Search Results (375)

Transformer reliability is crucial to grid security, with core loosening a common fault. This paper proposes a transformer core loosening fault diagnosis method based on a fusion feature extraction approach and Categorical Boosting (CatBoost) optimized by the Crested Porcupine Optimizer (CPO) algorithm. Firstly, the audio signal is decomposed into six Intrinsic Mode Functions (IMF) components through Variational Mode Decomposition (VMD). This paper utilizes Gaussian membership functions to quantify the energy proportion, central frequency, and kurtosis of IMF and constructs a fuzzy entropy discrimination function. Then, the IMF noise components are removed through an adaptive threshold. Subsequently, the denoised signal undergoes a wavelet packet transform instead of a short-time Fourier transform to optimize Mel-frequency cepstral coefficients (WPT-MFCC), combining time-domain statistical features and frequency-band energy distribution to form a 24-dimensional fusion feature. Finally, the CatBoost algorithm is employed to validate the effects of different feature schemes. The CPO is introduced to optimize its iteration number, learning rate, tree depth, and random strength parameters, thereby enhancing overall performance. The CPO-optimized CatBoost model had 99.0196% fault recognition accuracy in experimental testing, 15% better than the standard CatBoost. Accuracy exceeded 90% even under extreme 0 dB noise. This method makes fault diagnosis more accurate and reliable. Full article

An Adaptive Neuro-Fuzzy Inference System (ANFIS) framework for paediatric wrist injury classification (fracture versus sprain) was developed utilising infrared thermography (IRT). ANFIS combines artificial neural network (ANN) learning with interpretable fuzzy rules, mitigating the “black-box” limitation of conventional ANNs through explicit membership functions and Takagi–Sugeno rule consequents. Forty children (19 fractures, 21 sprains, confirmed by X-ray radiograph) provided thermal image sequences from which three statistically discriminative temperature distribution features namely standard deviation, inter-quartile range (IQR) and kurtosis were selected. A five-layer Sugeno ANFIS with Gaussian membership functions were trained using a hybrid least-squares/gradient descent optimisation and evaluated under three premise-parameter initialisation strategies: random seeding, K-means clustering, and fuzzy C-means (FCM) data partitioning. Five-fold cross-validation guided the selection of membership functions standard deviation (σ) and rule count, yielding an optimal nine-rule model. Comparative experiments show K-means initialisation achieved the best balance between convergence speed and generalisation versus slower but highly precise random initialisation and rapidly convergent yet unstable FCM. The proposed K-means–driven ANFIS offered data-efficient decision support, highlighting the potential of thermal feature fusion with neuro-fuzzy modelling to reduce unnecessary radiographs in emergency bone fracture triage. Full article

Bearings are core components in many types of industrial equipment, and their operating environment is often accompanied by strong background noise. This results in a low Signal-to-Noise Ratio (SNR) in the collected vibration signals, making it difficult for traditional methods to extract fault information effectively. Given that bearing failures often manifest as periodic impact signals, a Feature Mode Decomposition (FMD) method has been proposed by researchers which optimizes filter design through correlated kurtosis to enhance the ability to capture fault impact components. However, the decomposition performance of FMD is significantly affected by its parameters (mode number and filter length), and relies on manual settings, resulting in insufficient stability of the results. Therefore, this paper proposes a Dual-domain Impulse Complexity Index (DICI) that combines time-domain impulse characteristics and frequency-domain complexity as an evaluation criterion for FMD parameter optimization. Further, the projection-iterative-methods-based optimizer (PIMO) is adopted to achieve adaptive optimization of parameters. Subsequently, sensitive components are selected based on the maximum Fault Frequency Correlation (FFC) criterion, and their envelope spectra are calculated to recognize bearing fault modes. Simulation and real-signal verification show that the proposed method outperforms several established signal-processing approaches under low SNR conditions. Full article

