جامعة المعرفة
المصدر المعتمد للمعلومات البحثية والمخرجات هو قاعدة بيانات سكوبس (Elsevier Scopus) مع تدقيق بيانات الباحثين من عمادة البحث العلمي.

Research outputs

المخرجات البحثية

مقالات ومراجعات وأوراق مؤتمرات وفصول كتب، مفهرسة برقم DOI ومربوطة بباحثي الجامعة.

79 مُخرَج نتائج البحث عن «Architecture»

مقالوصول مفتوح20262استشهاد· Semantic Scholar

An AI-driven multimodal developmental disability detection and intervention framework using enhanced speech and behavioral analysis with BioNeuroFusionNet classification

Journal Of Big Data · مجلد 13 (1)

Abstract Effective intervention and detection of developmental disabilities (DDs) should be carried out early and accurately. This paper discusses a novel AI-based approach for detecting DDs precisely, by incorporating both speech and behavioral data in a unique manner. The new architecture comprises several key stages, with innovations attributed to the system: Enhanced Gaussian-Based Noise Filtering (EGNF) is used to enhance speech; Conv-BiLSTM and Transformer Embeddings for performing strong feature extraction from speech and behavior; Harmony-ReliefF Optimization (HRO) for feature selection, and BioNeuroFusionNet (BNFN) as a novel classification network integrating IndRNN and MnasNet, with GNA selected as the class superiors. The Harmony-ReliefF Optimization (HRO) strategy combines the exploration capabilities of the Harmony Search Algorithm (HSA) with the feature ranking ability of ReliefF for optimizing the feature space. The features selected and optimally extracted are then fed into the novel BioNeuroFusionNet (BNFN) intended for final classification. The BNFN is another significant contribution of this work, where Independently Recurrent Neural Networks (IndRNN) are incorporated for fast processing of sequential data set and MnasNet (MN) for extracting high-level abstract features. The experimental result indicates that this integrated architecture works effectively with high accuracy (0.98942 for voice; 0.99632 for behavior) and very high scores in precision, recall, specificity, and other relevant metrics. Indeed, these results indicate that the system proposed in this work represents a considerable improvement in the area of early diagnosis and intervention of developmental disorders.

multimodal analysisspeech analysisbehavior analysisearly diagnosis
مقالوصول مفتوح20262استشهاد· Semantic Scholar

Enhancing E-health system accuracy using Rendezvous Data Processing Model (RDPM) with IoT-cloud integration

Digital Health · مجلد 12 · 20552076251406312-20552076251406312

Objective: The study's overarching goal is to improve E-health monitoring systems' precision and performance by developing and implementing an Rendezvous Data Processing Model (RDPM) that is compatible with IoT-cloud architecture. The approach solves a problem with current E-health systems; these systems frequently make incorrect or redundant suggestions because they depend too much on static analytic methods and isolated data augmentation. Methods: The RDPM system recommended improves real-time decision-making by digesting historical suggestions and present analytical flaws. Divided features and data streams allow it to validate new hypotheses by comparing them to earlier observations. The state learning process has been improved by earlier efforts to avoid errors and data duplication, the model must distinguish intervening and non-intervening data. Internet-connected sensors collect massive volumes of patient and environment data. Cloud analytics evaluates the system's precision using these data. Results: Experimental results show that RDPM reduces data interruptions, analytical errors, and recommendation ratios while improving decision correctness. The model shows that it can quickly interpret many input streams without compromising accuracy. Compared to IoT-based healthcare analytics, the RDPM improves suggestion accuracy and reduces computing redundancy. Conclusion: IoT-cloud technologies with the RDPM system establish an adaptive and scalable platform for sophisticated E-health monitoring. State learning and dynamic data validation allow RDPM to make more accurate and convenient health recommendations. This approach allows a healthcare system to self-improve, understand context, and manage massive, real-time datasets.

e-health monitoringIoT-cloud integrationstate learningreal-time decision making
مقالوصول مغلق20261استشهاد· Semantic Scholar

