ABSTRACT The growing scale and heterogeneity of the Internet of Things (IoT) have intensified the need for intrusion detection systems that are not only accurate but also reliable under uncertainty. This study proposes NUTSGRAPH, a graph‐based Bayesian intrusion detection framework that integrates relational traffic modeling with No‐U‐Turn Sampling (NUTS) for uncertainty‐aware IoT security. NetFlow traffic from the NF‐ToN‐IoT‐v3 dataset is transformed into a directed communication graph in which endpoints are represented as nodes and flows are represented as attributed edges. A GraphSAGE encoder learns topology‐aware node embeddings, which are fused with flow attributes to form edge‐level representations for downstream classification. To improve confidence reliability beyond deterministic and approximate Bayesian baselines, a compact Bayesian multilayer perceptron is applied to the frozen graph embeddings, with posterior inference performed using NUTS. Experimental results show that NUTSGRAPH achieved 0.968150 accuracy, 0.003843 expected calibration error, and 0.950017 OoD AUROC, indicating strong predictive performance, improved calibration, and competitive novelty‐sensitive behavior within the NF‐ToN‐IoT‐v3 benchmark setting. The findings suggest that coupling graph‐based relational learning with sampling‐based Bayesian posterior inference can support more dependable intrusion detection by producing predictions with better‐calibrated uncertainty estimates. Although the Bayesian stage introduces higher memory demand, likelihood subset sampling provides a practical mechanism for balancing posterior quality and computational feasibility. Overall, NUTSGRAPH offers a focused benchmark‐based demonstration of how graph‐based relational learning and sampling‐based Bayesian inference can be combined to support more trustworthy, uncertainty‐aware intrusion detection in dynamic IoT environments.

Background:Brain tumour detection and analysis using medical imaging requires the extraction of both local spatial features and global contextual representations. Although convolutional neural networks (CNNs) excel at capturing local spatial patterns and Transformer-based architectures model long-range dependencies effectively, the optimal architectural paradigm for clinical deployment remains unresolved. This systematic review and meta-analysis evaluates hybrid CNN-Transformer architectures for brain tumour detection, focusing on the integration of local and global feature learning, diagnostic accuracy and computational efficiency. The roles of generative adversarial networks (GANs) for addressing data scarcity and multimodal imaging fusion for diagnostic completeness are also critically examined. Methods:A systematic search was conducted across IEEE Xplore, PubMed, Scopus and Google Scholar for studies published between January 2021 and May 2025. From 1876 initially identified articles, 94 met the prespecified inclusion criteria following quality assessment using the QUADAS-2 and ROBINS-I frameworks. A random-effects meta-analysis of diagnostic accuracy was performed using the DerSimonian-Laird estimator, with statistical heterogeneity quantified using I2 and publication bias assessed using funnel plot asymmetry and Egger's test. Computational efficiency was standardised to GigaFLOPs using a reference input of 240 × 240 × 155 voxels (BraTS benchmark), with FLOP estimates derived from primary publications where available and bounded by theoretical complexity formulas otherwise, with estimated values explicitly distinguished throughout. Results:Across all 94 included studies, the pooled diagnostic accuracy was 93.5% (95% CI: 92.7%-94.4%); however, confirmed publication bias (Egger's p = 0.043) indicates this represents an upper-bound approximation rather than an unbiased population estimate. Because subgroup study counts were insufficient for formal random-effects pooling (CNN-only: n = 3; Transformer-only: n = 2; CNN-Transformer hybrid: n = 4; minimum recommended n = 10 per subgroup), no subgroup meta-analysis was performed. Instead, descriptive mean accuracies are reported as hypothesis-generating observations only: CNN-only models 91.7%, Transformer-only models 93.6% and CNN-Transformer hybrid models 94.6%. These figures must not be interpreted as pooled meta-analytic estimates; they reflect mean observed accuracy across a small number of included studies and are reported solely to illustrate directional trends consistent with the mechanistic rationale for hybridisation. Substantial heterogeneity was observed (I2 = 78.3%; p < 0.001). Three integration paradigms were identified: sequential (45% of models; 93.8% accuracy; 1.8 GFLOPs), parallel (32%; 94.3%; 2.8 GFLOPs) and hierarchical (23%; 94.9%; 3.5 GFLOPs). Parallel architectures demonstrated optimal clinical viability, balancing accuracy with a mean inference time of 2.1 s. GAN-based augmentation improved rare tumour class detection by 7%-10%, with conditional GANs outperforming vanilla architectures. Multimodal MRI + PET fusion achieved 94.2% accuracy at 2.8 GFLOPs, whereas triple-modality integration yielded marginal additional gains (95.1%) at substantially elevated computational cost (9.1 GFLOPs). Notably, 65% of included studies used the BraTS benchmark exclusively, and hybrid model accuracy declined from 94.6% on high-grade gliomas to 88.3% on low-grade gliomas, with hybrid architectures exhibiting 2.3× greater susceptibility to Gaussian noise than CNN-only equivalents, limitations that constrain generalisation to real-world clinical settings. Conclusions:Descriptive comparison of mean observed accuracies based on study counts is insufficient for confirmatory meta-analysis, suggesting hybrid CNN-Transformer architectures may offer diagnostic accuracy advantages over CNN- and Transformer-only approaches; this observation is hypothesis-generating only and requires validation in a larger, more balanced evidence base. Among integration strategies, parallel architectures demonstrated the most favourable accuracy efficiency balance in the reviewed evidence. GANs and multimodal imaging function as essential architectural enablers, addressing data scarcity and diagnostic incompleteness, respectively. Significant challenges remain in computational efficiency, noise robustness and generalisation to rare tumour subtypes, representing priority directions for future research.

