Search Results (1,096)

Background: The intracranial space has limited capacity; thus, volume changes in any component can raise intracranial pressure and cause mass effect. This mechanism underlies many neurological disorders. Artificial Intelligence, increasingly applied in medicine and diagnostic imaging, may support the evaluation of such conditions. This systematic review investigates AI-based models for cerebrospinal fluid segmentation and analysis on computed tomography. Methods: In December 2024, a systematic review was conducted across MEDLINE (PubMed), Scopus, Web of Science, Embase, and Cochrane Library. From 559 identified studies, 14 were included after independent review by two evaluators. Extracted data covered study characteristics, AI model design, dataset composition, and performance metrics for CSF segmentation. Quality assessment followed PRISMA 2020 and used JBI, AMSTAR 2, and CASP checklists. Results: The 14 studies demonstrated applications of AI in CSF segmentation and volumetric assessment, primarily for hydrocephalus diagnosis, mass effect evaluation, and stroke outcome prediction. Convolutional Neural Networks and Random Forests were the most frequent approaches. Reported segmentation accuracy was high, with Dice Similarity Coefficient values ranging from 0.75 to 0.95 and strong volumetric correlations (r up to 0.99) between AI-based and manual measurements. Conclusions: AI-assisted CSF segmentation from CT images shows promising accuracy and efficiency, with potential to enhance neurological diagnostics. Remaining challenges include dataset variability, inconsistent algorithm performance, and limited clinical validation. Future research should prioritize standardization of methods, larger and more diverse training datasets, and integration of AI tools into clinical workflows. Full article

As carbon cycling and global environmental protection gain increasing attention, forest disturbance research has intensified worldwide. Constrained by limited data availability, existing frameworks often rely on extracting individual spectral bands for simple binary disturbance detection, lacking systematic approaches to visualize and classify causes of disturbance over large areas. Accurately identifying disturbance types is critical because different disturbances (e.g., fires, logging, pests) exhibit vastly different impacts on forest structure, successional pathways and, consequently, forest carbon sequestration and storage capacities. This study proposes an integrated remote sensing and deep learning (DL) method for forest disturbance type identification, enabling high-precision monitoring in Northeast China from 1992 to 2023. Leveraging the Google Earth Engine platform, we integrated Landsat time-series data (30 m resolution), Global Forest Change data, and other multi-source datasets. We extracted four key vegetation indices (NDVI, EVI, NBR, NDMI) to construct long-term forest disturbance feature series. A comparative analysis showed that the proposed convolutional neural network (CNN) model with six feature bands achieved 5.16% higher overall accuracy and a 6.92% higher Kappa coefficient than a random forest (RF) algorithm. Remarkably, even with only six features, the CNN model outperformed the RF model trained on fifteen features, achieving a 0.4% higher overall accuracy and a 0.58% higher Kappa coefficient, while utilizing 60% fewer parameters. The CNN model accurately classified forest disturbances—including fires, pests, logging, and geological disasters—achieving a 92.26% overall accuracy and an 89.04% Kappa coefficient. This surpasses the 81.4% accuracy of the Global Forest Change product. The method significantly improves the spatiotemporal accuracy of regional-scale forest monitoring, offering a robust framework for tracking ecosystem dynamics. Full article

Fine-grained identification of crop planting structures provides key data for precision agriculture, thereby supporting scientific production and evidence-based policy making. This study selected a representative experimental farmland in Qingyang, Gansu Province, and acquired Unmanned Aerial Vehicle (UAV) multi-temporal data (six epochs) from multiple sensors (multispectral [visible–NIR], thermal infrared, and LiDAR). By fusing 59 feature indices, we achieved high-accuracy extraction of cropland and planting structures and identified the key feature combinations that discriminate among crops. The results show that (1) multi-source UAV data from April + June can effectively delineate cropland and enable accurate plot segmentation; (2) July is the optimal time window for fine-scale extraction of all planting-structure types in the area (legumes, millet, maize, buckwheat, wheat, sorghum, maize–legume intercropping, and vegetables), with a cumulative importance of 72.26% for the top ten features, while the April + June combination retains most of the separability (67.36%), enabling earlier but slightly less precise mapping; and (3) under July imagery, the SAM (Segment Anything Model) segmentation + RF (Random Forest) classification approach—using the RF-selected top 10 of the 59 features—achieved an overall accuracy of 92.66% with a Kappa of 0.9163, representing a 7.57% improvement over the contemporaneous SAM + CNN (Convolutional Neural Network) method. This work establishes a basis for UAV-based recognition of typical crops in the Qingyang sector of the Loess Plateau and, by deriving optimal recognition timelines and feature combinations from multi-epoch data, offers useful guidance for satellite-based mapping of planting structures across the Loess Plateau following multi-scale data fusion. Full article

