Search Results (591)

Deep brain stimulation (DBS) is an established neurosurgical treatment for Parkinson’s disease (PD), mainly targeting motor symptoms resistant to pharmacological therapy. This review examines strategies to optimize DBS using advanced anatomical, functional, and imaging approaches. The subthalamic nucleus (STN) remains the principal target for alleviating bradykinesia and rigidity, while recent evidence highlights the dentato-rubro-thalamic tract (DRTt) as an additional promising target, especially for tremor control. Clinical data demonstrate that co-stimulation of both STN and DRTt via electrode electric fields results in superior motor outcomes, including greater reductions in UPDRS-III scores and lower levodopa requirements. The review highlights the use of high-resolution MRI and diffusion tensor imaging tractography in visualizing STN and DRTt with high precision. These methods support accurate targeting and individualized treatment planning. Electric field modelling is discussed as a tool to quantify stimulation overlap with target structures and predict clinical efficacy. Anatomical variability in DRTt positioning relative to the STN is emphasized, supporting the need for patient-specific DBS approaches. Alternative and emerging DBS targets—including the pedunculopontine nucleus, zona incerta, globus pallidus internus, and nucleus basalis of Meynert—are discussed for their potential in treating axial and cognitive symptoms. The review concludes with a forward-looking discussion on network-based DBS paradigms, the integration of adaptive stimulation technologies, and the potential of multimodal imaging and electrophysiological biomarkers to guide therapy. Together, these advances support a paradigm shift from focal to network-based neuromodulation in PD management. Full article

Drone-based target detection using visible and thermal (RGB-T) images is critical in disaster rescue, intelligent transportation, and wildlife monitoring. However, persons typically occupy fewer pixels and exhibit more varied postures than vehicles or large animals, making them difficult to detect in unmanned aerial vehicle (UAV) remote sensing images with complex backgrounds. We propose a novel progressive target-aware network (PTANet) for person detection using RGB-T images. A global adaptive feature fusion module (GAFFM) is designed to fuse the texture and thermal features of persons. A progressive focusing strategy is used. Specifically, we incorporate a person segmentation auxiliary branch (PSAB) during training to enhance target discrimination, while a cross-modality background mask (CMBM) is applied in the inference phase to suppress irrelevant background regions. Extensive experiments demonstrate that the proposed PTANet achieves high accuracy and generalization performance, reaching 79.5%, 47.8%, and 97.3% mean average precision (mAP)@50 on three drone-based person detection benchmarks (VTUAV-det, RGBTDronePerson, and VTSaR), with only 4.72 M parameters. PTANet deployed on an embedded edge device with TensorRT acceleration and quantization achieves an inference speed of 11.177 ms (640 × 640 pixels), indicating its promising potential for real-time onboard person detection. The source code is publicly available on GitHub. Full article

Current medical informatics systems face critical challenges, including limited scalability across distributed institutions, insufficient real-time AI-driven decision support, and lack of standardized interoperability for heterogeneous medical data exchange. To address these challenges, this paper proposes a novel distributed web system architecture for medical informatics, integrating artificial intelligence techniques and cloud-based services. The system ensures interoperability via HL7 FHIR standards and preserves data privacy and fault tolerance across interconnected medical institutions. A hybrid AI pipeline combining principal component analysis (PCA), K-Means clustering, and convolutional neural networks (CNNs) is applied to diffusion tensor imaging (DTI) data for early detection of neurological anomalies. The architecture leverages containerized microservices orchestrated with Docker Swarm, enabling adaptive resource management and high availability. Experimental validation confirms reduced latency, improved system reliability, and enhanced compliance with medical data exchange protocols. Results demonstrate superior performance with an average latency of 94 ms, a diagnostic accuracy of 91.3%, and enhanced clinical workflow efficiency compared to traditional monolithic architectures. The proposed solution successfully addresses scalability limitations while maintaining data security and regulatory compliance across multi-institutional deployments. This work contributes to the advancement of intelligent, interoperable, and scalable e-health infrastructures aligned with the evolution of digital healthcare ecosystems. Full article

Using systems with machine learning technologies for process automation is a global trend in agriculture. However, implementing this technology comes with challenges, such as the need for a large amount of computing resources under conditions of limited energy consumption and the high cost of hardware for intelligent systems. This article presents the possibility of applying a modern ESP32 microcontroller platform in the agro-industrial sector to create intelligent devices based on the Internet of Things. CNN models are implemented based on the TensorFlow architecture in hardware and software solutions based on the ESP32 microcontroller from Espressif company to classify objects in crop fields. The purpose of this work is to create a hardware–software complex for local energy-efficient classification of images with support for IoT protocols. The results of this research allow for the automatic classification of field surfaces with the presence of “high attention” and optimal growth zones. This article shows that classification accuracy exceeding 87% can be achieved in small, energy-efficient systems, even for low-resolution images, depending on the CNN architecture and its quantization algorithm. The application of such technologies and methods of their optimization for energy-efficient devices, such as ESP32, will allow us to create an Intelligent Internet of Things network. Full article