Comparative molecular dynamics reveal ligand-dependent structural and dynamic determinants of dipeptidyl peptidase IV recognition and inhibition

Journal of Biomolecular Structure and Dynamics · مجلد 44 (15) · 6664-6676

Dipeptidyl peptidase 4 (DPP4/CD26) is a clinically relevant serine protease involved in metabolic, immunological and viral processes. Although several crystal structures of DPP4-ligand complexes have been reported, the dynamic features that distinguish substrates from inhibitors remain less understood. In this study, 250 ns all-atom molecular dynamics simulations were performed for four structurally distinct ligand-bound states, a synthetic construct, a Neuropeptide Y (NPY) fragment, Diprotin A and an HIV-1 TAT peptide, to examine how ligand identity affects the conformational behavior of DPP4. All complexes maintained the global α/β-hydrolase and β-propeller architecture, with secondary-structure content remaining stable throughout the simulations. Ligand-dependent differences were mainly observed in the flexible loop regions near the catalytic cavity. The inhibitory ligands (Diprotin A and TAT) promoted reduced mobility and lower backbone RMSD values (1.5-2.0 Å), whereas the substrate and the synthetic construct allowed greater local flexibility (2.0-2.5 Å). Contact analyses showed a conserved anchoring network involving Glu205, Glu206 and Tyr662 across all systems, along with ligand-specific peripheral interactions, including Trp659, Tyr752 and Phe357, that influence the binding modes. These observations reveal a relationship among structure-dynamics-function, in which DPP4 maintains a stable interaction and modulates loop movements, including peripheral contacts, to differentiate substrates from inhibitors. The dynamic behavior of the enzyme offers mechanistic insights that allow for specific ligand recognition, which could support the design and development of new DPP4 modulators.

dipeptidyl peptidase IVCD26substrate recognitionenzyme inhibition
مقالوصول مغلق2026

Diagnostic Accuracy of Artificial Intelligence versus Clinical Diagnosis in Detecting Acute Otitis Media: A Systematic Review and Meta-analysis

Saudi Journal of Otorhinolaryngology Head and Neck Surgery

Abstract Acute otitis media (AOM) represents a prevalent infection in children globally, leading to considerable morbidity and imposing a substantial burden on healthcare systems. Prompt and precise diagnosis is critical to reducing complications, such as tympanic membrane perforation, hearing impairment, and excessive antibiotic prescribing. Advances in artificial intelligence (AI), particularly deep learning, have created opportunities to enhance diagnostic accuracy and broaden access to care, especially in low-resource settings. This systematic review and meta-analysis aimed to evaluate the diagnostic performance of AI-based systems for detecting AOM compared with conventional clinical diagnosis and to analyze their accuracy relative to human clinicians across different model architectures and diagnostic methods. Following PRISMA-DTA standards and PROSPERO registration, systematic searches were conducted in PubMed, Web of Science, and IEEE Xplore up to July 30, 2025. Eligible studies included human participants with suspected AOM, employed AI-driven diagnostic approaches, and used a clinical reference standard. The Quality Assessment of Diagnostic Accuracy Studies-2 tool assessed bias, and pooled sensitivity and specificity were estimated using a bivariate random-effects model. Nine studies published between 2020 and 2025 met the inclusion criteria. The pooled sensitivity and specificity were 91% (95% confidence interval [CI]: 77%–97%) and 99% (95% CI: 94%–100%), respectively. AI models overall matched or outperformed general practitioners and pediatricians, approaching otolaryngologist-level accuracy. AI-based tools have the potential to improve AOM detection, mitigate inequalities in healthcare access, and promote more effective diagnosis, particularly in underserved regions.

acute otitis mediaartificial intelligencedeep learningdiagnostic accuracy
مقالمفتوح (مستودع)2026

Effects of Roasting on the Antioxidant and Lipid‐Modulating Activities of Argania spinosa Oil Under Diabetic Conditions