Early warning systems on edge devices such as satellites and unmanned aerial vehicles (UAVs) are essential for effective forest fire prevention. Edge Intelligence (EI) enables deploying deep learning models on edge devices; however, traditional convolutional neural networks (CNNs)/Transformer-based models struggle to balance local-global context integration and computational efficiency in such constrained environments. To address these challenges, this paper proposes HybriDet, a novel hybrid-architecture neural network for wildfire detection. This architecture integrates the strengths of both CNNs and Transformers to effectively capture both local and global contextual information. Furthermore, we introduce efficient attention mechanisms—Windowed Attention and Coordinate-Spatial (CS) Attention—to simultaneously enhance channel-wise and spatial-wise features in high-resolution imagery, enabling long-range dependency modeling and discriminative feature extraction. Additionally, to optimize deployment efficiency, we also apply model pruning techniques to improve generalization performance and inference speed. Extensive experimental evaluations demonstrate that HybriDet achieves superior feature extraction capabilities while maintaining high computational efficiency. The optimized lightweight variant of HybriDet has a compact model size of merely 6.45 M parameters, facilitating seamless deployment on resource-constrained edge devices. Comparative evaluations on the FASDD-UAV, FASDD-RS, and VOC datasets demonstrate that HybriDet achieves superior performance over state-of-the-art models, particularly in processing highly heterogeneous remote sensing (RS) imagery. When benchmarked against YOLOv8, HybriDet demonstrates a 6.4% enhancement in mAP50 on the FASDD-RS dataset while maintaining comparable computational complexity. Meanwhile, on the VOC dataset and the FASDD-UAV dataset, our model improved by 3.6% and 0.2%, respectively, compared to the baseline model YOLOv8. These advancements highlight HybriDet’s theoretical significance as a novel hybrid EI framework for wildfire detection, with practical implications for disaster emergency response, socioeconomic security, and ecological conservation. Full article

Background/Objectives: Non-variceal upper gastrointestinal bleeding (NVUGIB) is associated with considerable morbidity and mortality, particularly in emergency department (ED) settings. While traditional clinical scores such as the Glasgow-Blatchford Score (GBS), AIMS65, and Pre-Endoscopic Rockall are widely used for risk stratification, their accuracy in mortality prediction is limited. This study aimed to evaluate the performance of multiple supervised machine learning (ML) models in predicting 30-day all-cause mortality in NVUGIB and to compare these models with established risk scores. Methods: A retrospective cohort study was conducted on 1233 adult patients with NVUGIB who presented to the ED of a tertiary center between January 2022 and January 2025. Clinical and laboratory data were extracted from electronic records. Seven supervised ML algorithms—logistic regression, ridge regression, support vector machine, random forest, extreme gradient boosting (XGBoost), naïve Bayes, and artificial neural networks—were trained using six feature selection techniques generating 42 distinct models. Performance was assessed using AUROC, F1-score, sensitivity, specificity, and calibration metrics. Traditional scores (GBS, AIMS65, Rockall) were evaluated in parallel. Results: Among the cohort, 96 patients (7.8%) died within 30 days. The best-performing ML model (XGBoost with univariate feature selection) achieved an AUROC > 0.80 and F1-score of 0.909, significantly outperforming all traditional scores (highest AUROC: Rockall, 0.743; p < 0.001). ML models demonstrated higher sensitivity and specificity, with improved calibration. Key predictors consistently included age, comorbidities, hemodynamic parameters, and laboratory markers. The best-performing ML models demonstrated very high apparent AUROC values (up to 0.999 in internal analysis), substantially exceeding conventional scores. These results should be interpreted as apparent performance estimates, likely optimistic in the absence of external validation. Conclusions: While machine-learning models showed markedly higher apparent discrimination than conventional scores, these findings are based on a single-center retrospective dataset and require external multicenter validation before clinical implementation. Full article

