapping, effectively extracting key time–frequency features. Finally, an improved variational Bayesian autoencoder is utilized for the unsupervised learning of the image features, establishing a GIS defect classification method with an associated confidence level by integrating probabilistic features. Validation based on measured data demonstrates the effectiveness of the proposed method. Full article

(1) Background: Pain assessment still relies primarily on subjective self-report. To address these limitations, we developed Piq, an EEG-based index derived from beta-band brain activity (Piq_β) aimed at providing objective pain identification and quantification. (2) Methods: The study combined cross-sectional and longitudinal designs. Resting-state brain activity was recorded for five minutes, and EEG signals were preprocessed using a dedicated algorithm. Piq_β performance was assessed by identifying an optimal cutoff to discriminate pain from no pain, evaluating its association with VNRS, and estimating agreement using a modified concordance criterion (exact match or ±1 category). A graded scale was also established to classify pain into distinct categories, according to intensity. (3) Results: An optimal cutoff of 10% for Piq_β yielded 97.8% sensitivity and 88.2% specificity. Piq_β correlated with self-reported scores (ρ = 0.60, p < 0.0001) with acceptable agreement (mean bias −1.02), accounting for clinically acceptable discrepancies. Five levels of pain were proposed, and Piq_β demonstrated the ability to track intra-individual fluctuations over time, accounting for clinically acceptable discrepancies. (4) Conclusions: These findings provide strong evidence to support the Piq_β index as a valuable complement to subjective pain ratings. Full article

The nonlinear Schrödinger equation is a classical nonlinear evolution equation with wide applications. This paper explores the asymptotic behavior of solutions to the nonlinear Schrödinger equation with non-zero boundary conditions in the presence of a pair of second-order discrete spectra. We analyze the Riemann–Hilbert problem in the inverse scattering transform by the Deift–Zhou nonlinear steepest descent method. Then we propose a proper deformation to deal with the growing time term and give the conditions for the series in the process of deformation by the Laurent expansion. Finally, we provide the characterization of the interactions between the solitary waves corresponding to second-order discrete spectra and the coherent oscillations produced by the perturbation. Numerical verifications are also performed. Full article

The Wigner representation for quantum mechanics of particles is generalized to Bose fields. The standard Hilbert space quantization becomes, via the Weyl transform, a quantization method that consists of adding a Gaussian zeropoint field distribution to the vacuum. I comment on the possible advantages of the method in order to study quantum fields in curved spaces. I study a unified formulation of non-relativistic quantum electrodynamics in the Weyl–Wigner formalism, in terms of (classical-like) c-numbers. Full article

Intelligent fault diagnosis is a set of methods for the health monitoring of mechanical bearings. To address the problem of insufficient applicability of diagnostic models due to differences in the domain distribution between laboratory data and actual working conditions, this study constructs a complete transfer learning diagnostic system. Firstly, the Hilbert transform technique was introduced to extract time-domain and frequency-domain features, as well as periodic correlations and other indicators; then, three models, i.e., transfer learning (TL), gradient boosting machine (GBM), and random forest (RF), were used to classify the data and compare their accuracy. It was found that TL had the highest accuracy in testing, with an F1 score of 0.9631. In the transfer task of the target domain samples, compared with the direct application of the source domain model with a classification accuracy of 70.3%, the transfer learning method achieved a classification accuracy of 97.6%, and the transfer gain increased by 27.3 percentage points, proving the superiority of the model constructed in this paper. Finally, SHapley Additive exPlanations (SHAP) was used to provide a detailed explanation of the transfer learning model, and the basis for model decision making was revealed through feature importance analysis. Full article

