Search Results (2,132)

Epilepsy is a chronic disease and one of the most common neurological disorders worldwide. Electroencephalogram (EEG) signals are widely used to detect epileptic seizures, which provide specialists with essential information about the brain’s functioning. However, manual screening of EEG signals is laborious, time-consuming, and subjective. The rapid detection of epilepsy seizures is important to reduce the risk of seizure-related implications. The existing automatic machine learning techniques based on deep learning techniques are characterized by automatic extraction and selection of the features, leading to better performance and increasing the robustness of the systems. These methods do not consider the multiscale nature of EEG signals, eventually resulting in poor sensitivity. In addition, the complexity of deep models is relatively high, leading to overfitting issues. To overcome these problems, we proposed an efficient and lightweight multiscale convolutional neural network model (LMPSeizNet), which performs multiscale temporal and spatial analysis of an EEG trial to learn discriminative features relevant to epileptic seizure detection. To evaluate the proposed method, we employed 10-fold cross-validation and three evaluation metrics: accuracy, sensitivity, and specificity. The method achieved an accuracy of 97.42%, a sensitivity of 99.33%, and a specificity of 96.51% for inter-ictal and ictal classes outperforming the state-of-the-art methods. The analysis of the features and the decision-making of the method shows that it learns the features that clearly discriminate the two classes. It will serve as a useful tool for helping neurologists and epilepsy patients. Full article

To overcome the problems in existing infrared remote sensing image generation methods, which make it difficult to combine high fidelity and high efficiency, we propose a High-Fidelity Infrared Remote Sensing Image Generation Method Coupled with the Global Radiation Scattering Mechanism and Pix2PixGAN (HFIRSIGM_GRSMP) in this paper. Firstly, based on the global radiation scattering mechanism, the HFIRSIGM_GRSMP model is constructed to address the problem of accurately characterizing factors that affect fidelity—such as the random distribution of the radiation field, multipath scattering, and nonlinear changes—through the innovative fusion of physical models and deep learning. This model accurately characterizes the complex radiation field distribution and the image detail-feature mapping relationship from visible-to-infrared remote sensing. Then, 8000 pairs of image datasets were constructed based on Landsat 8 and Sentinel-2 satellite data. Finally, the experiment demonstrates that the average SSIM of images generated using HFIRSIGM_GRSMP reaches 89.16%, and all evaluation metrics show significant improvement compared to the contrast models. More importantly, this method demonstrates high accuracy and strong adaptability in generating short-wave, mid-wave, and long-wave infrared remote sensing images. This method provides a more comprehensive solution for generating high-fidelity infrared remote sensing images. Full article

This study facilitates the extraction of foveal avascular zone (FAZ) metrics from optical coherence tomography angiography (OCTA) images, offering valuable clinical insights and enabling detailed statistical analysis of FAZ size and shape across three patient groups: healthy, type II diabetes mellitus and both type II diabetes mellitus (DM) and high blood pressure (HBP). Additionally, it evaluates the performance of four deep learning (DL) models—U-Net, U-Net with DenseNet121, U-Net with MobileNetV2 and U-Net with VGG16—in automating the segmentation of the FAZ. Manual segmentation of the images by ophthalmological clinicians was performed initially, and data augmentation was used to enhance the dataset for robust model training and evaluation. Consequently, the original set of 103 full retina OCTA images was extended to 672 cases, including 42 images from normal patients, 357 images from DM patients, and 273 images from patients with both DM and HBP. Among the models, U-Net with DenseNet outperformed the others, achieving the highest accuracy, Intersection over Union (IoU), and Dice coefficient across all patient groups. This research is distinct in its focus on full retina OCTA images and its inclusion of patients with both hypertension and diabetes, an area that is less studied in the existing literature. Full article