In response to the limitations of traditional single-signal approaches, which fail to comprehensively reflect fault conditions, and the difficulties of existing feature extraction methods in capturing subtle fault patterns in transformer fault diagnosis, this paper proposes an innovative fault diagnosis methodology. Initially, to address common severe faults in excitation transformers, Principal Component Analysis (PCA) is applied to reduce the dimensionality of multi-source feature data, effectively eliminating redundant information. Subsequently, to mitigate the impact of non-stationary noise interference in voiceprint signals, a Deep Belief Network (DBN) optimized using the Hunter–Prey Optimization (HPO) algorithm is employed to automatically extract deep features highly correlated with faults, thus enabling the detection of complex, subtle fault patterns. For temperature and electrical parameter signals, which contain abundant time-domain information, the Random Forest algorithm is utilized to evaluate and select the most relevant time-domain statistics. Nonlinear dimensionality reduction is then performed using an autoencoder to further reduce redundant features. Finally, a multi-classifier model based on Adaptive Boosting with Support Vector Machine (Adaboost-SVM) is constructed to fuse multi-source heterogeneous information. By incorporating a pseudo-label self-training strategy and integrating a working condition awareness mechanism, the model effectively analyzes feature distribution differences across varying operational conditions, selecting potential unseen condition samples for training. This approach enhances the model’s adaptability and stability, enabling real-time fault diagnosis. Experimental results demonstrate that the proposed method achieves an overall accuracy of 96.89% in excitation transformer fault diagnosis, outperforming traditional models such as SVM, Extreme Gradient Boosting with Support Vector Machine (XGBoost-SVM), and Convolutional Neural Network (CNN). The method proves to be highly practical and generalizable, significantly improving fault diagnosis accuracy. Full article

This study develops an innovative unmanned aerial vehicle (UAV)-based intelligent system for winter wheat yield prediction, addressing the inefficiencies of traditional manual counting methods (with approximately 15% error rate) and enabling quantitative analysis of water–fertilizer interactions. By integrating an enhanced Faster Region-Based Convolutional Neural Network (Faster R-CNN) architecture with multi-source data fusion and machine learning, the system significantly improves both spike detection accuracy and yield forecasting performance. Field experiments during the 2022–2023 growing season captured high-resolution multispectral imagery for varied irrigation regimes and fertilization treatments. The optimized detection model incorporates ResNet-50 as the backbone feature extraction network, with residual connections and channel attention mechanisms, achieving a mean average precision (mAP) of 91.2% (calculated at IoU threshold 0.5) and 88.72% recall while reducing computational complexity. The model outperformed YOLOv8 by a statistically significant 2.1% margin (p < 0.05). Using model-generated spike counts as input, the random forest (RF) model regressor demonstrated superior yield prediction performance (R² = 0.82, RMSE = 324.42 kg·ha⁻¹), exceeding the Partial Least Squares Regression (PLSR) (R² +46%, RMSE-44.3%), Least Squares Support Vector Machine (LSSVM) (R² + 32.3%, RMSE-32.4%), Support Vector Regression (SVR) (R² + 30.2%, RMSE-29.6%), and Backpropagation (BP) Neural Network (R²+22.4%, RMSE-24.4%) models. Analysis of different water–fertilizer treatments revealed that while organic fertilizer under full irrigation (750 m³ ha⁻¹) conditions achieved maximum yield benefit (13,679.26 CNY·ha⁻¹), it showed relatively low water productivity (WP = 7.43 kg·m⁻³). Conversely, under deficit irrigation (450 m³ ha⁻¹) conditions, the 3:7 organic/inorganic fertilizer treatment achieved optimal WP (11.65 kg m⁻³) and WUE (20.16 kg∙ha⁻¹∙mm⁻¹) while increasing yield benefit by 25.46% compared to organic fertilizer alone. This research establishes an integrated technical framework for high-throughput spike monitoring and yield estimation, providing actionable insights for synergistic water–fertilizer management strategies in sustainable precision agriculture. Full article