Quantum-enhanced machine learning, encompassing both quantum algorithms and quantum-inspired classical methods such as tensor networks, offers promising tools for extracting structure from complex, high-dimensional data. In this work, we study the training dynamics of Matrix Product State (MPS) classifiers applied to three-class problems, using both fashion MNIST and hyperspectral satellite imagery as representative datasets. We investigate the phenomenon of grokking, where generalization emerges suddenly after memorization, by tracking entanglement entropy, local magnetization, and model performance across training sweeps. Additionally, we employ information-theory tools to gain deeper insights: transfer entropy is used to reveal causal dependencies between label-specific quantum masks, while O-information captures the shift from synergistic to redundant correlations among class outputs. Our results show that grokking in the fashion MNIST task coincides with a sharp entanglement transition and a peak in redundant information, whereas the overfitted hyperspectral model retains synergistic, disordered behavior. These findings highlight the relevance of high-order information dynamics in quantum-inspired learning and emphasize the distinct learning behaviors that emerge in multi-class classification, offering a principled framework to interpret generalization in quantum machine learning architectures. Full article

As 5G and beyond networks grow in heterogeneity, complexity, and scale, traditional Intrusion Detection Systems (IDS) struggle to maintain accurate and precise detection mechanisms. A promising alternative approach to this problem has involved the use of Deep Learning (DL) techniques; however, DL-based IDS suffer from issues relating to interpretation, performance variability, and high computational overheads. These issues limit their practical deployment in real-world applications. In this study, CiNeT is introduced as a novel DL-based IDS employing Convolutional Neural Networks (CNN) within a bijective encoding–decoding framework between network traffic features (such as IPv6, IPv4, Timestamp, MAC addresses, and network data) and their RGB representations. This transformation facilitates our DL IDS in detecting spatial patterns without sacrificing fidelity. The bijective pipeline enables complete traceability from detection decisions to their corresponding network traffic features, enabling a significant initiative towards solving the ‘black-box’ problem inherent in Deep Learning models, thus facilitating digital forensics. Finally, the DL IDS has been evaluated on three datasets, UNSW NB-15, InSDN, and ToN_IoT, with analysis conducted on accuracy, GPU usage, memory utilisation, training, testing, and validation time. To summarise, this study presents a new CNN-based IDS with an end-to-end pipeline between network traffic data and their RGB representation, which offers high performance and enhanced interpretability through revisable transformation. Full article

To address the limitations of traditional single-point detection methods in monitoring the sulfur content of ship fuel (FSC), which are inadequate in meeting the regulatory demands of high-traffic ports, this study proposes an integrated analytical approach based on artificial intelligence. This approach synthesizes multi-source heterogeneous data, including historical fuel testing records, Automatic Identification System (AIS) trajectory data, ship and operator profiles, technical specifications, fuel supply chain documentation, fundamental ship attributes and so on. Following rigorous data cleaning and preprocessing procedures, a refined dataset comprising 3046 records collected between 2017 and 2024 from the Port of Ningbo was utilized. Initially, multiple linear regression analysis was con-ducted to identify key factors influencing sulfur emissions, resulting in an R² value of 0.67. Based on these findings, a deep neural network model was developed using TensorFlow to enable real-time estimation of FSC and classification of compliance risk levels. The results indicate that the proposed method exhibits high estimated accuracy and robustness. An AI-based intelligent monitoring module, developed based on this research, has been integrated into the ship exhaust gas detection system at the Port of Ningbo. This module enables real-time analysis of inbound ships and intelligent identification of potentially non-compliant ships, thereby significantly improving the precision and efficiency of port regulatory operations. This study not only contributes to the theoretical framework for ship fuel compliance monitoring but also provides a practical and scalable technical solution for intelligent port governance. Full article

Recent publications have introduced the concept of quantum-like (QL) bits, along with their associated QL states and QL gate operations, which emerge from the dynamics of complex, synchronized networks. The present work extends these ideas to multi-level QL resources, referred to as QL dits, as higher-dimensional analogs of QL bits. We employ systems of k-regular graphs to construct QL-dits for arbitrary dimensions, where the emergent eigenspectrum of their adjacency matrices defines the QL-state space. The tensor product structure of multi-QL dit systems is realized through the Cartesian product of graphs. Furthermore, we examine the potential computational advantages of employing d-nary QL systems over two-level QL bit systems, particularly in terms of classical resource efficiency. Overall, this study generalizes the paradigm of using synchronized network dynamics for QL information processing to include higher-dimensional QL resources. Full article