Background/Objectives: Germ Cell Neoplasia In Situ (GCNIS) is considered the precursor lesion for the majority of testicular germ cell tumors (TGCTs). The aim of this study was to evaluate whether first-order radiomics features derived from volumetric diffusion tensor imaging (DTI) metrics—specifically apparent diffusion coefficient (ADC) and fractional anisotropy (FA) histogram parameters—can detect GCNIS. Methods: This study included 15 men with TGCTs and 10 controls. All participants underwent scrotal MRI, including DTI. Volumetric ADC and FA histogram metrics were calculated for the following tissues: group 1, TGCT; group 2: testicular parenchyma adjacent to tumor, histologically positive for GCNIS; and group 3, normal testis. Non-parametric statistics were used to assess differences in ADC and FA histogram parameters among the three groups. Pearson’s correlation analysis was followed by ordinal regression analysis to identify key predictive histogram parameters. Results: Widespread distributional differences (p < 0.05) were observed for many ADC and FA variables, with both TGCTs and GCNIS showing significant divergence from normal testes. Among the ADC statistics, the 10th percentile and skewness (p = 0.042), range (p = 0.023), interquartile range (p = 0.021), total energy (p = 0.033), entropy and kurtosis (p = 0.027) proved the most significant predictors for tissue classification. FA_energy (p = 0.039) was the most significant fingerprint of the carcinogenesis among the FA metrics. These parameters correctly characterized 88.8% of TGCTs, 87.5% of GCNIS tissues and 100% of normal testes. Conclusion: Radiomics features derived from volumetric ADC and FA histograms have promising potential to differentiate TGCTs, GCNIS, and normal testicular tissue, aiding early detection and characterization of pre-cancerous lesions. Full article

Clinical application of AI/DL-aided acquisitions for quantitative bi-parametric (q-bp)MRI requires validation and harmonization across vendor platforms. An AI/DL-accelerated q-bpMRI, including 5-echo T₂ and 4-b-value apparent diffusion coefficient (ADC) mapping, was implemented on two 3T clinical scanners by two vendors alongside the qualitative standard-of-care (SOC) MRI protocols for six patients with biopsy-confirmed prostate cancer (PCa). AI/DL versus SOC bpMRI image quality was compared for MR-visible PCa lesions on a 4-point Likert-like scale. Quantitative validation and protocol bias assessment were performed using a multiparametric phantom with reference T₂ and diffusion kurtosis values mimicking prostate tissue ranges. Six-minute q-bpMRI achieved acceptable diagnostic quality comparable to the SOC. Better SNR was observed for DL/AI versus SOC ADC with method-dependent distortion susceptibility and resolution enhancement. The measured biases were unaffected by AI/DL reconstruction and related to acquisition protocol parameters: constant for spin-echo T₂ (−7 ms to +5 ms) and ADC (4b-fit: −0.37 µm²/ms and 2b-fit: −0.19 µm²/ms), while nonlinear for echo-planar T₂ (−37 ms to +14 ms). Measured phantom ADC bias dependence on b-value range was consistent with that observed for PCa lesions. Bias correction harmonized lesion T₂ and ADC values across different AI/DL-aided q-bpMRI acquisitions. The developed workflow enables harmonization of AI/DL-accelerated quantitative T₂ and ADC mapping in multi-vendor clinical settings. Full article

This study addresses the detection of gunshot residue (GSR) around a bullet hole, which is one of the key forensic procedures for estimating the firing distance. GSR was inspected using flash-pulse thermography (FPT) with Kurtosis statistical processing. The result of such an inspection is a pattern composed of numerous small indications distributed around the hole, attributed to gunshot residue particles. The number and spatial distribution of these indications depend on the firing distance. Analyzing such results based on individual indications is impractical, as the pattern must be evaluated as a whole. Therefore, quantifying the overall result can significantly improve the analysis of the firing distance estimation. This study presents a quantification procedure based on threshold-based mass-marking of indications and evaluation of several statistical characteristics. The correlation between these characteristics and firing distance is then analyzed. A strong but distinctly nonlinear correlation was found between the firing distance and some simple quantitative characteristics, such as the total number of indications. However, the study shows that some derived characteristics, such as the contrast between marked areas and background, exhibit a near-linear correlation. These parameters are, therefore, promising for firing distance analysis based on FPT inspection of GSR on through-shot targets. Full article

