Background: Deep learning (DL)-based medical image classification is becoming increasingly reliable, enabling physicians to make faster and more accurate decisions in diagnosis and treatment. A plethora of algorithms have been developed to classify and analyze various types of medical images. Among them, Convolutional Neural Networks (CNNs) have proven highly effective, particularly in medical image analysis and disease detection. Methods: To further enhance these capabilities, this research introduces MediVision, a hybrid DL-based model that integrates a vision backbone based on CNNs for feature extraction, capturing detailed patterns and structures essential for precise classification. These features are then processed through Long Short-Term Memory (LSTM), which identifies sequential dependencies to better recognize disease progression. An attention mechanism is then incorporated that selectively focuses on salient features detected by the LSTM, improving the model’s ability to highlight critical abnormalities. Additionally, MediVision utilizes a skip connection, merging attention outputs with LSTM outputs along with Grad-CAM heatmap to visualize the most important regions of the analyzed medical image and further enhance feature representation and classification accuracy. Results: Tested on ten diverse medical image datasets (including, Alzheimer’s disease, breast ultrasound, blood cell, chest X-ray, chest CT scans, diabetic retinopathy, kidney diseases, bone fracture multi-region, retinal OCT, and brain tumor), MediVision consistently achieved classification accuracies above 95%, with a peak of 98%. Conclusions: The proposed MediVision model offers a robust and effective framework for medical image classification, improving interpretability, reliability, and automated disease diagnosis. To support research reproducibility, the codes and datasets used in this study have been publicly made available through an open-access repository.

In this paper, we propose a new method based on the dynamic mode decomposition (DMD) to find a distinctive contrast between the ictal and interictal patterns in epileptic electroencephalography (EEG) data. The features extracted from the method of DMD clearly capture the phase transition of a specific frequency among the channels corresponding to the ictal state and the channel corresponding to the interictal state, such as direct current shift (DC-shift or ictal slow shifts) and high-frequency oscillation (HFO). By performing classification tests with Electrocorticography (ECoG) recordings of one patient measured at different timings, it is shown that the captured phenomenon is the unique pattern that occurs in the ictal onset zone of the patient. We eventually explain how advantageously the DMD captures some specific characteristics to distinguish the ictal state and the interictal state. The method presented in this study allows simultaneous interpretation of changes in the channel correlation and particular information for activity related to an epileptic seizure so that it can be applied to identification and prediction of the ictal state and analysis of the mechanism on its dynamics.

We study Lp−Lr restriction estimates for algebraic varieties in d-dimensional vector spaces over finite fields. Unlike the Euclidean case, if the dimension d is even, then it is conjectured that the L(2d+2)/(d+3)−L2 Stein–Tomas restriction result can be improved to the L(2d+4)/(d+4)−L2 estimate for both spheres and paraboloids in finite fields. In this paper we show that the conjectured Lp−L2 restriction estimate holds in the specific case when test functions under consideration are restricted to d-coordinate functions or homogeneous functions of degree zero. To deduce our result, we use the connection between the restriction phenomena for our varieties in d dimensions and those for homogeneous varieties in (d+1) dimensions

This paper presents a hybrid adaptive approach based on machine learning (ML) for classifying incoming traffic, feature selection and thresholding, aimed at enhancing downgrade attack detection in Wi-Fi Protected Access 3 (WPA3) networks. The fast proliferation of WPA3 is regarded critical for securing modern Wi-Fi systems, which have become integral to 5G and Beyond (5G&B) Radio Access Networks (RAN) architecture. However, the wireless communication channel remains inherently susceptible to downgrade attacks, where adversaries intentionally cause networks to revert from WPA3 to WPA2, with the malicious intent of exploiting known security flaws. Traditional Intrusion Detection Systems (IDS), which rely on fixed-threshold statistical methods, often fail to adapt to changing network environments and new, sophisticated attack strategies. To address this limitation, we introduce a novel ML-based Feature Selection and Thresholding for Downgrade Attacks Detection (MFST-DAD) approach, which comprises three stages: traffic data preprocessing, baseline adaptive feature selection, and real-time detection and prevention using ML algorithms. Experimental results on a specially generated dataset demonstrate that the proposed approach detects downgrade attacks in WPA3 networks, achieving 99.8% accuracy with a Naive Bayes classifier in both WPA3 personal and enterprise transition modes. These findings confirm the effectiveness of our proposed approach in securing next-generation Wi-Fi systems.

Fifth-generation (5G) networks are the fifth generation of mobile networks and are regarded as a global standard, following 1G, 2G, 3G, and 4G networks. Fifth-generation, with its large available bandwidth provided by mmWave, not only provides the end user with higher spectrum efficiency, massive capacity, low latency, and high speed but is also a network designed to connect virtually everyone and everything together, including machines, objects, and devices. Therefore, studies of such systems’ performance evaluation and capacity bounds are critical for the research community. Furthermore, the performance of these systems should be investigated in realistic contexts while considering signal strength and restricted uplink power to maintain system coverage and capacity, which are also affected by the environment and the value of the service factor parameter. However, any proposed application should include a multiservice case to reflect the true state of 5G systems. As an extension of previous work, the capacity bounds for 5G networks are derived and analyzed in this research, considering both single and multiservice cases with mobility. In addition, the influence of different parameters on network performance, such as the interference, service factor, and non-orthogonality factors, and cell radii, is also discussed. The numerical findings and analysis reveal that the type of environment and service factor parameters have the greatest influence on system capacity and coverage. Subsequently, it is shown that the investigated parameters have a major impact on cell performance and therefore can be considered key indicators for mobile designers and operators to consider in planning and designing future networks. To validate these findings, some results are evaluated against ITU-T standards, while others are compared with related studies from the literature.

Background/Objectives: Detecting Alzheimer’s disease (AD) from normal aging using eyes-open (EO) EEG is challenging due to stimulus-driven nonstationarity and fragmented oscillatory responses. This study aims to determine whether prototype-based representations derived from Dynamic Mode Decomposition (DMD) can improve AD detection from EO photostimulation EEG. Methods: We propose a DMD-based framework termed DMD-based Clustered Pattern Projection (DMD-CPP). Segment-wise DMD representations were clustered to learn class-specific medoid prototypes, and each EEG epoch was encoded as a vector of cosine-similarity coordinates with respect to these prototypes. A linear SVM classifier was trained on the resulting DMD-CPP features and evaluated under strict leave-one-subject-out validation. Results: The DMD-CPP model achieved competitive classification accuracy and, importantly, enhanced margin-based reliability. In EO photostimulation, AD versus healthy control classification showed a pronounced improvement, characterized by wider and more asymmetric decision margins, particularly assigning low confidence to normal epochs misclassified as AD. Tasks involving frontotemporal dementia also showed improvement, although the effect was less pronounced than for AD. Conclusions: Clustering-based pattern projection has been shown to stabilize EEG dynamics and provide an interpretable, confidence-aware feature representation. These findings suggest that DMD-CPP offers a promising framework for reliable AD detection from EO EEG, where conventional spectral methods typically struggle.