Grassland classification via remote sensing is essential for ecosystem monitoring and precision management, yet conventional satellite-based approaches are fundamentally constrained by coarse spatial resolution. To overcome this limitation, we harness high-resolution UAV multi-sensor data, integrating multi-scale image fusion with deep learning to achieve fine-scale grassland classification that satellites cannot provide. First, four categories of UAV data, including RGB, multispectral, thermal infrared, and LiDAR point cloud, were collected, and a fused image tensor consisting of 10 channels (NDVI, VCI, CHM, etc.) was constructed through orthorectification and resampling. For feature-level fusion, four deep fusion networks were designed. Among them, the MultiScale Pyramid Fusion Network, utilizing a pyramid pooling module, effectively integrated spectral and structural features, achieving optimal performance in all six image fusion evaluation metrics, including information entropy (6.84), spatial frequency (15.56), and mean gradient (12.54). Subsequently, training and validation datasets were constructed by integrating visual interpretation samples. Four backbone networks, including UNet++, DeepLabV3+, PSPNet, and FPN, were employed, and attention modules (SE, ECA, and CBAM) were introduced separately to form 12 model combinations. Results indicated that the UNet++ network combined with the SE attention module achieved the best segmentation performance on the validation set, with a mean Intersection over Union (mIoU) of 77.68%, overall accuracy (OA) of 86.98%, F1-score of 81.48%, and Kappa coefficient of 0.82. In the categories of Leymus chinensis and Puccinellia distans, producer’s accuracy (PA)/user’s accuracy (UA) reached 86.46%/82.30% and 82.40%/77.68%, respectively. Whole-image prediction validated the model’s coherent identification capability for patch boundaries. In conclusion, this study provides a systematic approach for integrating multi-source UAV remote sensing data and intelligent grassland interpretation, offering technical support for grassland ecological monitoring and resource assessment. Full article

Purpose: Focal cortical dysplasia (FCD) is a common developmental malformation disease of the cerebral cortex. Although mounting evidence has suggested that FCD lesions have variable locations and topographies throughout the cortex, few studies have explored consistencies in structural connectivity among different lesion types. In this study, we analyzed microscopic structural changes via lesion analysis and explored structural changes in nonlesion regions across the brain. Methods: Diffusion tensor imaging (DTI) and magnetization transfer imaging were used to compare FCD lesions and contralateral normal appearing gray/white matter (cNAG/WM). Voxel-based morphometry was calculated for 28 children with FCD and 34 sex- and age-matched healthy participants. DTI indices of the FCD and healthy control groups were analyzed via the tract-based spatial statistic method to evaluate the microstructure abnormalities of WM fiber tracts in individuals with FCD. Results: In terms of FCD lesions, compared with those of the cNAG, the fractional anisotropy (FA) values were decreased, and the mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD) values were increased; the magnetization transfer ratios were also decreased. In terms of whole-brain changes due to FCD, compared with the healthy control group, the FCD group showed a decrease in the volume of the right hippocampus and left anterior cingulate cortex. FCD patients had lower FA values, higher MD values, lower AD values, and mainly increased RD values in relation to WM microstructure. Conclusions: Microstructural abnormalities outside lesion regions may be related to injury to the epileptic network, and the identification of such abnormalities may complement diagnoses of FCD in pediatric patients. Full article

With the rising adoption of machine learning (ML) and deep learning (DL) applications, the demand for deploying these algorithms closer to sensors has grown significantly, particularly in sensor-driven use cases such as predictive maintenance (PM) and condition monitoring (CM). This study investigated a novel application-oriented approach to representing interpretable ML inference as deep neural networks (DNNs) regarding the latency and energy efficiency on the edge, to tackle the problem of inefficient, high-effort, and uninterpretable-implementation ML algorithms. For this purpose, the interpretable deep neural network representation (IDNNRep) was integrated into an open-source interpretable ML toolbox to demonstrate the inference time and energy efficiency improvements. The goal of this work was to enable the utilization of generic artificial intelligence (AI) accelerators for interpretable ML algorithms to achieve efficient inference on edge hardware in smart sensor applications. This novel approach was applied to one regression and one classification task from the field of PM and validated by implementing the inference on the neural processing unit (NPU) of the QXSP-ML81 Single-Board Computer and the tensor processing unit (TPU) of the Google Coral. Different quantization levels of the implementation were tested against common Python and C++ implementations. The novel implementation reduced the inference time by up to 80% and the mean energy consumption by up to 76% at the lowest precision with only a 0.4% loss of accuracy compared to the C++ implementation. With the successful utilization of generic AI accelerators, the performance was further improved with a 94% reduction for both the inference time and the mean energy consumption. Full article