The Rivest–Shamir–Adleman (RSA) cryptosystem is one of the most prevalently utilized public-key cryptographic systems in current practice. Prior investigations into vulnerabilities of this cryptosystem have concentrated on diminishing the complexity associated with the integer factorization challenge, which is integral to the RSA modulus, expressed as 𝑁=𝑝𝑞. Possessing partial knowledge about the least significant digits (LSDs) of both p and q is a common assumption attacker’s advantage to enable the polynomial-time factorization of N, ultimately undermining the security of RSA. This paper presents a novel heuristic algorithm predicated on the Constraint Satisfaction Problem (CSP) principles, which estimates k-LSD pairs of the RSA prime factors, and . The proposed Generate and Test (GT) and Backtracking with Heuristic Variable Ordering (BHVO) solver guarantees polynomial-time factorization of known bits by iteratively refining candidate pairs and eliminating invalid combinations through effective constraint propagation. The proposed approach obviates the requirement for specialized hardware for side-channel attacks to reveal a portion of and . In our results, we have successfully estimated up to 5-LSDs of and with a reduced number of iterations and factored 2048 bits, N based on the known 4-LSDs of the prime in polynomial time. Our research lays the groundwork for factorization algorithms that require partial knowledge of the prime factors. We have highlighted the possible vulnerabilities linked to existing RSA key generation techniques. These may make RSA moduli susceptible to the attacks discussed in this study and proposed countermeasures to ensure secure prime generation.

At present, prime numbers are an essential part of online security for secure key generation and digital signatures due to their unpredictabil-‎ity and resistance to factorization attacks, which has led to considerable interest in research on distributions of primes, especially prime gaps ‎and hidden patterns. Traditional prime analytical methods lack structural insight at the scale of cryptographic systems. This has led to greater ‎interest in applying machine learning (ML) techniques to prime analysis. However, due to a lack of publicly available high-quality datasets, ‎the evaluation of prime behavior using ML is often impeded. Furthermore, current prime generation techniques tend to become computa‎tionally inefficient when scaled to produce large quantities of high-bit-length prime numbers. In response, this paper proposes a novel generative pipeline, GANLPN, that leverages Generative Adversarial Networks (GANs) combined with the Learning Parity with Noise (LPN) ‎problem to generate bulk primes that are cryptographically secure. In this way, a publicly accessible dataset of 1,115,000 1024-bit primes ‎has been generated. The generated sequences passed all NIST SP800-22 statistical randomness tests with better inference time and throughput. The present work outlines a path for future ML-assisted studies of prime patterns, gaps, distributions, and the presence of weak keys in ‎cryptography.