High-resolution remote sensing (HRRS) imagery enables the extraction of cropland information with high levels of detail, especially when combined with the impressive performance of deep convolutional neural networks (DCNNs) in understanding these images. Comprehending the factors influencing DCNNs’ performance in HRRS cropland extraction is of considerable importance for practical agricultural monitoring applications. This study investigates the impact of classifier selection and different training data characteristics on the HRRS cropland classification outcomes. Specifically, Gaofen-1 composite images with 2 m spatial resolution are employed for HRRS cropland extraction, and two county-wide regions with distinct agricultural landscapes in Shandong Province, China, are selected as the study areas. The performance of two deep learning (DL) algorithms (UNet and DeepLabv3+) and a traditional classification algorithm, Object-Based Image Analysis with Random Forest (OBIA-RF), is compared. Additionally, the effects of different band combinations, crop growth stages, and class mislabeling on the classification accuracy are evaluated. The results demonstrated that the UNet and DeepLabv3+ models outperformed OBIA-RF in both simple and complex agricultural landscapes, and were insensitive to the changes in band combinations, indicating their ability to learn abstract features and contextual semantic information for HRRS cropland extraction. Moreover, compared with the DL models, OBIA-RF was more sensitive to changes in the temporal characteristics. The performance of all three models was unaffected when the mislabeling error ratio remained below 5%. Beyond this threshold, the performance of all models decreased, with UNet and DeepLabv3+ showing similar performance decline trends and OBIA-RF suffering a more drastic reduction. Furthermore, the DL models exhibited relatively low sensitivity to the patch size of sample blocks and data augmentation. These findings can facilitate the design of operational implementations for practical applications. Full article

Background: Hepatopancreatobiliary (HPB) surgery is among the most complex domains in oncologic care, where decisions entail significant risk–benefit considerations. Artificial intelligence (AI) has emerged as a promising tool for improving individualized decision-making through enhanced risk stratification, complication prediction, and survival modeling. However, its role in HPB oncologic surgery has not been comprehensively assessed. Methods: This systematic review was conducted in accordance with PRISMA guidelines and registered with PROSPERO ID: CRD420251114173. A comprehensive search across six databases was performed through 30 May 2025. Eligible studies evaluated AI applications in risk–benefit assessment in HPB cancer surgery. Inclusion criteria encompassed peer-reviewed, English-language studies involving human s ubjects. Two independent reviewers conducted study selection, data extraction, and quality appraisal. Results: Thirteen studies published between 2020 and 2024 met the inclusion criteria. Most studies employed retrospective designs with sample sizes ranging from small institutional cohorts to large national databases. AI models were developed for cancer risk prediction (n = 9), postoperative complication modeling (n = 4), and survival prediction (n = 3). Common algorithms included Random Forest, XGBoost, Decision Trees, Artificial Neural Networks, and Transformer-based models. While internal performance metrics were generally favorable, external validation was reported in only five studies, and calibration metrics were often lacking. Integration into clinical workflows was described in just two studies. No study addressed cost-effectiveness or patient perspectives. Overall risk of bias was moderate to high, primarily due to retrospective designs and incomplete reporting. Conclusions: AI demonstrates early promise in augmenting risk–benefit assessment for HPB oncologic surgery, particularly in predictive modeling. However, its clinical utility remains limited by methodological weaknesses and a lack of real-world integration. Future research should focus on prospective, multicenter validation, standardized reporting, clinical implementation, cost-effectiveness analysis, and the incorporation of patient-centered outcomes. Full article

Children with Autism Spectrum Disorder (ASD) often experience cognitive difficulties that impact learning. This study explores the use of electroencephalogram data collected with the MUSE 2 headband during task-based cognitive sessions to understand how cognitive states in children with ASD change across three structured tasks: Shape Matching, Shape Sorting, and Number Matching. Following signal preprocessing using Independent Component Analysis (ICA), power across various frequency bands was extracted using the Welch method. These features were used to analyze cognitive states in children with ASD in comparison to typically developing (TD) peers. To capture dynamic changes in attention over time, Morlet wavelet transform was applied, revealing distinct brain signal patterns. Machine learning classifiers were then developed to accurately distinguish between ASD and TD groups using the EEG data. Models included Support Vector Machine, K-Nearest Neighbors, Random Forest, an Ensemble method, and a Neural Network. Among these, the Ensemble method achieved the highest accuracy at 0.90. Feature importance analysis was conducted to identify the most influential EEG features contributing to classification performance. Based on these findings, an ASD map was generated to visually highlight the key EEG regions associated with ASD-related cognitive patterns. These findings highlight the potential of EEG-based models to capture ASD-specific neural and attentional patterns during learning, supporting their application in developing more personalized educational approaches. However, due to the limited sample size and participant heterogeneity, these findings should be considered exploratory. Future studies with larger samples are needed to validate and generalize the results. Full article