Hyperspectral anomaly detection (HAD) aims to separate subtle anomalies of a given hyperspectral image (HSI) from its background, which is a hot topic as well as a challenging inverse problem. Despite the significant success of the deep learning-based HAD methods, they are hard to interpret due to their black-box nature. Meanwhile, deep learning methods suffer from the identity mapping (IM) problem, referring to the network excessively focusing on the precise reconstruction of the background while neglecting the appropriate representation of anomalies. To this end, this paper proposes a self-attention-aided deep unfolding tensor representation network (SATRNet) for interpretable HAD by solving the tensor representation (TR)-based optimization model within the framework of deep networks. In particular, a Self-Attention Learning Module (SALM) was first designed to extract discriminative features of the input HSI. The HAD problem was then formulated as a tensor representation problem by exploring both the low-rankness of the background and the sparsity of the anomaly. A Weight Learning Module (WLM) exploring local details was also generated for precise background reconstruction. Finally, a deep network was built to solve the TR-based problem through unfolding and parameterizing the iterative optimization algorithm. The proposed SATRNet prevents the network from learning meaningless mappings, making the network interpretable to some extent while essentially solving the IM problem. The effectiveness of the proposed SATRNet is validated on 11 benchmark HSI datasets. Notably, the performance of SATRNet against adversarial attacks is also investigated in the experimentation, which is the first work exploring adversarial robustness in HAD to the best of our knowledge. Full article

Convolutional Neural Networks (CNNs) are a highly used machine learning architecture in various fields. Typical descriptions of CNNs are based on low-dimension and tensor representations in the feature extraction part. In this article, we extend the setting of CNNs to any arbitrary dimension and linearize the whole setting via the typical layers of neurons. In essence, a partial and a full network construct the entire process of a standard CNN, with the partial network being used to linearize the feature extraction. By doing so, we link the tensor-style representation of CNNs with the pure network representation. The outcomes serve two main purposes: to relate CNNs with other machine learning frameworks and to facilitate intuitive representations. Full article

Background: The epidermal growth factor receptor (EGFR) is an important factor in the behavior of diffuse glioma, serving as a potential biomarker for tumor aggressiveness and a therapeutic target. Diffusion tensor imaging (DTI) provides insights into the microstructural integrity of brain tissues, allowing for detailed visualization of tumor-induced changes in white matter tracts. This imaging technique can complement molecular pathology by correlating imaging findings with molecular markers and genetic profiles, potentially enhancing the understanding of tumor behavior and aiding in the formulation of targeted therapeutic strategies. The present study aimed to investigate the molecular properties of diffuse glioma based on DTI sequences. Methods: A total of 27 patients with diffuse glioma (in accordance with the WHO 2021 classification) were investigated using preoperative DTI sequences. The study was conducted using the tractography software DSI Studio (Hou versions 2025.04.16). Following the preprocessing of the raw data, volumes of the arcuate fasciculus (AF), frontal aslant tract (FAT), inferior fronto-occipital fasciculus (IFOF), superior longitudinal fasciculus (SLF), and uncinate fasciculus (UF) were reconstructed, and fractional anisotropy (FA) was derived. Molecular pathological examination was conducted to assess the presence of EGFR amplifications. Results: The mean age of patients was 56 ± 13 years, with 33% females. EGFR amplification was observed in 8/27 (29.6%) of cases. Following correction for multiple comparisons, FA in the left AF (p = 0.025) and in the left FAT (p = 0.020) was found to be significantly lowered in EGFR amplified glioma. In the right language network, however, no statistically significant changes were observed. Conclusions: EGFR amplification may be associated with lower white matter integrity of left hemispheric language tracts, possibly impairing neurological function and impacting surgical outcomes. The underlying molecular and cellular mechanisms driving this association require further investigation. Full article

This work presents the development of a web system using deep learning (DL) neural networks to solve the inverse kinematics problem of the Quetzal robotic arm, designed for academic and research purposes. Two architectures, LSTM and CNN, were designed, trained, and evaluated using data generated through the Denavit–Hartenberg (D-H) model, considering the robot’s workspace. The evaluation employed the mean squared error (MSE) as the loss metric and mean absolute error (MAE) and accuracy as performance metrics. The CNN model, featuring four convolutional layers and an input of 4 timesteps, achieved the best overall performance (95.9% accuracy, MSE of 0.003, and MAE of 0.040), significantly outperforming the LSTM model in training time. A hybrid web application was implemented, allowing offline training and real-time online inference under one second via an interactive interface developed with Streamlit 1.16. The solution integrates tools such as TensorFlow™ 2.15, Python 3.10, and Anaconda Distribution 2023.03-1, ensuring portability to fog or cloud computing environments. The proposed system stands out for its fast response times (1 s), low computational cost, and high scalability to collaborative robotics environments. It is a viable alternative for applications in educational or research settings, particularly in projects focused on industrial automation. Full article