To address the challenge of rapid geometric model development in the digital twin industry, this paper presents a comprehensive pipeline for constructing 3D models from images using monocular vision imaging principles. Firstly, a structure-from-motion (SFM) algorithm generates a 3D point cloud from photographs. The feature detection methods scale-invariant feature transform (SIFT), speeded-up robust features (SURF), and KAZE are compared across six datasets, with SIFT proving the most effective (matching rate higher than 0.12). Using K-nearest-neighbor matching and random sample consensus (RANSAC), refined feature point matching and 3D spatial representation are achieved via antipodal geometry. Then, the Poisson surface reconstruction algorithm converts the point cloud into a mesh model. Additionally, texture images are enhanced by leveraging a visual geometry group (VGG) network-based deep learning approach. Content images from a dataset provide geometric contours via higher-level VGG layers, while textures from style images are extracted using the lower-level layers. These are fused to create texture-transferred images, where the image quality assessment (IQA) metrics SSIM and PSNR are used to evaluate texture-enhanced images. Finally, texture mapping integrates the enhanced textures with the mesh model, improving the scene representation with enhanced texture. The method presented in this paper surpassed a LiDAR-based reconstruction approach by $20\%$ in terms of point cloud density and number of model facets, while the hardware cost was only $1\%$ of that associated with LiDAR. Full article

Numerical simulation in fluid dynamics can be computationally expensive and difficult to achieve. To enhance efficiency, developing high-performance and accurate surrogate models is crucial, where deep learning shows potential. This paper introduces geometric attention (GeoAttention), a method that leverages attention mechanisms to encode geometry represented by point cloud, thereby enhancing the neural network’s generalizability across different geometries. Furthermore, by integrating GeoAttention with neural field, we propose the geometric attention neural field (GeoANF), specifically for learning representations of airfoil flow fields. The GeoANF embeds observational data independently of the specific discretization process into a latent space, constructing a mapping that relates geometric shape to the corresponding flow fields under given initial conditions. We use the public dataset AirfRANS to evaluate our approach, GeoANF significantly surpasses the baseline models on four key performance metrics, particularly in volume flow field and surface pressure measurements, achieving mean squared errors of 0.0038 and 0.0089, respectively. Full article

The main objective of a machine learning (ML) system is to obtain a trained model from input data in such a way that it allows predictions to be made on new i.i.d. (Independently and Identically Distributed) data with the lowest possible error. However, how can we assess whether the training and test data have a similar distribution? To answer this question, this paper presents a proposal to determine the degree of distribution shift of two datasets. To this end, a metric for evaluating complexity in datasets is used, which can be applied in multi-class problems, comparing each pair of classes of the two sets. The proposed methodology has been applied to three well-known datasets: MNIST, CIFAR-10 and CIFAR-100, together with corrupted versions of these. Through this methodology, it is possible to evaluate which types of modification have a greater impact on the generalization of the models without the need to train multiple models multiple times, also allowing us to determine which classes are more affected by corruption. Full article

The challenge of fracture segmentation remains a significant obstacle in imaging logging interpretation within the current oil and gas exploration and development field. However, existing image segmentation algorithms still encounter issues related to accuracy, speed, and robustness, as well as a tendency to misdetect or overlook small fractures when applied to logging image fracture segmentation tasks. To address these challenges comprehensively, this paper proposes an end-to-end fracture segmentation algorithm named SWSDS-Net. This algorithm is built upon the UNet architecture and incorporates the SimAM with slicing (SWS) attention mechanism along with the deformable strip convolution (DSCN) module. The SWS introduces a fully 3D attention mechanism that effectively learns the weights of each neuron in the feature map, enabling better capture of fracture features while ensuring fair attention and enhancement for both large and small objects. Additionally, the deformable properties of DSCN allow for adaptive sampling based on fracture shapes, effectively tackling challenges posed by varying fracture shapes and enhancing segmentation robustness. Experimental results demonstrate that SWSDS-Net achieves optimal performance across all evaluation metrics in this task, delivering superior visual results in fracture segmentation while successfully overcoming limitations present in existing algorithms such as complex shapes, noise interference, and low-quality images. Moreover, serving as a lightweight network solution enables SWSDS-Net’s deployment on mobile devices at remote sites—an advancement that lays a solid foundation for interpreting logging data and promotes deep learning technology application within traditional industrial scenarios. Full article