This paper presents an automatic system for the classification of musical instruments from audio recordings. The project leverages deep learning (DL) techniques to achieve its objective, exploring three different classification approaches based on distinct input representations. The first method involves the extraction of Mel-Frequency Cepstral Coefficients (MFCCs) from the audio files, which are then fed into a two-dimensional convolutional neural network (Conv2D). The second approach makes use of mel-spectrogram images as input to a similar Conv2D architecture. The third approach employs conventional machine learning (ML) classifiers, including Logistic Regression, K-Nearest Neighbors, and Random Forest, trained on MFCC-derived feature vectors. To gain insight into the behavior of the DL model, explainability techniques were applied to the Conv2D model using mel-spectrograms, allowing for a better understanding of how the network interprets relevant features for classification. Additionally, t-distributed stochastic neighbor embedding (t-SNE) was employed on the MFCC vectors to visualize how instrument classes are organized in the feature space. One of the main challenges encountered was the class imbalance within the dataset, which was addressed by assigning class-specific weights during training. The results, in terms of classification accuracy, were very satisfactory across all approaches, with the convolutional models and Random Forest achieving around 97–98%, and Logistic Regression yielding slightly lower performance. In conclusion, the proposed methods proved effective for the selected dataset, and future work may focus on further improving class balance techniques. Full article

Machine learning (ML) algorithms have promoted the development of predictive modeling of mineral prospectivity, enabling data-driven decision-making processes by integrating multi-source geological information, leading to efficient and accurate prediction of mineral exploration targets. However, it is challenging to conduct ML-based mineral prospectivity mapping (MPM) in under-explored areas where scarce data are available. In this study, the Narigongma district of the Qiangtang block in the Himalayan–Tibetan orogen was chosen as a case study. Five typical alterations related to porphyry mineralization in the study area, namely pyritization, sericitization, silicification, chloritization and propylitization, were extracted by remote sensing interpretation to enrich the data source for MPM. The extracted alteration evidences, combined with geological, geophysical and geochemical multi-source information, were employed to train the ML models. Four machine learning models, including artificial neural network (ANN), random forest (RF), support vector machine and logistic regression, were employed to map the Cu-polymetallic prospectivity in the study area. The predictive performances of the models were evaluated through confusion matrix-based indices and success-rate curves. The results show that the classification accuracy of the four models all exceed 85%, among which the ANN model achieves the highest accuracy of 96.43% and a leading Kappa value of 92.86%. In terms of predictive efficiency, the RF model outperforms the other models, which captures 75% of the mineralization sites within only 3.5% of the predicted area. A total of eight exploration targets were delineated upon a comprehensive assessment of all ML models, and these targets were further ranked based on the verification of high-resolution geochemical anomalies and evaluation of the transportation condition. The interpretability analyses emphasize the key roles of spatial proxies of porphyry intrusions and geochemical exploration in model prediction as well as significant influences everted by pyritization and chloritization, which accords well with the established knowledge about porphyry mineral systems in the study area. The findings of this study provide a robust ML-based framework for the exploration targeting in greenfield areas with good outcrops but low exploration extent, where fusion of a remote sensing technique and multi-source geo-information serve as an effective exploration strategy. Full article

Natural latex harvest is pushing the boundaries of unmanned agricultural production in rubber milk collection via integrated robots in hilly and mountainous regions, such as the fixed and mobile tapping robots widely deployed in forests. As there are bad working conditions and complex natural environments surrounding rubber trees, the real-time and precision assessment of rubber milk yield status has emerged as a key requirement for improving the efficiency and autonomous management of these kinds of large-scale automatic tapping robots. However, traditional manual rubber milk yield status detection methods are limited in their ability to operate effectively under conditions involving complex terrain, dense forest backgrounds, irregular surface geometries of rubber milk, and the frequent occlusion of rubber milk bowls (RMBs) by vegetation. To address this issue, this study presents an unmanned aerial vehicle (UAV) imagery rubber milk yield state detection method, termed YOLOv8n-RMB, in unstructured field environments instead of manual watching. The proposed method improved the original YOLOv8n by integrating structural enhancements across the backbone, neck, and head components of the network. First, a receptive field attention convolution (RFACONV) module is embedded within the backbone to improve the model’s ability to extract target-relevant features in visually complex environments. Second, within the neck structure, a bidirectional feature pyramid network (BiFPN) is applied to strengthen the fusion of features across multiple spatial scales. Third, in the head, a content-aware dynamic upsampling module of DySample is adopted to enhance the reconstruction of spatial details and the preservation of object boundaries. Finally, the detection framework is integrated with the BoT-SORT tracking algorithm to achieve continuous multi-object association and dynamic state monitoring based on the filling status of RMBs. Experimental evaluation shows that the proposed YOLOv8n-RMB model achieves an AP@0.5 of 94.9%, an AP@0.5:0.95 of 89.7%, a precision of 91.3%, and a recall of 91.9%. Moreover, the performance improves by 2.7%, 2.9%, 3.9%, and 9.7%, compared with the original YOLOv8n. Plus, the total number of parameters is kept within 3.0 million, and the computational cost is limited to 8.3 GFLOPs. This model meets the requirements of yield assessment tasks by conducting computations in resource-limited environments for both fixed and mobile tapping robots in rubber plantations. Full article