Food Science & Nutrition · مجلد 14 (10) · e72441-e72441

ABSTRACT Argania spinosa L. is endemic to a small region in southwestern Morocco and has been used for food and medicinal purposes for decades. This study evaluates the hepatoprotective and potential cardioprotective effects of roasted Argan seed oil (HGAT) and non‐roasted Argan seed oil (HGANT) in a streptozotocin‐induced experimental diabetic model. An experimental approach in diabetic rats treated with HGAT or HGANT was used to assess oxidative stress, liver enzymes, and lipid metabolism. Diabetic rats were administered 2 mL/kg body weight/day of HGAT or HGANT. Biochemical analyses measured blood glucose levels, liver enzyme activity (ALT, AST, LDH), oxidative stress markers (MDA, GSH, CAT), lipid profiles, and hepatic glycogen content. Histopathological examinations assessed collagen deposition and fibrosis in liver and heart tissues. Both HGAT and HGANT significantly reduce oxidative stress and improve lipid profiles in diabetic rats. Notably, 2 mL/kg body weight/day of HGAT treatments decreased blood sugar levels to values comparable to those observed with metformin ( p > 0.05). Histopathological analysis revealed that both oils improved hepatic architecture, reduced lipid droplet accumulation, and decreased collagen deposition in the liver. In cardiac tissue, HGAT and HGANT improved myocardial structure and significantly reduced collagen deposition, indicating attenuation of cardiac fibrosis. The roasting process did not affect total cholesterol, triglycerides, hepatic glycogen, or oxidative stress markers. Argan oil demonstrates promising hepatoprotective and lipid‐modulating effects in diabetic conditions, with additional cardioprotective potential, supporting its ethnomedicinal use in metabolic disorders. Further studies are warranted to clarify its underlying mechanisms, long‐term effects, and potential interactions with standard antidiabetic therapies.

Argania spinosaroastingdiabetes mellitushepatoprotection
مقالوصول مفتوح2026

Artificial Intelligence-Driven Sensor Network Approach for Optimizing Halted Product Delivery in E-Commerce Platforms

International Journal of Computational Intelligence Systems · مجلد 19 (1)

Sensor network architectures incorporated with artificial intelligence are exploited for process automation in e-commerce platforms. Specifically, product management is automated through sale-based tracking, enabled by the rapid exchange of information within the sensor network. A Paused Product-based Data Management Scheme is proposed in this article to track and improve the delivery of halted products through e-commerce platforms. The sensor network architecture performs individual identification and tracking of halted/delayed products at any hub through synchronized data updates. The synchronization of product information, delay time, and its associated paused information is performed by identifying the product delivery time. For this purpose, a deep neural network is employed to compute the difference between actual and paused delivery intervals. The higher the difference, the synchronization and delivery re-initialization processes are through the sensor network’s interconnected data exchange. Besides, the network is trained until the delay time is reduced with the re-scheduled time as the base. This infers precise product tracking under fewer missing order complaints in an e-commerce platform.

e-commercesensor networksdeep neural networkdelivery delay prediction
مقالوصول مفتوح2026

Dose-dependent vitamin D3 supplementation attenuates endothelial dysfunction and renal injury in an obstructive sleep apnea rat model