Schizophrenia is a complex neuropsychiatric disorder characterized by disruptions in brain connectivity and cognitive functioning. Continuous monitoring of neural activity is essential, as it allows for the detection of subtle changes in brain connectivity patterns, which could provide early warnings of cognitive decline or symptom exacerbation, ultimately facilitating timely therapeutic interventions. This paper proposes a novel approach for detecting schizophrenia-related abnormalities using deep learning (DL) techniques applied to electroencephalogram (EEG) data. Using an openly available EEG dataset on schizophrenia, the focus is on preprocessed event-related potentials (ERPs) from key electrode sites and applied transfer entropy (TE) analysis to quantify the directional flow of information between brain regions. TE matrices were generated to capture neural connectivity patterns, which were then used as input for a hybrid DL model, combining convolutional neural networks (CNNs) and Bidirectional Long Short-Term Memory (BiLSTM) networks. The model achieved a performant accuracy of 99.94% in classifying schizophrenia-related abnormalities, demonstrating its potential for real-time mental health monitoring. The generated TE matrices revealed significant differences in connectivity between the two groups, particularly in frontal and central brain regions, which are critical for cognitive processing. These findings were further validated by correlating the results with EEG data obtained from the Muse 2 headband, emphasizing the potential for portable, non-invasive monitoring of schizophrenia in real-world settings. The final model, integrated into the NeuroPredict platform, offers a scalable solution for continuous mental health monitoring. By incorporating EEG data, heart rate, sleep patterns, and environmental metrics, NeuroPredict facilitates early detection and personalized interventions for schizophrenia patients. Full article

Background: Prostate cancer is a leading cause of cancer-related deaths in men worldwide, making accurate diagnosis critical for effective treatment. Recent advancements in artificial intelligence (AI) and machine learning (ML) have shown promise in improving the diagnostic accuracy of prostate cancer. Objectives: This systematic review aims to evaluate the effectiveness of AI-based tools in diagnosing prostate cancer using MRI, with a focus on accuracy, specificity, sensitivity, and clinical utility compared to conventional diagnostic methods. Methods: A comprehensive search was conducted across PubMed, Embase, Ovid MEDLINE, Web of Science, Cochrane Library, and Institute of Electrical and Electronics Engineers (IEEE) Xplore for studies published between 2019 and 2024. Inclusion criteria focused on full-text, English-language studies involving AI for Magnetic Resonance Imaging (MRI) -based prostate cancer diagnosis. Diagnostic performance metrics such as area under curve (AUC), sensitivity, and specificity were analyzed, with risk of bias assessed using the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool. Results: Seven studies met the inclusion criteria, employing various AI techniques, including deep learning and machine learning. These studies reported improved diagnostic accuracy (with AUC scores of up to 97%) and moderate sensitivity, with performance varying based on training data quality and lesion characteristics like Prostate Imaging Reporting and Data System (PI-RADS) scores. Conclusions: AI has significant potential to enhance prostate cancer diagnosis, particularly when used for second opinions in MRI interpretations. While these results are promising, further validation in diverse populations and clinical settings is necessary to fully integrate AI into standard practice. Full article

