The advent of Knowledge-Based Planning (KBP) models has introduced a transformative approach to Intensity-Modulated Radiation Therapy (IMRT) treatment planning in breast cancer and lung cancer cases. This paper explores the application of KBP models in these specific cancer types, highlighting their potential to enhance treatment accuracy, efficiency, and patient outcomes. By leveraging historical treatment data and machine learning techniques, KBP-IMRT offers a data-driven framework for optimizing dose distributions, minimizing radiation exposure to healthy tissues, and improving overall treatment plan quality. Through a comprehensive review of the literature and clinical case studies, this paper underscores the advantages of KBP-IMRT, such as streamlined planning processes and improved plan consistency, while acknowledging the challenges associated with model development and implementation. As the field of radiotherapy continues to evolve, KBP models hold the promise of shaping the future of personalized and precise cancer treatment strategies.

The scale-up of AI models for analyzing largescale power systems necessitates a thorough understanding of their scaling properties. Existing studies on these properties provide only partial insights, showing that loss function decreases predictably with increased model scales; yet no scaling law for power system AI models has been established, and model performance remains unpredictable due to “emergent abilities”. This study pioneers the discussion on the emergent abilities of graph neural network (GNN) for analyzing large-scale power systems, revealing that model performance improves dramatically once model scale exceeds a threshold. Furthermore, we introduce an empirical power-law formula to quantify the relationship between this threshold and the power system size. Our theory accurately predicts the threshold for the appearance of emergent ability in large-scale power systems, including a synthetic 10,000- bus and a real-world 19,402-bus systems.

Convolutional Neural Network (CNN) models often exhibit sub-optimal performance without labeled datasets. For diseases like brain tumors, the availability of labels is challenging, and obtaining scans for specific grades of tumors is even more difficult. To tackle these issues, this paper introduces an innovative approach that utilizes self-supervised learning (SSL) to learn insights from unlabeled data and few-shot learning to handle the small sample size problem. Current methods targeting small datasets yield poor, often over-segmented results, rendering them ineffective for clinical applications. Particularly for brain tumor MR images, we propose to learn more information via registration as a pretext task. We register deformed images using an encoder-decoder model, allowing us to learn a more adequate representation of the segmentation target. The trained model extracts prototypes from the support images, employing them as guidance to generate segmentation masks for the query images. The use of registration as a pretext task in the few-shot model sees a dice score of 75.18%, a 1.28% increase over the baselines for the BRATS2021 in the 1-way-1-shot setting. Furthermore, our method demonstrates a remarkable ability to delineate tumor boundaries. The Hausdorff Distance (HD) of the model drops to 10.03 from 14.29, showing the model's ability to identify tumor boundaries. This novel approach holds significant potential for addressing the scarcity of labeled data in MR imaging, particularly in scenarios where obtaining precise annotations is challenging. With this approach, we aim to equip clinicians with valuable tools for more precise and effective diagnosis and treatment of brain tumors.

This paper presents a robust bag encoding strategy based on an MLP mixer. The proposed approach introduces the mixing concept to MIL applications, which helps to generate robust bag encoding. The existing bag encoding strategies for MIL applications consider instances in the bag as independent. This assumption may restrict the performance of these algorithms. Therefore, this paper proposes MIL-Mixer, which utilizes the information between the instances to generate a robust bag encoding. We also extend MLP-Mixture to use classification token similar to vision transformers which diversify the encoding generation process. In this study, three benchmark MIL datasets are used to assess the performance of the proposed MIL-Mixer. In comparison with existing MIL approaches, the proposed MIL-Mixer achieves better performance.

With the rapid integration of electronically interfaced renewable energy resources and loads into smart grids, there is increasing interest in power quality disturbances (PQD) classification to enhance the security and efficiency of these grids. This paper introduces a new approach to PQD classification based on the Vision Transformer (ViT) model. When a PQD occurs, the proposed approach first converts the power quality signal into an image and then utilizes a pre-trained ViT to accurately determine the class of the PQD. Unlike most previous works, which were limited to a few disturbance classes or small datasets, the proposed method is trained and tested on a large dataset with 17 disturbance classes. Our experimental results show that the proposed ViT-based approach achieves PQD classification precision and recall of 98.28% and 97.98%, respectively, outperforming recently proposed techniques applied to the same dataset.