Frontiers in Pharmacology · مجلد 17 · 1847011-1847011

Background Endothelial dysfunction driven by intermittent hypoxia (IH)-induced oxidative stress, nitric oxide depletion, and adhesion molecule upregulation underlies obstructive sleep apnea (OSA)-related cardiovascular complications. Vitamin D deficiency, which frequently coexists with OSA, may further amplify this endothelial injury through impaired endothelial nitric oxide synthase (eNOS) regulation and heightened renin-angiotensin-aldosterone system (RAAS) activity; however, the dose-dependent therapeutic potential of vitamin D3 on combined endothelial and renal outcomes in an IH model remains uncharacterized. Methods Fifty-six adult male Sprague-Dawley rats were allocated to seven groups: control, vitamin D-deficient (VDD), VDD with severe IH (SIH-VDD), and four supplementation subgroups receiving high-dose (37.5 μg/kg/day) or low-dose (25 μg/kg/day) cholecalciferol under severe or moderate IH for 14 days. Serum E-selectin, ICAM-1, ET-1, NO, and PCSK9 were quantified by ELISA. Renal histopathology was assessed on H&E sections by a blinded pathologist. Data were analyzed by one-way MANOVA, univariate ANOVA, and LSD post hoc tests (p < 0.05). Results The SIH-VDD group showed the most severe endothelial activation: peak E-selectin (4.92 ± 0.90 ng/mL), ICAM-1 (201.50 ± 21.55 ng/mL), ET-1 (5.31 ± 0.81 pg/mL), and PCSK9 (213.50 ± 14.71 ng/mL), with the lowest NO (14.88 ± 2.16 μmol/L; all p < 0.001 vs. control). High-dose supplementation produced substantial recovery across all markers; PCSK9 normalized to near-control (162.50 ng/mL) under severe IH. Low-dose supplementation yielded incomplete normalization. Renal histopathology revealed mild interstitial lymphocytic infiltration in IH-exposed groups, with proximal tubular vacuolisation exclusively in the SIH-VDD group and entirely absent in all supplemented groups. Glomerular architecture remained preserved throughout. Conclusion Vitamin D deficiency combined with IH upregulates endothelial dysfunction markers and induces renal tubular injury in an experimental OSA model. High-dose vitamin D 3 attenuates these changes dose- and severity-dependently, supporting its role as a pharmacological adjunct for OSA-associated endothelial and renal complications.

obstructive sleep apneavitamin D3 supplementationvitamin D3endothelial dysfunction
مقالوصول مفتوح2026

Protective effect of hesperetin against myelosuppression and splenic inflammation via modulating TLR4/MYD88/NF-κB signaling: molecular docking and experimental validation

Frontiers in Pharmacology · مجلد 17 · 1826774-1826774

Introduction: Myelotoxicity is a common adverse effect of vincristine and other chemotherapeutic drugs. Hesperetin is a bioactive flavanone with anti-inflammatory activity. This study was planned to examine the hesperetin protecting effect against vincristine induced myelosuppression and splenic inflammation in mice and test whether this effect is mediated through the suppression of Toll-like receptor 4 (TLR4)/myeloid differentiation primary response-88 (MYD88)/nuclear factor-kappa B (NF-κB) signaling pathways. Methods: Bioinformatics tools were used to establish the rationale of the study and molecular docking defined the potential binding of hesperetin to TLR4. Mice were allocated into different experimental groups: Group 1: was the saline mice, Group 2; mice were injected with vincristine (0.1 mg/kg, i. p) for 5 days per week, Group 3; mice received vincristine (as in Group 2) plus hesperetin (100 mg/kg, p. o.) and Group 4; mice received hesperetin (100 mg/kg, p. o.) only; the experiment continued for 2 weeks. Results: 9/L in the saline group, p less than 0.05) and bone marrow hypocellularity (dropped to ∼30-40% compared to 95%-100% in the saline group). The spleens showed white pulp lymphoid depletion, red pulp mild hypocellularity, greater immunostaining for MYD88, and greater content for TLR4, NF-κB, MYD88, TNF-α and IL-1β proteins. Results of molecular docking showed that hesperetin is an interesting complementarity with the TLR4/MD2 interface which was a good rationale for the experimental validation. The mouse experiment showed that hesperetin prevented the decrease hematological parameters, enhanced the bone marrow cellularity to 80%-90%, and the splenic architecture. Hesperetin further inhibited the increases in the target proteins (TLR4, NF-κB, MYD88, TNF-α and IL-1β) significantly. Discussion: The present data documents the novel role of hesperetin in mitigating myelotoxicity of vincristine through modulation of TLR4/MYD88/NF-κB signaling and a direct binding to TLR4 was proposed by molecular docking. This highlights possible usefulness of this combination after adequate clinical studies.

hesperetinvincristinemyelosuppressionsplenic inflammation
مقالمفتوح (مستودع)2026

A Topological and Spatial Analysis of FDA‐Approved Drugs, Blood‐Brain Barrier Permeants, Natural Products, and Metabolites