Rice is one of the most important staple crops in China, and the rapid and accurate extraction of rice planting areas plays a crucial role in the agricultural management and food security assessment. However, the existing rice field identification methods faced the significant challenges in mountainous regions due to the severe cloud contamination, insufficient utilization of multi-dimensional features, and limited classification accuracy. This study presented a novel rice field identification method based on the Graph Convolutional Networks (GCN) that effectively integrated multi-source remote sensing data tailored for the complex mountainous terrain. A coarse-to-fine cloud removal strategy was developed by fusing the synthetic aperture radar (SAR) imagery with temporally adjacent optical remote sensing imagery, achieving high cloud removal accuracy, thereby providing reliable and clear optical data for the subsequent rice mapping. A comprehensive multi-feature library comprising spectral, texture, polarization, and terrain attributes was constructed and optimized via a stepwise selection process. Furthermore, the 19 key features were established to enhance the classification performance. The proposed method achieved an overall accuracy of 98.3% for the rice field identification in Huoshan County of the Dabie Mountains, and a 96.8% consistency compared to statistical yearbook data. The ablation experiments demonstrated that incorporating terrain features substantially improved the rice field identification accuracy under the complex topographic conditions. The comparative evaluations against support vector machine (SVM), random forest (RF), and U-Net models confirmed the superiority of the proposed method in terms of accuracy, local performance, terrain adaptability, training sample requirement, and computational cost, and demonstrated its effectiveness and applicability for the high-precision rice field distribution mapping in mountainous environments. Full article

Automated quality control is critical in modern manufacturing, especially for metallic cast components, where fast and accurate surface defect detection is required. This study evaluates classical Machine Learning (ML) algorithms using extracted statistical parameters and deep learning (DL) architectures including ResNet50, Capsule Networks, and a 3D Convolutional Neural Network (CNN3D) using 3D image inputs. Using the Dataset Original, ML models with the selected parameters achieved high performance: RF reached 99.4 ± 0.2% precision and 99.4 ± 0.2% sensitivity, GB 96.0 ± 0.2% precision and 96.0 ± 0.2% sensitivity. ResNet50 trained with extracted parameters reached 98.0 ± 1.5% accuracy and 98.2 ± 1.7% F1-score. Capsule-based architectures achieved the best results, with ConvCapsuleLayer reaching 98.7 ± 0.2% accuracy and 100.0 ± 0.0% precision for the normal class, and 98.9 ± 0.2% F1-score for the affected class. CNN3D applied on 3D image inputs reached 88.61 ± 1.01% accuracy and 90.14 ± 0.95% F1-score. Using the Dataset Expanded with ML and PCA-selected features, Random Forest achieved 99.4 ± 0.2% precision and 99.4 ± 0.2% sensitivity, K-Nearest Neighbors 99.2 ± 0.0% precision and 99.2 ± 0.0% sensitivity, and SVM 99.2 ± 0.0% precision and 99.2 ± 0.0% sensitivity, demonstrating consistent high performance. All models were evaluated using repeated train-test splits to calculate averages of standard metrics (accuracy, precision, recall, F1-score), and processing times were measured, showing very low per-image execution times (as low as $3.69\times {10}^{-4}$ s/image), supporting potential real-time industrial application. These results indicate that combining statistical descriptors with ML and DL architectures provides a robust and scalable solution for automated, non-destructive surface defect detection, with high accuracy and reliability across both the original and expanded datasets. Full article