AC-coupled amplifiers are widely used in analog signal processing, particularly in audio and biopotential amplifiers. Applications that require very low cutoff frequencies 𝒇𝑪 lead to large time constants 𝝉 = 𝑹𝑪, thus resulting in large 𝑹 values because the capacitance of high-quality non-polarized capacitors is limited to a few 𝝁𝑭. Moreover, in a classic approach using a passive 𝑹𝑪 network before the amplifier, the noise of this high value resistor forces to set 𝒇𝑪 well below the bandwidth of interest to reduce its noise contribution, thus leading to very long transient time responses. The presented analysis shows that the use of DC servo loop circuits (DCSL) allows becoming negligible the noise contribution of the resistor, that works as a “cold resistor”: it behaves as a resistor of value 𝑹 to set 𝒇𝑪 but with a noise significantly lower than its theoretical thermal noise. The inverting and non-inverting DCSL topologies were analyzed showing that the cost to pay for this noise reduction is a limitation on the DC voltages the DCSLs admit at their inputs. The non-inverting scheme allows for achieving the lowest noise but the lowest DC input range. The circuits' analytic noise transfer functions and their experimental validation are provided. As an example, a DCSL optimized to achieve a very-low voltage noise, a gain of 40 dB and 𝒇𝑪 = 𝟏. 𝟔 𝐇𝐳 was designed, built, and tested showing that its input referred noise 𝒆𝒊 is exclusively due to the input operational amplifier. Using an OPA221 of Texas InstrumentsTM the DCSL featured 𝒆𝒊=𝟏. 𝟐 𝐧𝐕⁄√𝐇𝐳 @𝟏 𝐤𝐇𝐳.

Kinematics, kinetics and biomechanics of human gait are widely investigated fields of research. The biomechanics of locomotion has been described characterizing muscle activations and synergistic control, i.e, spatial and temporal patterns of coordinated muscle groups and joints. Both kinematic synergies and muscle synergies have been extracted from locomotion data, showing that in healthy people 4-5 synergies underlie human locomotion; such synergies are in general robust across subjects and might be altered in pathological gait, depending on the severity of the impairment. In this work, for the first time, we apply the mixed matrix factorization algorithm to locomotion data of 15 healthy participants to extract hybrid kinematic-muscle synergies, and show that they allow to directly link task space variables (i.e., kinematics) to the neural structure of muscle synergies. We show that kinematic-muscle synergies can describe the biomechanics of motion at a better extent than muscle synergies or kinematic synergies alone. Morevoer, this study shows that at a functional level, modular control of the lower limb during locomotion is underlied by an increased number of functional synergies with respect to standard muscle synergies and account for different biomechanical roles that each synergy may have within the movement. Kinematic-muscular synergies may have impact in future work for a deeper understanding of modular control and neuro-motor recovery in the medical and rehabilitation fields, as they associate neural and task space variables in the same factorization, including the evaluation of post-stroke, Parkinson and cerebral palsy patients, and in other fields such as sports.

Dynamic metasurface antennas (DMAs) is a new antenna technology comprising multiple microstrips, each encompassing several reconfigurable metamaterial elements acting as radiating and receiving antennas. The whole unit is usually connected via the input/output port of each microstrip to the radio frequency (RF) chains of the baseband unit. Point-to-point (P2P) multiple-input multiple-output (MIMO) system is one of the most prominent wireless communication systems, with its use cases recently increasing with the booming of machine-to-machine communication. However, it is yet to be explored with DMAs. This work thus studies the achievable rate maximization of a DMA-based P2P MIMO system whose transmitter or receiver is equipped with DMA instead of traditional antennas. Noteworthy, this is the first work to analyze the deployment where both the BS and the user are equipped with DMA. Additionally, we numerically demonstrate how uplink and downlink communication can be achieved in each DMA deployment strategy. Finally, through the simulation results, the performance of our proposed designs is compared with various benchmark systems, such as conventional P2P and IRS-aided MIMO systems, where in both cases, it is shown that the proposed designs provide the best tradeoff in achieving high performance with significant RF chains reduction.