ChemMedChem · مجلد 21 (17) · e70462-e70462

In this study, we present a comprehensive structural analysis of four distinct chemical libraries: FDA‐approved drugs, blood‐brain barrier (BBB) permeable agents, natural products (UNPDA), and human metabolites. Our multidimensional profiling reveals fundamental divergences in molecular architecture. While synthetic drugs (FDA) and CNS‐active agents (BBB) follow continuous, Gaussian‐like distributions for carbon hybridization, natural products exhibit biosynthetic quantization, discrete, high‐intensity abundance spikes at C19 and C30, corresponding to steroid and triterpene scaffolds. Stereochemical analysis identifies a chirality‐structure lock in natural products, where stereoisomer density peaks significantly at these same integers (Avg. 3.53 isomers/mol), a complexity feature absent in the diffuse distribution of synthetic libraries. Topologically, we identify hyper‐globularity as a distinct signature for CNS permeability. Radial distribution function (RDF) analysis demonstrates that BBB agents possess a highly compressed aliphatic core, packing 93.2% of their sp 3 mass within a 3 Å radius, compared to 84.4% for general FDA therapeutics. Furthermore, elemental profiling reveals a heteroatom gap: synthetic drugs exhibit a high electronegative atom‐to‐carbon ratio (0.43), whereas BBB agents (0.33) converge with the lipophilic profile of natural products (0.32), suggesting that CNS penetration requires a biomimetic reduction in polarity. These findings establish new quantitative metrics for designing neuro‐compatible and biomimetic pharmacophores.

blood-brain barrier permeabilitynatural productshuman metabolitesFDA-approved drugs
مقالوصول مغلق2026

The effect of losartan nanoparticles vs losartan on TIMP1 and α-SMA immunohistochemical and gene expression in carbon tetrachloride-induced liver fibrosis

Biotechnic & Histochemistry · مجلد 101 (7) · 405-418

Liver fibrosis is characterized by scar formation as a response to injury. Although many drugs have demonstrated antifibrotic efficacy in vitro and in animal models, effectively translating these into effective clinical therapies remains a challenge. Nanotechnology offers potential therapeutic approaches for treating liver fibrosis. This study aimed to assess and compare the antifibrotic effects of losartan and losartan nanoparticles (LP-NPs) on CCl4-induced liver fibrosis in rats.Thirty-two adult male Sprague-Dawley rats were divided into four groups: a control group, a CCl4-induced fibrosis group, a losartan-treated group, and an LP-NPs-treated group. Liver fibrosis was induced by CCl4 injections over six weeks. Losartan or LP-NPs were orally administered for four weeks. At the conclusion of the study, rats were euthanized, and blood samples were collected for liver enzyme assessment. Liver specimens were subjected to histopathological examination (H&E, Sirius red staining), immunohistochemistry (α-SMA and TIMP1), and quantitative real-time PCR (qRT-PCR) for gene expression analysis of α-SMA and TIMP1. Morphometrical studies quantified collagen areas in Sirius red-stained sections and optical density of α-SMA and TIMP1 in immunohistochemically stained sections.The results showed that CCl4 induced significantly elevated liver enzymes, severe distortion of hepatic architecture, marked steatosis and fibrosis. Additionally, there was an increase in collagen content, accompanied by strong positive immunostaining and upregulated gene expression of α-SMA and TIMP1. Both losartan and LP-NPs treatments significantly improved these parameters compared to the CCl4-only group. Notably, LP-NPs demonstrated significantly greater improvements in liver enzyme levels, collagen percentage area, and the gene expression of α-SMA and TIMP1 compared to the standard losartan group. However, no significant difference was observed between the two treatments in α-SMA optical density.In conclusion, this investigation demonstrates that nanoformulation significantly enhanced losartan’s antifibrotic efficacy. LP-NPs improved liver function and histological architecture, reduced collagen content after CCl4 administration, and demonstrated greater therapeutic efficacy than conventional losartan.

losartanliver fibrosiscarbon tetrachlorideantifibrotic therapy
مقالوصول مفتوح2026