Integrating advanced technologies such as artificial intelligence (AI) with traditional agricultural practices has changed how activities are developed in agriculture, with the aim of automating manual processes and improving the efficiency and quality of farming decisions. With the advent of deep learning models such as convolutional neural network (CNN) and You Only Look Once (YOLO), many studies have emerged given the need to develop solutions to problems and take advantage of all the potential that this technology has to offer. This systematic literature review aims to present an in-depth investigation of the application of AI in supporting the management of weeds, plant nutrition, water, pests, and diseases. This systematic review was conducted using the PRISMA methodology and guidelines. Data from different papers indicated that the main research interests comprise five groups: (a) type of agronomic problems; (b) type of sensor; (c) dataset treatment; (d) evaluation metrics and quantification; and (e) AI technique. The inclusion (I) and exclusion (E) criteria adopted in this study included: (I1) articles that obtained AI techniques for agricultural analysis; (I2) complete articles written in English; (I3) articles from specialized scientific journals; (E1) articles that did not describe the type of agrarian analysis used; (E2) articles that did not specify the AI technique used and that were incomplete or abstract; (E3) articles that did not present substantial experimental results. The articles were searched on the official pages of the main scientific bases: ACM, IEEE, ScienceDirect, MDPI, and Web of Science. The papers were categorized and grouped to show the main contributions of the literature to support agricultural decisions using AI. This study found that AI methods perform better in supporting weed detection, classification of plant diseases, and estimation of agricultural yield in crops when using images captured by Unmanned Aerial Vehicles (UAVs). Furthermore, CNN and YOLO, as well as their variations, present the best results for all groups presented. This review also points out the limitations and potential challenges when working with deep machine learning models, aiming to contribute to knowledge systematization and to benefit researchers and professionals regarding AI applications in mitigating agronomic problems. Full article

The widespread use of machine and deep learning algorithms for anomaly detection has created a critical need for robust explanations that can identify the features contributing to anomalies. However, effective evaluation methodologies for anomaly explanations are currently lacking, especially those that compare the explanations against the true underlying causes, or ground truth. This paper aims to address this gap by introducing a rigorous, ground-truth-based framework for evaluating anomaly explanation methods, which enables the assessment of explanation correctness and robustness—key factors for actionable insights in anomaly detection. To achieve this, we present an innovative benchmark dataset of digital circuit truth tables with model-based anomalies, accompanied by local ground truth explanations. These explanations were generated using a novel algorithm designed to accurately identify influential features within each anomaly. Additionally, we propose an evaluation methodology based on correctness and robustness metrics, specifically tailored to quantify the reliability of anomaly explanations. This dataset and evaluation framework are publicly available to facilitate further research and standardize evaluation practices. Our experiments demonstrate the utility of this dataset and methodology by evaluating common model-agnostic explanation methods in an anomaly detection context. The results highlight the importance of ground-truth-based evaluation for reliable and interpretable anomaly explanations, advancing both theory and practical applications in explainable AI. This work establishes a foundation for rigorous, evidence-based assessments of anomaly explanations, fostering greater transparency and trust in AI-driven anomaly detection systems. Full article

Urban tree canopy (UTC) coverage, or area, is an important metric for monitoring changes in UTC over large areas within a municipality. Several methods have been used to obtain these data, but remote sensing image classification is one of the fastest and most reliable over large areas. However, most studies have tested only one or two classification methods to accomplish this while using costly satellite imagery or LiDAR data. This study seeks to compare three urban tree canopy cover classifiers by testing a deep learning U-Net convolutional neural network (CNN), support vector machine learning classifier (SVM) and a random forests machine learning classifier (RF) on cost-free 2012 aerial imagery over a small southern USA city and midsize, growing southern USA city. The results of the experiment are then used to decide the best classifier and apply it to more recent aerial imagery to determine canopy changes over a 10-year period. The changes are subsequently compared visually and statistically with recent urban heat maps derived from thermal Landsat 9 satellite data to compare the means of temperatures within areas of UTC loss and no change. The U-Net CNN classifier proved to provide the best overall accuracy for both cities (89.8% and 91.4%), while also requiring the most training and classification time. When compared spatially with city heat maps, city periphery regions were most impacted by substantial changes in UTC area as cities grow and the outer regions get warmer. Furthermore, areas of UTC loss had higher temperatures than those areas with no canopy change. The broader impacts of this study reach the urban forestry managers at the local, state/province, and national levels as they seek to provide data-driven decisions for policy makers. Full article