Objectives: To address the urgent need for forensic systems capable of detecting and analyzing advanced persistent threats in hybrid quantum-classical communication infrastructures, particularly those that may compromise quantum key distribution environments. Method: The study introduces a Quantum-Aware Forensics Investigation Framework, a multi-layered forensic architecture combining quantum telemetry, classical metadata analysis, and machine learning-driven threat classification. Experimental validation was conducted using a simulated testbed built with SimulaQron, Wireshark, and custom scripting tools. Various quantum attack scenarios were emulated, including intercept-resend, entanglement flooding, and control-plane hijacking. Machine learning models Random Forest, SVM, and Autoencoder were tested as standalone classifiers. A stacked ensemble model, with Random Forest and SVM as base learners and Logistic Regression as meta-classifier, was implemented for performance optimization. We used an experimentally generated, cross-layer dataset from a SimulaQron BB84 QKD emulation by combining quantum logs and classical control-plane captures under benign and scripted attacks such as intercept–resend, entanglement flooding, payload obfuscation, session hijacking, spoofing. Parameters studied were quantum - QBER, event inter-arrival jitter, event/count rate and classical - packet/flow statistics, inter-arrival mean/variance, latency proxy, TCP SYN/RST flags, byte-level Shannon entropy, with labels for benign vs. attack class. Findings: The standalone models achieved moderate performance on the held-out test set for Random Forest: ROC AUC = 0.93, F1 = 0.90, MCC = 0.86, Brier = 0.072; SVM (RBF): ROC AUC = 0.91, F1 = 0.88, MCC = 0.82, Brier = 0.081; Autoencoder (one-class): ROC AUC = 0.87, F1 = 0.83, MCC = 0.74, Brier = 0.094. By contrast, the stacked ensemble delivered perfect detection metrics for ROC AUC = 1.00, F1 = 1.00, MCC = 1.00, and Brier = 0.014. The study further emphasized the need for forensic systems to support explainability and continuous adaptability via Explainable AI and online learning with drift detection. Novelty: This study presents a cross-layer forensic framework for quantum–classical hybrid networks that fuses QKD telemetry with classical control-plane evidence and machine-learning analytics. Unlike prior work that treats these planes separately, our design unifies event-level QKD signals such QBER, arrival-time jitter with packet/flow features to produce timestamp-aligned, explainable alerts. In evaluation, the stacked-ensemble detector achieved perfect detection metrics for ROC AUC, F1, MCC and Brier on held-out data, which distinctly outperformed single-model baselines. The framework couples these gains with an XAI layer and an online, drift-aware learning loop, providing a scalable, auditable, and resilient foundation for forensic intelligence in the quantum era. Keywords: Quantum network forensics, QKD security, Advanced threat detection, Hybrid quantum-classical networks, Quantum-safe evidence, SimulaQron, Quantum cybersecurity