NeuroTLEDNet: a novel artificial intelligence with squirrel reptilian optimization (SRO) and TLE-SegNet+ for early detection of temporal lobe epilepsy from medical imaging

Journal Of Big Data · مجلد 13 (1)

Temporal Lobe Epilepsy (TLE), a very common type of focal epilepsy, requires early and precise identification to enable prompt treatment and enhance patient outcomes. Traditional methods of detection, most notably manual MRI interpretation, tend to be time-consuming, subjective, and dependent on expert specialists, hindering their scalability and reliability. Due to the limitations of conventional and current automated systems—e.g., poor feature extraction, poor segmentation, and low accuracy in classification—this work presents NeuroTLEDNet, a new end-to-end artificial intelligence model for early TLE detection from medical images. The new methodology integrates multiple series of innovations across every processing phase. In preprocessing, skull stripping, Gaussian blurring, and color normalization are used to improve image uniformity and suppress irrelevant structures. In segmentation, the improved TLE-SegNet + is suggested—based on Mask R-CNN and supplemented with DarkNet-53, DenseNet, HetConv layers, and Attention Feature Pyramid Networks (FPN)—to facilitate highly accurate delineation of epileptogenic areas. For further describing the segmented areas, a rich feature space encompassing color histograms, Local Binary Patterns (LBP), and geometric features (eccentricity, area, perimeter) is developed. To resolve inefficiencies in feature selection, this paper presents the Squirrel Reptilian Optimization (SRO) algorithm, a new hybrid metaheuristic that incorporates both the Squirrel Search Algorithm (SSA) and the Reptile Search Algorithm (RSA). The hybrid enhances discriminative feature selection by handling exploration and exploitation dynamics better compared to individual methods. For categorization, NeuroTLEDNet utilizes an in-depth hybrid ensemble of 3D Convolutional Neural Networks (3D-CNN), Recurrent Neural Networks (RNN), Radial Basis Function Networks (RBFN), and SqueezeNet, facilitating powerful modeling of spatial-temporal relationships in brain MRI scans. Tested on an 80/30 split MRI dataset, the designed model attains a remarkable accuracy of 98.6%, outperforming current methods like CNN-LSTM (94.2%), Mask R-CNN (92.7%), and ResNet-based architectures (93.1%). Indeed, NeuroTLEDNet provides an end-to-end solution for the shortcomings of existing TLE detection methods by utilizing advanced preprocessing, segmentation, feature optimization, and hybrid classification, with state-of-the-art performance in terms of both accuracy and explainability. The suggested model is a promising candidate for clinical application, and it has the potential to assist neurologists in faster and more accurate TLE diagnosis.

temporal lobe epilepsybrain MRIMask R-CNNDarknet-53
ورقة مؤتمرمفتوح (مستودع)2026

AI-Driven Digital Transformation for Sustainable Architecture and Engineering Using Integrated Smart Systems

International Conference on Frontiers of Engineering and Emerging Technologies (FET) · 1-7

Built environment faces unprecedented challenges from climate change, rapid urbanization, and resource scarcity, necessitating innovative approaches to sustainable design and construction. This paper presents an integrated framework that combines artificial intelligence, digital twin technology, wireless sensor networks, and generative design tools to address these challenges. The proposed platform enables real-time monitoring, predictive analytics, and adaptive optimization of building systems throughout their entire lifecycle. By integrating AI-powered generative design with continuous performance assessment, the framework facilitates climate-responsive architecture, enhanced resource efficiency, and improved indoor environmental quality. The methodology incorporates human-centric design principles, passive design strategies, and advanced material selection to create resilient and sustainable built environments. Results from simulated and pilot implementations demonstrate significant reductions in energy consumption, improved thermal comfort, and enhanced operational efficiency compared to traditional building management approaches. The framework's ability to continuously evaluate sustainability metrics while adapting to changing environmental conditions represents a transformative approach to architectural and engineering practice. This paper illustrates how digital transformation through intelligent systems can enable the development of future-ready solutions that balance environmental sustainability with occupant well-being and operational performance.