The exposure parameters in stereolithography with liquid crystal display (SLA-LCD) influence the functional properties of photopolymers, which is particularly important for tribological applications. In this study, the influence of the exposure time of the layers (2–8 s) on the surface topography (ISO 25178), Brinell hardness (HB) and chemical structure (FTIR spectroscopy) of UV-cured resin samples is investigated. Both insufficient and excessive UV irradiation led to undesirable effects ranging from incomplete cross-linking and surface irregularities to excessive curing, micro-cracking and increased surface kurtosis (high Sku values). The most balanced mechanical and topographical performance was observed at a layer exposure time of 6 s, characterised by low Spk values, uniform surface texture and high cohesion between layers. FTIR analysis confirmed the progressive cross-linking with increasing exposure time. The results show that precise control of irradiation parameters enables optimisation of the interrelationships between microstructure, mechanical properties and surface functionality, which is critical for improving the durability and performance of components operating under boundary or mixed lubrication. Full article

Methane concentration anomalies during coal mining operations are identified as important factors triggering major safety accidents. This study aimed to address the key issues of insufficient adaptability of existing detection methods in dynamic and complex underground environments and limited characterization capabilities for non-uniform sampling data. Specifically, an intelligent diagnostic model was proposed by integrating the improved Dung Beetle Optimization Algorithm (SGDBO) with Transformer-SVM. A dual-path feature fusion architecture was innovatively constructed. First, the original sequence length of samples was unified by interpolation algorithms to adapt to deep learning model inputs. Meanwhile, statistical features of samples (such as kurtosis and differential standard deviation) were extracted to deeply characterize local mutation characteristics. Then, the Transformer network was utilized to automatically capture the temporal dependencies of concentration time series. Additionally, the output features were concatenated with manual statistical features and input into the LSSVM classifier to form a complementary enhancement diagnostic mechanism. Sine chaotic mapping initialization and a golden sine search mechanism were integrated into DBO. Subsequently, the SGDBO algorithm was employed to optimize the hyperparameters of the Transformer-LSSVM hybrid model, breaking through the bottleneck of traditional parameter optimization falling into local optima. Experiments reveal that this model can significantly improve the classification accuracy and robustness of anomaly curve discrimination. Furthermore, core technical support can be provided to construct coal mine safety monitoring systems, demonstrating critical practical value for ensuring national energy security production. Full article

Background: Thyroid nodules are highly prevalent, affecting up to 75% of the population, yet most are benign. The limited specificity of ultrasound-based workup leads to substantial overdiagnosis and overtreatment, underscoring the need for improved imaging-based classification. Diffusion-weighted MRI (DWI), quantified via the apparent diffusion coefficient (ADC), has emerged as a promising imaging biomarker. This meta-analysis updates pooled diagnostic performance metrics and systematically evaluates which DWI acquisition techniques, imaging parameters, and combinations with other MRI modalities are most promising for clinical translation. Methods: PubMed, Web of Science, Scopus, and ProQuest were systematically searched. Pooled sensitivity, specificity, and area under the curve (AUC) were calculated using bivariate random-effects models. The effects of b-value, magnetic field strength, echo time, and diffusion model on diagnostic accuracy and ADC values were examined through subgroup and meta-regression analyses. Results: Forty-six studies (3003 nodules) were included. Pooled sensitivity and specificity were 0.84 (95% CI: 0.81–0.86) and 0.88 (95% CI: 0.85–0.90), with an AUC of 0.912. Intravoxel incoherent motion and diffusion kurtosis imaging showed no added value over the mono-exponential model. For the mono-exponential model, a negative association between b-values and reported ADCs was observed, whereas no association was found between b-values and diagnostic accuracy. Magnetic field strength and echo time did not affect ADCs. Combining DWI with morphological imaging showed the potential to further enhance diagnostic performance. Conclusions: DWI holds strong potential to improve the diagnostic workup of thyroid nodules. Technical standardization, particularly of key acquisition parameters, should be pursued to enable clinical implementation. Full article