In January 2021 and January 2022, COVID-19 caused roughly 13,000 and 6,000 deaths respectively per day. In August 2022, 26,000 deaths per day were estimated to be caused by COVID-19, followed by 13,000 deaths per day in February 2024. The timely identification and treatment of malignant diseases can potentially lower the mortality rate. Nonetheless, the use of manual methods for diagnosing these conditions requires a meticulous and comprehensive examination, making it susceptible to errors, burdensome for healthcare professionals, and timeintensive. Hence, the objective of this study is to design and deploy a novel deep-learning model for the detection of COVID-19 thoracic diseases. A Convolutional Neural Network (CNN) with less trainable parameters was implemented. This proposed model was deployed on a mobile device using Android Studio and Flutter for the detection of COVID-19 thoracic diseases. Specificity, accuracy, precision, sensitivity, f1-score, ROC, and PR curves were used to evaluate the model's performance. Moreover, the carbon footprint as well as how responsible the proposed model is according to Responsible AI rules was also assessed. The model's evaluation results show an overall accuracy of 93.27 %, specificity of 97.33 %, precision of 93.75 %, sensitivity of 94.42 %, F1-Score of 94.06 %, ROC rate of 98.0 %, PR rate of 96.8 %. The evaluation of the mobile application shows higher generalizability on the COVID-19 dataset. Also, the overall FACETS Score representing responsible AI is 83 % and the carbon footprint (representing the amount of carbon emission emitted into the environment during model training and testing) of 416.73 g with equivalent tree months of 0.45 was obtained. This application with better performance and a low carbon footprint was deployed using Android Studio and Flutter and can assist physicians in the diagnosis of COVID-19 and related diseases.

Download This Paper Open PDF in Browser Add Paper to My Library Share: Permalink Using these links will ensure access to this page indefinitely Copy URL Copy DOI

Artificial Intelligence (AI) has been evident in the agricultural sector recently. The objective of AI in agriculture is to control crop pests/diseases, reduce cost, and improve crop yield. In developing countries, the agriculture sector faces numerous challenges in the form of knowledge gap between farmers and technology, disease and pest infestation, lack of storage facilities, among others. In order to resolve some of these challenges, this paper presents crop pests/disease datasets sourced from local farms in Ghana. The dataset is presented in two folds; the raw images which consists of 24,881 images (6,549-Cashew, 7,508-Cassava, 5,389-Maize, and 5,435-Tomato) and augmented images which is further split into train and test sets. The latter consists of 102,976 images (25,811-Cashew, 26,330-Cassava, 23,657-Maize, and 27,178-Tomato), categorized into 22 classes. All images are de-identified, validated by expert plant virologists, and freely available for use by the research community.

Phishing remains a significant and ever-growing threat in our digitalized society. Its impact on industries crucial to human welfare, such as healthcare, can be devastating and even life-threatening. This paper presents a comprehensive exploration of various types of simulated phishing exercises in the healthcare sector, encompassing incentive methods, tools, risks, and ethical considerations through a scoping review.In this study, a total of six phishing simulation studies and sixteen simulated phishing tools were identified. Among the three types of simulated phishing studies (survey, lab-based experiment, in-the-wild study), a combination of survey and in-the-wild-field study, or a hybrid approach, has been commonly employed for conducting phishing simulations in healthcare settings. While the protection motivation theory has been examined to enhance phishing practices, no measures of incentives have been evaluated to improve resistance against phishing attacks.Regarding risk and ethics, the contentious issue of deception in research has sparked considerable debate across various communities. Nevertheless, employing the "to deceive" approach in in-the-wild-field studies has proven to be the most effective and efficient. However, it is imperative to adopt appropriate ethical measures to mitigate the risks associated with deceitfulness in such studies.Based on the findings of this review, a framework has been developed to guide future studies focusing on effective incentive-related approaches to reduce susceptibility among healthcare workers. This framework serves as a valuable resource for researchers and practitioners aiming to address the ongoing threat of phishing in the healthcare industry.

This study aims to investigate the relationship between stress levels among hospital staff and their risky cybersecurity practices. A web-based survey was conducted with a sample of 353 hospital staff from Ghana, Norway, and Indonesia. The results indicate a statistically significant positive correlation between the stress levels of hospital staff and their engagement in unsafe cybersecurity practices (r = 0.201, p < 0.01). Specifically, the study finds that staff members' inclination to click on links from unknown sources is the cybersecurity practice most strongly influenced by stress levels. The study did not observe any significant differences in cybersecurity practices based on gender, age, job, position level, or work experience. However, it does highlight notable differences in cybersecurity practices across countries, with Norwegian hospital staff exhibiting better cybersecurity practices than their counterparts from Ghana and Indonesia.