artificial intelligencedigital transformationsustainable architecturesustainable engineering
مقالوصول مفتوح2026

Federated Learning in Smart Home Intrusion Detection: Securing the Learning Process with Blockchain

Network and Communication Technologies · مجلد 10 (1) · 1-1

The rapid expansion of Internet of Things (IoT) devices has fundamentally reshaped traditional homes into highly interconnected smart environments, driving unprecedented convenience while simultaneously exposing critical cybersecurity vulnerabilities. Conventional centralized Intrusion Detection Systems (IDS) are increasingly inadequate for smart homes, owing to inherent privacy issues, latency, and susceptibility to single points of failure. Federated Learning (FL) offers a promising, privacy-preserving paradigm by enabling collaborative model training without the exchange of raw user data; however, FL remains vulnerable to model poisoning and integrity attacks. This comprehensive review systematically examines the integration of blockchain technology as a means to secure FL-based IDS in smart home contexts. We critically evaluate diverse blockchain frameworks, consensus protocols, and enhancement strategies that underpin the development of trustworthy, transparent, and resilient security architectures. Specifically, the review addresses two core research questions: (1) In what ways are blockchain frameworks and consensus mechanisms leveraged to secure FL-based IDS? and (2) What enhancement techniques and hybrid models are implemented to optimize the performance and efficiency of these integrated systems? Our synthesis reveals that permissioned blockchains—particularly those employing Proof of Authority (PoA) and Practical Byzantine Fault Tolerance (PBFT) consensus mechanisms—when combined with edge computing and lightweight architectural designs, deliver robust security while meeting the resource constraints characteristic of IoT devices. This technological synergy not only strengthens the resilience of the FL process but also advances the development of next-generation, secure smart home ecosystems.

federated learningintrusion detection systemshome automationblockchain
مقالوصول مفتوح202532استشهاد· Semantic Scholar

Multi-scale feature fusion of deep convolutional neural networks on cancerous tumor detection and classification using biomedical images

Scientific Reports · مجلد 15 (1) · 1105-1105

In the present scenario, cancerous tumours are common in humans due to major changes in nearby environments. Skin cancer is a considerable disease detected among people. This cancer is the uncontrolled evolution of atypical skin cells. It occurs when DNA injury to skin cells, or a genetic defect, leads to an increase quickly and establishes malignant tumors. However, in rare instances, many types of skin cancer occur from DNA changes tempted by infrared light affecting skin cells. This disease is a worldwide health problem, so an accurate and appropriate diagnosis is needed for efficient treatment. Current developments in medical technology, like smart recognition and analysis utilizing machine learning (ML) and deep learning (DL) techniques, have transformed the analysis and treatment of these conditions. These approaches will be highly effective for the recognition of skin cancer utilizing biomedical imaging. This study develops a Multi-scale Feature Fusion of Deep Convolutional Neural Networks on Cancerous Tumor Detection and Classification (MFFDCNN-CTDC) model. The main aim of the MFFDCNN-CTDC model is to detect and classify cancerous tumours using biomedical imaging. To eliminate unwanted noise, the MFFDCNN-CTDC method initially utilizes a sobel filter (SF) for the image preprocessing stage. For the segmentation process, Unet3+ is employed, providing precise localization of tumour regions. Next, the MFFDCNN-CTDC model incorporates multi-scale feature fusion by combining ResNet50 and EfficientNet architectures, capitalizing on their complementary strengths in feature extraction from varying depths and scales of the input images. The convolutional autoencoder (CAE) model is utilized for the classification method. Finally, the parameter tuning process is performed through a hybrid fireworks whale optimization algorithm (FWWOA) to enhance the classification performance of the CAE model. A wide range of experiments is performed to authorize the performance of the MFFDCNN-CTDC approach. The experimental validation of the MFFDCNN-CTDC approach exhibited a superior accuracy value of 98.78% and 99.02% over existing techniques under ISIC 2017 and HAM10000 datasets.

skin cancer detectionskin lesion classificationmulti-scale feature fusionconvolutional neural networks