Snoring, a common sleep-disordered breathing phenomenon, impairs sleep quality for both the sufferer and any bed partner. While mild snoring primarily disrupts sleep continuity, severe cases often indicate obstructive sleep apnea (OSA), a disorder affecting 9–17% of the global population, linked to significant comorbidities and socioeconomic burden (see Introduction for supporting data). Here, we propose a low-cost, real-time snoring detection and intervention system that integrates a multiple-spectrum deep learning framework with an Internet of Things (IoT)-enabled smart pillow. The modified Parallel Convolutional Spatiotemporal Network (PCSN) combines three parallel convolutional neural network (CNN) branches processing Constant-Q Transform (CQT), Synchrosqueezing Wavelet Transform (SWT), and Hilbert–Huang Transform (HHT) features with a Long Short-Term Memory (LSTM) network to capture spatial and temporal characteristics of sounds associated with snoring. The smart pillow prototype incorporates two Micro-Electro-Mechanical System (MEMS) microphones, an ESP8266 off-shelf board, a speaker, and two vibration motors for real-time audio acquisition, cloud-based processing via Arduino cloud, and closed-loop haptic/audio feedback that encourages positional changes without fully awakening the snorers. Experiments demonstrated that the modified PCSN model achieves 98.33% accuracy, 99.29% sensitivity, 98.34% specificity, 98.3% recall, and 98.32% F1-score, outperforming existing systems. Hardware costs are under USD 8 and a smartphone app provides authorized users with real-time visualization and secure data access. This solution offers a cost-effective and accurate approach for home-based OSA screening and intervention. Full article

This paper comprehensively reviews advanced signal processing methods for partial discharge (PD) analysis, covering traditional time-frequency techniques, wavelet transform, Hilbert–Huang transform, and artificial intelligence-based methods. This paper critically examines the principles, advantages, limitations, and applicable scenarios of each method. A key contribution of this review is the systematic comparison of these methods, highlighting their evolution and complementary roles in processing non-stationary and noisy PD signals. However, a significant gap in current research remains the lack of standardized, explainable, and embeddable AI solutions for real-time, fine-grained PD classification. Future trends point to hybrid approaches and edge AI systems that combine physical insights with lightweight deep learning models to improve diagnostic accuracy and deployability. Full article

Near-fault ground motions exhibit significant characteristics such as velocity pulses, rupture directivity, and strong vertical components, which pose serious threats to structural safety. However, near-fault ground motion records remain scarce and have not been adequately accounted for in current seismic design codes. This paper proposes a data-driven simulation method for non-stationary near-fault ground motions based on Stationary Wavelet Transform (SWT) combined with Hilbert’s instantaneous frequency estimation. First, to address the baseline drift issue commonly observed in measured seismic motions, a baseline correction technique combining the least squares method and the Iwan method is proposed to enhance the reliability of seismic time histories. Subsequently, statistical distributions of velocity pulses and vertical-to-horizontal (V/H) acceleration ratios, along with their relationships with fault distance and magnitude, are analyzed based on more than 900 ground motion records. The results show that these near-fault motions generally contain pronounced long-period components, which will have significant implications for the seismic response of long-period structures. Additionally, unidirectional pulses dominate in near-fault records. Among the 107 selected long-period pulse records, unidirectional pulses account for 69.2%. Based on this, seismic motions are decomposed using SWT, and stochastic reconstruction is performed, combined with multivariate response spectrum matching to optimize the generation of near-fault time histories consistent with the target spectrum. Compared with the results obtained without optimization, the proposed method reduces the mean square error by about 40% or more, demonstrating a clear improvement in accuracy and reliability. This method provides reliable seismic input support for seismic analysis and performance-based design of bridges in near-fault regions. Full article