Several topics, problems, and established legal principles are already being challenged using artificial intelligence (AI) in numerous applications. The powers of AI have been snowballing to the point where it is evident that AI applications in law and various economic sectors aid in promoting a good society. However, questions such as who the prolific authors, papers, and institutions are, as well as what the specific and thematic areas of application are, remain unanswered. In the current study, 177 papers on AI applications in law published between 1960 and April 29, 2022, were pulled from Scopus using keywords and analysed scientometrically. We identified the strongest citation bursts, the most prolific authors, countries/regions, and primary research interests, as well as their evolution trends and collaborative relationships over the past 62 years. The analysis also identified co-authorship networks, collaboration networks of countries/regions, co-occurrence networks of keywords, and timeline visualisation of keywords. This study concludes that systematic study and enough attention are still lacking in AI application in law (AIL). The methodical design of the required platforms, as well as the collecting, cleansing, and storage of data, and the collaboration of many stakeholders, researchers, and nations/regions are all problems that AIL must still overcome. Both researchers and industry professionals who are devoted to AIL will find value in the findings.

In the contemporary wave of digital transformation, the implementation of electronic health records (EHRs) has become a pivotal undertaking for numerous nations. However, amidst this technological advancement, a critical facet deserving heightened attention is the security and privacy of these electronic health systems. Regrettably, this crucial concern often finds itself eclipsed by other aspects of digitalization. Consequently, these oversight lapses create vulnerabilities within the EHR framework, leaving them open and exposed to an array of malicious cyber intrusions. In response to this pressing issue, our study delves into a comprehensive evaluation of security measures within the ambit of African digital health strategies. Remarkably, among the number of approximately 42 nations that have embarked on digital health strategy formulation, a mere 2 countries have taken cognizance of the imperative to integrate robust security and privacy policies into their healthcare-oriented digital transformation initiatives. In light of this disconcerting revelation, we present an actionable roadmap that endeavours to fortify EHR security, aligning with the progressive "shift-left" paradigm. By advocating for the proactive integration of security measures from the inception of EHR development, we strive to curtail vulnerabilities and enhance the overall resilience of these systems. Our proposed roadmap stands as a clarion call for governments, healthcare authorities, and technology stakeholders to collectively prioritize security in tandem with digital health advancement, thereby fostering a safeguarded and privacy-respecting electronic healthcare landscape.

Quantum blockchain is a distributed database that is decentralized, encrypted, and based on quantum information theory and computation. This comes as a result of the recent progress made in quantum computing and the need for quantum equivalents of classical blockchains. Algorithms, frameworks, models, tools, architectures, and databases, of which quantum blockchain is a part, are still being standardized. Recently, the growth of quantum information theory and computation has resulted in a rise in the number of research ongoing in this domain. This chapter presents an insight into quantum blockchain using the PRISMA technique with results registered and analyzed. The literature is analyzed based on some parameters or categorizations accompanied by graphical and tabular representations.

Data breaches in healthcare have become common in recent times due to the weakness of the human element. As a result, intrinsic and extrinsic motivations were identified, analyzed, and assessed through a literature survey. After a critical gap analysis of the related studies, a framework was designed. This can be used to practically assess the effectiveness of various motivational methods for incentivizing healthcare staff toward strengthening the human aspect of security practice.