Rolling bearings are prone to failure, meaning that research on intelligent fault diagnosis is crucial in relation to this key transmission component in rotating machinery. The application of deep learning (DL) has significantly advanced the development of intelligent fault diagnosis. This paper proposes a novel method for rolling bearing fault diagnosis based on the fractional constant Q non-stationary Gabor transform (FCO-NSGT) and VMamba-Conv. Firstly, a rolling bearing fault experimental platform is established and the vibration signals of rolling bearings under various working conditions are collected using an acceleration sensor. Secondly, a kurtosis-to-entropy ratio (KER) method and the rotational kernel function of the fractional Fourier transform (FRFT) are proposed and applied to the original CO-NSGT to overcome the limitations of the original CO-NSGT, such as the unsatisfactory time–frequency representation due to manual parameter setting and the energy dispersion problem of frequency-modulated signals that vary with time. A lightweight fault diagnosis model, VMamba-Conv, is proposed, which is a restructured version of VMamba. It integrates an efficient selective scanning mechanism, a state space model, and a convolutional network based on SimAX into a dual-branch architecture and uses inverted residual blocks to achieve a lightweight design while maintaining strong feature extraction capabilities. Finally, the time–frequency graph is inputted into VMamba-Conv to diagnose rolling bearing faults. This approach reduces the number of parameters, as well as the computational complexity, while ensuring high accuracy and excellent noise resistance. The results show that the proposed method has excellent fault diagnosis capabilities, with an average accuracy of 99.81%. By comparing the Adjusted Rand Index, Normalized Mutual Information, F1 Score, and accuracy, it is concluded that the proposed method outperforms other comparison methods, demonstrating its effectiveness and superiority. Full article

The abrasion resistance of fabrics is one of the basic properties determining the utility performance and durability. The abrasion resistance of textile materials is measured using the Martindale device according to appropriate standards. The sample breakage method is the most commonly used of the three methods. The method is based on organoleptic assessment of fabric breakage. The method is time-consuming, and results may be subject to error resulting from the subjective nature of the assessment. The aim of the presented work was to check the possibility of the application of contactless 3D surface geometry measurement using an optical profilometer in an assessment of changes in fabrics’ surface due to the abrasion process. The obtained results confirmed that some parameters of the geometric structure of fabric surfaces, such as the highest height of the roughness profile Rz, the height of the highest pick of the roughness profile Rp, the depth of the lowest valley of the roughness profile Rv, the depth of the total height of the roughness profile Rt, and the kurtosis Rku, can be used to assess the abrasion resistance of fabrics. It is also stated that using the non-contact optical measurement of fabric surface geometry allows for an assessment of the directionality of surface texture. For this purpose, the autocorrelation function and angle distribution function can be applied. Full article

To address the challenges posed by weak early fault signal features, strong noise interference, low diagnostic accuracy, poor reliability when using single information sources, and the limited availability of high-quality samples in practical applications for permanent magnet synchronous motor (PMSM) bearings, this paper proposes an early bearing fault diagnosis method based on Hippopotamus Optimization Variational Mode Decomposition (HO-VMD) and weighted evidence fusion of current–vibration signals. The HO algorithm is employed to optimize the parameters of VMD for adaptive modal decomposition of current and vibration signals, resulting in the generation of intrinsic mode functions (IMFs). These IMFs are then selected and reconstructed based on their kurtosis to suppress noise and harmonic interference. Subsequently, the reconstructed signals are demodulated using the Teager–Kaiser Energy Operator (TKEO), and both time-domain and energy spectrum features are extracted. The reliability of these features is utilized to adaptively weight the basic probability assignment (BPA) functions. Finally, a weighted modified Dempster–Shafer evidence theory (WMDST) is applied to fuse multi-source feature information, enabling an accurate assessment of the PMSM bearing health status. The experimental results demonstrate that the proposed method significantly enhances the signal-to-noise ratio (SNR) and enables precise diagnosis of early bearing faults even in scenarios with limited sample sizes. Full article

