The increasing complexity of digital image manipulation has rendered content authentication a significant issue in multimedia forensics. To enhance detection systems, it is essential to go beyond just semantic-level clues and also identify subtle, low-level inconsistencies that can help distinguish between authentic and altered content. Traditional methods that combine manually designed features with CNN-based or hybrid models often lack the necessary flexibility or granularity. Convolutional neural networks focus on fundamental semantics, frequently overlooking sophisticated indicators of forgery, whereas manually crafted approaches require an understanding of manipulation styles, which limits their scalability. This study tackles this limitation by introducing an innovative multi-perspective detection framework that separates the image analysis process into three complementary visual pathways. The initial module examines high-frequency components, where manipulation artefacts are frequently observed. The second focuses on non-semantic cues, such as residuals and noise patterns, whereas the third investigates spatial context to enhance overall comprehension. Every branch utilizes ResNet18 for feature enhancement, supplemented by global-axis connections to achieve thorough representation learning. Furthermore, we implement edge supervision through a Sobel-based edge extraction module to improve the precision of boundary detection and manipulation localization. The proposed method attains an F1 score of up to 98.26
The rapid adoption of the Industrial Internet of Things (IIoT) in smart manufacturing and critical infrastructure has significantly increased the exposure of industrial networks to sophisticated cyber threats. Ensuring secure communication and reliable threat detection in IIoT environments has therefore become a critical challenge. This study proposes an intelligent Cyber Threat Detection and Response System that integrates a Hybrid Deep Neural Network with the Grey Wolf Optimizer to enhance security in IIoT networks. The proposed framework utilizes CyberTec IIoT Malware Dataset (CIMD‑2024) on Kaggle containing network traffic characteristics, device communication patterns, and anomaly indicators. A comprehensive data preprocessing phase is employed, including noise removal, normalization, and missing value handling, to improve data quality and model reliability. The hybrid deep learning architecture combines Convolutional Neural Networks for spatial feature extraction with Long Short-Term Memory networks to capture temporal dependencies in network behavior. Additionally, a dual-attention mechanism is incorporated to emphasize significant spatial and temporal features, thereby improving the accuracy of cyber threat classification. The Grey Wolf Optimizer is applied to optimize key hyperparameters such as learning rate, dropout rate, and batch size, leading to improved model performance. Experimental results demonstrate that the proposed model achieves an accuracy of 96.5
Chronic kidney disease (CKD) is a progressive medical condition with significant public health impact, where early detection is critical for effective intervention. This work presents a clearly defined and structured data-driven framework designed to improve CKD prediction through robust preprocessing and optimized ensemble learning. This study proposes a novel hybrid framework that combines Random Forest (RF)-based imputation for missing value handling, categorical feature encoding, and synthetic minority oversampling technique (SMOTE) for addressing class imbalance, integrated with a Grey wolf optimizer (GWO)-based weighted ensemble of top-performing classifiers (Decision Tree, Logistic Regression, and Gaussian Naïve Bayes). The ensemble weights are optimized using the Grey Wolf optimizer (GWO) to enhance predictive accuracy. We evaluate the proposed framework on the UCI CKD dataset, demonstrating that it outperforms individual classifiers and conventional ensemble methods, achieving an accuracy of 98.75%, precision of 98.8%, recall of 98.6%, and F1-score of 98.7%. Additionally, explainable AI (XAI) techniques including SHAP and LIME are employed to analyze feature contributions, providing interpretable insights and confirming the clinical relevance of the predictions. Overall, the proposed framework offers a transparent, reliable, and computationally efficient clinical decision support model that bridges the gap between data-driven AI and nephrology practice.
This paper presents a novel design approach to implement higher fan-in logic functions in the Positive Feedback Source Coupled Logic (PFSCL) style. The existing two design approaches use either NOR gates or fundamental cells that infer the cascading of multiple gates. The new design approach realizes three input logic functions in a single gate and can be further extended for high fan-in circuits. The circuits based on the proposed design approach reduce the cascading of gates by using a new circuit element called a quad-coupled (QC) cell. The design of a 3-input exclusive-OR (XOR3) gate is proposed and compared with the existing XOR3 gates using LT SPICE simulations with PTM 90nm CMOS technology parameters. The performance is compared in terms of propagation delay, power consumption, power delay product (PDP), gate count and transistor count. The proposed PFSCL QC XOR3 gate outperforms existing PFSCL XOR3 gates in all performance parameters. It is observed that the proposed PFSCL QC XOR3 gate achieves a maximum and minimum reduction of 81%, 86%, 97% and 87.5% and 73%, 50%, 86%, and 50% in propagation delay, power consumption, PDP and gate count values as compared to the conventional PFSCL XOR3 gate respectively. Furthermore, the Monte Carlo analysis and performance across various process corners is performed to ensure robustness. Additionally, the design example of a full adder is added to demonstrate the versatility and applicability of the proposed QC design approach in complex logic circuitry.
Medical image fusion is essential for merging functional metabolic data and structural tissue properties, which are exceptionally difficult to record in a single image because of intrinsic sensor constraints. Through the provision of more comprehensive images, this fusion improves healthcare diagnosis and patient care. Despite the widespread usage of deep learning-based methods, their inability to simultaneously learn long-range and shortrange interdependence impairs fusion capabilities. This study introduces new hybrid multimodal healthcare image fusion system that combines Convolutional Neural Network (CNN) with Swin Transformer (ST) architectures to tackle this issue and improved feature representation. In addition, a new PCA-based block is shown to choose high-gradient features, therefore enhancing image quality and generating better fusion results with 6% fewer feature component. An attention-based fusion technique guarantees the best combination of diverse characteristics from different modalities of healthcare imaging. The proposed method outperforms all existing comparison models with the lowest mean metric error score of 9.99. It outperforms current benchmark approaches by a minimum of 2.25%. Furthermore, the system is robust and reliable, as evidenced by statistical testing and Explainable Artificial Intelligence (XAI) approaches. The stability of our model guarantees a substantial enhancement in diagnostic accuracy, making it a priceless asset for the healthcare industry, especially for radiologists and physicians.