A phishing attack is one of the less complicated ways to circumvent sophisticated technical security measures. It is often used to exploit psychological (as as well as other) factors of human users to succeed in social engineering attacks including ransomware. Guided by the state-of-the-arts in a phishing simulation study in healthcare and after deeply assessing the ethical dilemmas, an SMS-based phishing simulation was conducted among healthcare workers in Ghana. The study adopted an in-the-wild study approach alongside quantitative and qualitative surveys. From the state-of-the-art studies, the in-the-wild study approach was the most commonly used method as compared to laboratory-based experiments and statistical surveys because its findings are generally reliable and effective. The attack results also showed that 61% of the targeted healthcare staff were susceptible, and some of the healthcare staff were not victims of the attack because they prioritized patient care and were not susceptible to the simulated phishing attack. Through structural equation modelling, the workload was estimated to have a significant effect on self-efficacy risk (r = 0.5, p-value = 0.05) and work emergency predicted a perceived barrier in the reverse direction at a substantial level of r = −0.46, p-value = 0.00. Additionally, Pearson’s correlation showed that the perceived barrier was a predictor of self-reported security behaviour in phishing attacks among healthcare staff. As a result, various suggestions including an extra workload balancing layer of security controls in emergency departments and better security training were suggested to enhance staff’s conscious care behaviour.

Capsule Networks have shown great promise in image recognition due to their ability to recognize the pose, texture, and deformation of objects and object parts. However, the majority of the existing capsule networks are deterministic with limited ability to express uncertainty. Many of them tend to be overconfident on out-of-distribution data, making them less trustworthy and hence reducing their suitability for practical adoption in safety-critical areas such as health and self-driving cars. In this work, we propose a capsule network based on a variational mixture of Gaussians to train distributions of network weights as opposed to a single set of weights and enable the model to express its predictive uncertainty on out-of-distribution data. Training distributions of weights have the added advantage of avoiding overfitting on smaller datasets which are common in health and other fields. Although Bayesian neural networks are known to exhibit slow training and convergence, experimental results show that the proposed model can retrieve only relevant features, converge faster, is less computationally complex, can effectively express its predictive uncertainties, and achieve performance values that are comparable to the state-of-the-art models. This is an indication that CapsNets can exhibit the transparency, credibility, reliability, and interpretability required for practical adoption.

The global impact of the COVID-19 pandemic has been felt in diverse ways. Although the death rate in Africa has not been as devastating as predicted by the World Health Organization (WHO), its economic and social impact has been fully felt by the African continent. As the world goes through the vaccination process to achieve herd immunity, Africa has not only faced problems like the inability to produce and procure vaccines, but some countries in the west are doubting the authenticity of the vaccination process and even vaccine certificates coming from various countries on the continent. The approach of using centralized systems to validate COVID-19 vaccine certificates makes these systems susceptible to Denial of Service (DoS), modification, and Man-in-the-Middle (MiTM) attacks. To curb this problem, we proposed a blockchain-based digital COVID-19 vaccination certificate verification system called BLOCOVID. The proposed system uses the decentralized approach of distributed ledgers to ensure that vaccine certificates are secured, immutable, and verifiable. Our proposed system stores vaccine serial numbers and their corresponding certificates as hash values. These hash values are stored on the blockchain network as transaction values. The authenticity of a vaccine certificate is determined by the availability of the hash values of the certificate and its corresponding vaccine serial number on the blockchain network. The proposed system was simulated using the BlockSim simulator. To begin with, the simulation results show that the proposed system can ensure system availability, thereby minimizing DoS attacks. Secondly, the proposed system can ensure the integrity of vaccine certificates by allowing third parties to verify the authenticity of these certificates. The simulation results show that even with 10240 nodes, the average transaction time was 137.2ms, with a total transaction rate of 9911.034 transactions per second.

Abstract Sentiment Analysis (SA) is a field of text mining research that is still evolving. SA is the algorithmic treatment of text's opinions, sentiments, and subjectivity to determine if a text contains negative, positive, or neutral feelings. We present a thorough introduction to sentiment analysis. This approach is deduced in a simple, intuitive manner, and implementation advice is provided. The many types, uses, challenges and techniques for sentiment analysis, as well as examples, are also explored in this study. The major goal of this survey is to provide a near-complete picture of SA techniques and related topics, compare sentiment analysis and social network analysis in the latter section of the paper, highlighting the distinctions and how they can both be used along with brief information.