This study introduces an algorithm for quantifying oceanic wind speed and identifying sound sources in the local underwater soundscape. Utilizing low-complexity metrics like one-minute spectral kurtosis and power spectral density levels, the algorithm categorizes different soundscapes and estimates wind speed. It detects rain, vessels, fin and blue whales, as well as clicks and whistles from dolphins. Positioned as a foundational tool for implementing the Ocean Sound Essential Ocean Variable (EOV), it contributes to understanding long-term trends in climate change for sustainable ocean health and predicting threats through forecasts. The proposed soundscape classification algorithm, validated using extensive acoustic recordings (≥32 kHz) collected at various depths and latitudes, demonstrates high performance, achieving an average precision of $89\%$ and an average recall of $86.59\%$ through optimized parameter tuning via a genetic algorithm. Here, wind speed is determined using a cubic function with power spectral density (PSD) at 6 kHz and the MASLUW method, exhibiting strong agreement with satellite data below 15 m/s. Designed for compatibility with low-power electronics, the algorithm can be applied to both archival datasets and real-time data streams. It provides a straightforward metric for ocean monitoring and sound source identification. Full article

To address the challenges of weak fault features and strong non-stationarity in early-stage vibration signals, this study proposes a novel fault diagnosis method combining enhanced variational mode decomposition (VMD) with a structurally improved GoogLeNet. Specifically, an improved wild horse optimizer (IWHO) with tent chaotic mapping is employed to automatically optimize critical VMD parameters, including the number of modes K and the penalty factor α, enabling precise decomposition of non-stationary signals to extract weak fault features. The vibration signal is decomposed, and the top five intrinsic mode functions (IMFs) are selected based on the kurtosis criterion. Time–frequency features are then extracted from these IMFs and input into a modified GoogLeNet classifier. The GoogLeNet structure is improved by replacing standard n × n convolution kernels with cascaded 1 × n and n × 1 kernels, and by substituting the ReLU activation function with a parameterized TReLU function to enhance adaptability and convergence. Experimental results on two public rolling bearing datasets demonstrate that the proposed method effectively handles non-stationary signals, achieving 99.17% accuracy across four fault types and maintaining over 95.80% accuracy under noisy conditions. Full article

Background and Objective: This study aimed to examine the possibility of using surface electromyography (sEMG) to aid in assessing the correctness of overhead deep squat performance. Electromyography signals were recorded for 20 athletes from the lower (rectus femoris (RF), vastus medialis (VM), biceps femoris (BF), and gluteus (GM)) and upper (deltoid (D), latissimus dorsi (L)) muscles. The sEMG signals were categorized into three groups based on physiotherapists’ evaluations of deep squat correctness. Methods: The raw sEMG signals were filtering at 10–250 Hz, and then the mean frequency, median frequency, and kurtosis were calculated. Next, the maximum excitation of the muscles expressed in percentage of maximum voluntary contraction (%MVC) and co-activation index (CAI) were estimated. To determine the muscle excitation level, the pulse interference filter and variance analysis of the sEMG signal derivative were applied. Next, analysis of variance (ANOVA) tests, that is, nonparametric Kruskal–Wallis and post hoc tests, were performed. Results: The parameter that most clearly differentiated the groups considered turned out to be %MVC. The statistically significant difference with a large effect size in the excitation of RF & GM (p = 0.0011) and VM & GM (p = 0.0002) in group 3, where the correctness of deep squat execution was the highest and ranged from 85% to 92%, was pointed out. With the decrease in the correctness of deep squat performance, an additional statistically significant difference appeared in the excitation of RF & BF and VM & BF for both groups 2 and 1, which was not present in group 3. However, in group 2, with the correctness of the deep squat execution at 62–77%, the statistically significant differences in muscle excitation found in group 3 were preserved, in contrast to group 1, with the lowest 23–54% correctness of the deep squat execution, where the statistical significance of these differences was not confirmed. Conclusions: The results indicate that sEMG can differentiate muscle activity and provide additional information for physiotherapists when assessing the correctness of deep squat performance. The proposed analysis can be used to evaluate the correctness of physical exercises when physiotherapist access is limited. Full article