Unmanned aerial vehicles (UAVs) have emerged as pivotal tools in emergency communication scenarios. In the aftermath of disasters, UAVs can be communication nodes to provide communication services for users in the area. In this paper, we establish a meticulously crafted virtual simulation environment and leverage advanced deep reinforcement learning algorithms to train UAVs agents. Notwithstanding, the development of reinforcement learning algorithms is beset with challenges such as sparse rewards and protracted training durations. To mitigate these issues, we devise an enhanced reward function aimed at bolstering training efficiency. Initially, we delineate a specific mountainous emergency communication scenario and integrate it with the particularized application of UAVs to undertake virtual simulations, constructing a realistic virtual environment. Furthermore, we introduce a supplementary shaped reward function tailored to alleviate the problem of sparse rewards. By refining the DQN algorithm and devising a reward structure grounded on potential functions, we observe marked improvements in the final evaluation metrics, substantiating the efficacy of our approach. The experimental outcomes underscore the prowess of our methodology in effectively curtailing training time while augmenting convergence rates. In summary, our work underscores the potential of leveraging sophisticated virtual environments and refined reinforcement learning techniques to optimize UAVs deployment in emergency communication contexts. Full article

Weighted k-nearest neighbor (WKNN)-based Wi-Fi fingerprinting is popular in indoor location-based services due to its ease of implementation and low computational cost. KNN-based methods rely on distance metrics to select the nearest neighbors. However, traditional metrics often fail to capture the complexity of indoor environments and have limitations in identifying non-linear relationships. To address these issues, we propose a novel WKNN-based Wi-Fi fingerprinting method that incorporates distance metric learning. In the offline phase, our method utilizes a Siamese network with a triplet loss function to learn a meaningful distance metric from training fingerprints (FPs). This process employs a unique triplet mining strategy to handle the inherent noise in FPs. Subsequently, in the online phase, the learned metric is used to calculate the embedding distance, followed by a signal-space distance filtering step to optimally select neighbors and estimate the user’s location. The filtering step mitigates issues from an overfitted distance metric influenced by hard triplets, which could lead to incorrect neighbor selection. We evaluate the proposed method on three benchmark datasets, UJIIndoorLoc, Tampere, and UTSIndoorLoc, and compare it with four WKNN models. The results show a mean positioning error reduction of 3.55% on UJIIndoorLoc, 16.21% on Tampere, and 16.49% on UTSIndoorLoc, demonstrating enhanced positioning accuracy. Full article

When monitoring crop growth using top-down images of the plant canopies, leaves in agricultural fields appear very dense and significantly overlap each other. Moreover, the image can be affected by external conditions such as background environment and light intensity, impacting the effectiveness of image segmentation. To address the challenge of segmenting dense and overlapping plant leaves under natural lighting conditions, this study employed a Bilayer Convolutional Network (BCNet) method for accurate leaf segmentation across various lighting environments. The major contributions of this study are as follows: (1) Utilized Fully Convolutional Object Detection (FCOS) for plant leaf detection, incorporating ResNet-50 with the Convolutional Block Attention Module (CBAM) and Feature Pyramid Network (FPN) to enhance Region of Interest (RoI) feature extraction from canopy top-view images. (2) Extracted the sub-region of the RoI based on the position of the detection box, using this region as input for the BCNet, ensuring precise segmentation. (3) Utilized instance segmentation of canopy top-view images using BCNet, improving segmentation accuracy. (4) Applied the Varifocal Loss Function to improve the classification loss function in FCOS, leading to better performance metrics. The experimental results on cucumber canopy top-view images captured in glass greenhouse and plastic greenhouse environments show that our method is highly effective. For cucumber leaves at different growth stages and under various lighting conditions, the Precision, Recall and Average Precision (AP) metrics for object recognition are 97%, 94% and 96.57%, respectively. For instance segmentation, the Precision, Recall and Average Precision (AP) metrics are 87%, 83% and 84.71%, respectively. Our algorithm outperforms commonly used deep learning algorithms such as Faster R-CNN, Mask R-CNN, YOLOv4 and PANet, showcasing its superior capability in complex agricultural settings. The results of this study demonstrate the potential of our method for accurate recognition and segmentation of highly overlapping leaves in diverse agricultural environments, significantly contributing to the application of deep learning algorithms in smart agriculture. Full article