This work proposes a vehicle number plate identification approach using convolutional neural networks and YOLOv8 for automated entry and exit tracking. Vehicle identification plays a crucial role in detecting and tracking vehicles on roads, parking areas, commercial spaces such as shopping malls, and also residential areas such as apartments. There are many approaches reported in the recent past on detecting vehicles and identifying the number plate but many of them are not appropriate to use in real practical scenarios. In this connection, this work proposes a new method for identifying the number plates using advanced deep-learning techniques. This project aims to create a comprehensive solution capable of automating the entry and exit logging process will streamline operations, reduce wait times at entry and exit points, and minimize the need for manual data entry and verification. When implemented in a real-world setup, the proposed method is found to be robust and outperforms some of the already-known approaches with a license plate detection precision of 98.45% and 96.06% recall which is comparable with many state-of-the-art approaches.

We analyse a lossy transmission line and the Johnson-Nyquist noise generated therein. A representation as a noisy two-port with a voltage and a current noise sources on one end of a noiseless two-port is given. An expression for the noise properties is given for an arbitrary temperature profile along the transmission line. Compatibility of the general expression found here with expectations for special cases calculable using thermodynamics is demonstrated.

Due to the high complexity of geometry-deterministic wireless channel modeling and the difficulty in its implementation, geometry-based stochastic channel modeling (GBSM) approaches have been used to evaluate system performance of wireless communications. This paper introduces a new method to model a GBSM by training a generative neural network using images formed by channel parameters. Toward this end, we process the data of channel parameters in the form of images and train the generative neural networks where the convolutional layers are mainly employed. Through a case study, we confirm that the use of channel images facilitates the training of the generative model and ensures that the model learns the correlations among multipath components. We show that the outputs of the generative model faithfully represent the distributions of the original data. Furthermore, to corroborate applicability of the trained model, we run simple system-level simulations and show the results from the trained model and ray-tracing data are well matched. Therefore, we argue that the proposed model trained by channel images will ease the burden of GBSM implementations with general wireless conditions and capture the statistical joint distributions of the original channel data.

Spam, defined as unwanted, malicious and improper content that led to various challenge in digital realm. This paper provides a comprehensive overview and analysis of the different type of spam such as social media spam, SMS spam, call spam and forum/blog spam. Each of these spams has its own characteristics and challenges. This paper analyzes the methods used by spammers such as obfuscation, social engineering, polymorphism and machine learning integration to make the spam message pass through the detection system. The analysis of the consequences of spam resulting financial losses, security breaches and social consequences. In address these challenges, the study proposes a comprehensive method in handling spam such as advanced filtering system, robust email authentication and enhanced the user awareness education about spam.

The feasibility of memristor devices as a controllable resistor element for making tunable analogue filters is investigated by employing the Stanford Resistive Random Access Memory (RRAM) memristor model. Replacing the resistor in a conventional filter with a memristor allows the filter to have a variable range, this would solve the limitation where the conventional analogue filters have only a single filter range due to the fixed values of their constituent elements. This paper reports the results of fitting the Al: HfO2 memristor using the Stanford RRAM model, and the fitted results can be perfectly matched to real experimental results. Memristor-based filters can obtain at least two different filter ranges with the same filtering effect as conventional filters.

The integration of unmanned aerial vehicles (UAVs), cognitive radio (CR), and non-orthogonal multiple access (NOMA) presents a promising solution to significantly enhance the performance of future wireless networks. Achieving this integration requires cognitive self-awareness for intelligent resource allocation. In this paper, we address the problem of sum rate maximization in UAV-enabled cognitive NOMA uplink systems through the joint optimization of subchannel assignment and power allocation, while considering the UAV's mobility. The traditional approach to finding the optimal solution requires an iterative or exhaustive search across all possible combinations of subchannel assignment, power allocation, and UAV position at each time slot, leading to excessive computational complexity. Furthermore, machine learning models, often trained on datasets that do not fully capture the complexity of real-world scenarios, struggle to handle non-stationary events effectively. To solve this nonconvex optimization challenge, we draw inspiration from active inference in cognitive neuroscience and propose a novel data-driven approach called the Active Generalized Dynamic Bayesian Network (Active-GDBN). The main idea is to process the unknown nonlinear input of an exhaustive search optimization algorithm using an Active-GDBN framework. This framework leverages a probabilistic generative model to learn the complex relationships and dependencies among subchannel assignments, power distributions, and the UAV's mobility. The model is facilitated by continuous neuronal message passing in both discrete and continuous states to predict the optimal configuration. Numerical results show that the proposed approach achieves sum rate performance near the optimal upper bound and surpasses other baseline approaches.