The tidal mudflat of the South Yellow Sea is characterized by complex sediment environments that preserve rich paleoenvironmental signals, making it an important area for understanding land–sea interactions and promoting sustainable coastal development. Thus, accurate identification of sediment sequences and layer thicknesses becomes crucial for interpreting sediment dynamics and paleoenvironmental reconstruction. While borehole data have elucidated local sediment facies, their spatially discontinuous nature hinders a holistic reconstruction of regional depositional history. To overcome this limitation, ground-penetrating radar (GPR) surveys were conducted across the tidal mudflat of the South Yellow Sea, enabling systematic correlation between radar reflection patterns and sediment architectures. Based on the relationship between the dielectric permittivity and wave velocity, short-time Fourier transform (STFT) was applied to derive the peak-weighted average frequency in the frequency domain for individual soil layers, revealing its dependence on dielectric properties. Sediment interfaces and layer thicknesses were determined using three methods: the radar image waveform method, the Hilbert spectrum instantaneous phase method, and the generalized S-transform time–frequency analysis method. The results indicate the following: (1) GPR enables high-fidelity imaging of subsurface stratigraphy, successfully resolving three distinct radar facies: F1: high-amplitude, horizontal, continuous reflections with parallel waveforms; F2: moderate-to-high-amplitude, sinuous continuous reflections with parallelism; and F3: medium-amplitude, discontinuous chaotic reflections. (2) All three methods effectively characterize subsurface soil stratification, but positioning accuracy decreases systematically with depth. Excluding anomalous errors at one site, the relative error for most layers within the 1 m depth is below 15%, and remains ≤25% at the 1–2 m depth. Beyond the 2 m depth, reliable stratification becomes unattainable due to severe signal attenuation. (3) Comparative analysis demonstrates that the Hilbert spectral instantaneous phase method significantly enhances GPR signals, achieving an optimal performance with positioning errors consistently below 5 cm for most soil layers. The application of this approach along the tidal mudflat of the South Yellow Sea significantly enhances the precision of sediment layer boundary identification. Our analysis systematically interpreted radar facies, demonstrating the effectiveness of the Hilbert spectrum instantaneous phase method in delineating soil stratification. These findings offer reliable technical support for interpreting GPR data in comparable sediment environments. Full article

This paper proposes a novel AC side series arc fault (SAF) identification method based on Generalized Lissajous Trajectory (GLT) learning for low-voltage residential circuits. The method addresses challenges in detecting SAFs—characterized by high concealment, random occurrence, and limitations in existing protection devices—by leveraging the Hilbert transform to map current signals into 2D Generalized Lissajous Trajectories. These trajectories amplify key SAF features (e.g., zero-break distortion and random pulses). A ResNet50-based image recognition model achieves high-precision fault detection under specific load types, with a validation accuracy of up to 99.91% for linear loads and 98.93% for nonlinear loads. The algorithm operates within 1.6 ms, enabling real-time circuit breaker tripping. The proposed method achieves higher recognition accuracy with lower computational cost compared to other image-based methods. In this paper, an adjustable load signal modeling approach is proposed to visualize the current signal using GLT and complete the lightweight identification based on ResNet network, which provides new ideas and methods for series arc fault detection. Full article

Corrosion of rock bolts in engineering exhibits random spatial distribution characteristics. To elucidate the influence mechanism of stochastic corrosion on the surface of rock bolts on the propagation behavior of ultrasonic guided waves, this study establishes a finite element model of rock bolts that incorporates stochastic corrosion characteristics. The coupled effects of corrosion depth and area ratio on guided wave propagation characteristics, time-domain response, energy distribution, and wave velocity variation are systematically investigated. Results indicate that corrosion depth and area ratio synergistically deteriorate guided wave morphology, transforming the stress field from symmetric and uniform to asymmetric and spiral. Reflections, scattering, and mode conversion induced by defects lead to a significant increase in the attenuation rate of pulse amplitude, with the two parameters governing the vertical interaction intensity and horizontal interference scope, respectively. Analysis of the Hilbert curve reveals that corrosion characteristics disrupt energy concentration. Under constant corrosion depth, an increase in area ratio disperses energy toward delayed scattered waves, while under constant area ratio, greater corrosion depth reduces the peak amplitude of the envelope curve. Overall, the energy integral exhibits an increasing trend with the degree of corrosion, whereas the peak-to-peak wave velocity shows a declining trend. The established multivariate nonlinear model accurately describes the coupled influence of corrosion parameters on wave velocity. This stochastic corrosion model overcomes the limitations of traditional simplified models and provides critical theoretical support for parameter calibration and engineering application of ultrasonic guided wave technology in the quantitative assessment of rock bolt corrosion. Full article