As digital economies continue to grow and the demand for wireless communication intensify, efficient and effective spectrum management becomes increasingly vital. In response, regulatory bodies worldwide are interest in leveraging advanced technologies to enhance and expedite their spectrum management efforts. This paper investigates the application of large language models (LLMs) for regulatory processes and their potential to improve operational efficiencies and decision-making. In addition, this paper demonstrates how LLMs enhance and accelerate policy and regulation processes for information extraction and question answering; focusing on spectrum management use cases. The results obtained indicate a robust document processing alternative, suggesting a transformative impact on spectrum regulatory practices.

Technology CAD (TCAD) tools are pivotal for transistor modeling, providing high-accuracy, physics-based simulations essential for the development of next-generation technology nodes. Nevertheless, their high computational cost and low throughput significantly impede large-scale design-space exploration. Addressing these limitations by accelerating TCAD processes is a critical research goal. One promising approach involves using graph-based representations of TCAD meshes fed into graph neural networks (GNNs), which offer rapid and accurate alternatives. However, traditional GNN solutions, such as Graph Convolutional Networks (GCNs) and Graph Attention Networks (GATs), often falter with noisy graphs like transistor meshes, where embeddings become cluttered and complex. To overcome these challenges, we introduce the first selfsupervised GNN model for TCAD device modeling, specifically employing the Self-Supervised GAT (SuperGAT). We demonstrate that self-supervised GNN models can, in contrast to GCNs and GATs, declutter noisy graphs by performing self-supervised edge prediction tasks to differentiate between informative and redundant connections. Our evaluation compares various GNNbased aggregation layers and demonstrates that the proposed SuperGAT model successfully reproduces fundamental transistor transfer characteristics (Id-Vg) curves, sub-threshold swing (SS), and threshold voltage (Vth). We generated our dataset using 14nm FDSOI [1] transistors, which were validated against industrial measurements. Our SuperGAT model achieves a minimum R 2 score of 0.992 across all transistor characteristics and an inference time of only 0.007 sec per transistor compared to TCAD simulation time of 13 min per transistor. Resulting in a speedup of more than 100 000 x, highlighting its potential to revolutionize TCAD transistor simulations in advanced technology development. This holds especially true for performing reliability and variability analysis for transistors, where expensive Monte-Carlo simulations are inevitable.

Combining machine learning with physical models can significantly impact retrieval algorithms designed to invert geophysical parameters from remote sensing data. Such hybrid models integrate physical knowledge with domain expertise through a joint architecture, potentially enhancing performance by increasing the efficiency and flexibility of the physical model as well as the generalization and interpretability of the machine learning predictions. This work introduces a hybrid model for estimating forest height using single-baseline, single-polarization TanDEM-X interferometric coherence measurements. In this model, the vertical reflectivity profile is derived as a function of input features, including topographic and acquisition geometry descriptors, using a multilayer perceptron network. This profile is then used to invert forest height by leveraging the established physical relationship connecting the vertical reflectivity profile to forest height. The developed model is applied and validated on several TanDEM-X acquisitions over tropical sites with different acquisition geometries, and its performance is assessed against reference data derived from airborne LiDAR measurements.

This paper delves into the integration of quantum technologies with semantic communications, addressing challenges related to transmitting knowledge graphs over conventional wireless channels. As traditional communications evolve towards the sharing and elaboration of knowledge between intelligent agents, both natural and artificial, they encounter critical challenges in terms of efficiency, accuracy, and the associated high complexity. To address these challenges, we propose leveraging quantum mechanical principles to communicate conceptual graph-based semantic representations of data between networked parties. The primary goal of the protocol is to transmit essential semantic features accurately while minimizing resource consumption and maintaining resilience against depolarizing errors. These include representing the semantic messages efficiently, using a quantum data encoding method, and employing remote state preparation to transmit the messages. Our work evaluates both the success probability of decoding semantic messages and investigates scalability analysis. Our results reveal that quantum semantic communication is robust against a noisy environment, and the evaluation of scalability analysis shows a trade-off regarding the complexity of communicating conceptual graphs.