Arab Academy for Science, Technology, and Maritime Transport

Lung cancer remains a global health challenge that requires early and accurate diagnosis through medical imaging analysis. This study introduces ARXAF-Net framework which integrates Active Reinforcement deep leaning with strategic feature engineering, selection, advanced classification techniques with Explainable AI. Firstly, The ARXAF-Net framework overcomes the challenges of labeled dataset limitation by leveraging active reinforcement learning where the model achieves a remarkable 99.0% training accuracy using reinforcement learning. After that, Traditional feature extraction techniques, including GLCM and LBP, are combined with CNN with attention fusion model features, forming comprehensive vectors. Advanced techniques pinpoint essential characteristics for classification. The experiments conducted on a dataset comprising 30,020 CT images categorized into two classes—15,010 non-cancer and 15,010 cancer—demonstrate that CNN models employing attention fusion with traditional feature extraction achieve remarkable consistency, reaching a testing accuracy of 99%. Basic CNN models with traditional feature extraction, following normalization, display commendable performance, nearing an accuracy of 95%. Additionally, integrating explainable AI (XAI) into the performance frameworks significantly enhances the outcomes by incorporating feedback from radiologists. This research offers valuable insights into the optimal combinations of preprocessing, feature engineering/selection, and classification algorithms aimed at maximizing lung cancer detection efficacy. It also recognizes the trade-offs between accuracy and efficiency when merging deep and traditional features, highlighting the importance of careful feature selection. Moreover, addressing the challenges in this integration and investigating hyper-parameter tuning for machine learning models may present avenues for future improvements.

Caviar/roe, widely valued in modern cuisine, is little characterized regarding its complete metabolite composition. Existing studies focused primarily on specific classes e.g., lipids and proteins. This study represents the first extensive GC-MS metabolite profiling of 48 caviar/roe samples from 10 commercially important taxa, including male and female aquatic animals. One hundred thirty-nine primary and secondary metabolites were identified and quantified belonging to fatty, amino, and organic acids, sugars, nitrogenous compounds, and steroids/terpenoids. Multivariate data analysis further uncovered clear interspecies and gender-specific metabolic differences. OPLS-DA highlighted palmitic acid and cholesterol as discriminative markers for sturgeon caviar, while serine and urea distinguished salmon roe. Gender differences were evident in Charybdis natator roe, with males enriched in amino acids and females in fatty metabolites. This comprehensive caviar/roe metabolite map proposes male gilt-head bream Sparus aurata and female common cuttlefish Sepia officinalis for further investigation of their potential functional food applications, driven by their rich omega-3 PUFA levels. Specifically, S. officinalis roe presents good fat source being rich in γ-tocopherol, with favorable n-3/n-6 ratio. The rich content of pyroglutamic acid in S. aurata may contribute to its characteristic umami taste encouraging further sensory analysis. The findings strengthen the molecular basis for improved quality assessment and nutritional labeling of caviar/roe products.

Large Language Models (LLMs) combined with Retrieval-Augmented Generation (RAG) frameworks greatly improve natural language processing performance, but they incur substantial computational overhead because many similar queries are processed repeatedly. To mitigate this, semantic caching has been introduced to store past responses and reuse them for semantically similar inputs, thereby reducing computation costs. Yet, semantic caching mechanisms that depend only on semantic similarity are vulnerable to adversarial exploitation: carefully engineered malicious queries with minor lexical variations can trigger incorrect cache hits, undermining both the reliability and the security of the system. This paper examines security vulnerabilities in semantic proximity caching systems such as GPTCache, a widely used open-source semantic cache that exemplifies these issues, and introduces a new approach called SAFE-CACHE, which is built to withstand adversarial attacks. SAFE-CACHE adopts a cluster-centroid-based caching strategy that is fundamentally distinct from GPTCache's single-query embedding method. It uses unsupervised clustering of historical query-answer pairs, statistical detection of noisy clusters, bi-encoder-based refinement, and conditional cluster enrichment driven by a fine-tuned lightweight LLM to infer the underlying intent of cached queries. During runtime, incoming queries are compared to cluster centroids instead of individual cached entries, enabling stronger semantic validation and improved resilience against adversarial behavior. Our experimental evaluations demonstrate that SAFE-CACHE dramatically reduces adversarial attack success rates from 52.77% to 14.27% compared to GPTCache, representing up to 72% improvement in adversarial resistance.

Abstract Background: HIV stigma among healthcare providers remains a barrier to prevention and effective patient care, particularly in Egypt. In a national cross-sectional survey (August 2025) we measured HIV knowledge, stigma, and perceived HIV curricular adequacy among medical students. Methods: Data were collected via a bilingual (Arabic/English) online questionnaire distributed by stratified convenience sampling across university type, region, and academic year. The survey included Arabic and English versions of the Brief HIV Knowledge Questionnaire (HIV-KQ-18) and the Healthcare Providers HIV/AIDS Stigma Scale (HPASS). Pearson’s correlation and multivariable linear regression were conducted to identify factors associated with HIV stigma. Results: A total of 1,503 students participated (mean age 20.6 years; 57.4% female), half of which (48.9%) rated curricular coverage of stigma and psychosocial aspects of HIV as inadequate. The mean HIV-KQ-18 score was 8.96/18 (SD 4.26). Only 39.9% recognized that HIV cannot be transmitted through kissing, 33.1% believed washing after sex is protective, and just 41.3% knew that not all infants born to mothers with HIV will have AIDS. The mean HPASS score was 60.1/108 (SD 17.6). Most students (76.8%) worried about contracting HIV from patients, 52% believed patients acquired HIV through risky behaviors, and 43.6% endorsed a right to refuse providing care. Knowledge and stigma were inversely but weakly correlated (r = −0.17, p < .001); and higher knowledge predicted lower stigma on multivariable regression (B = -0.68, p < .001). Despite higher knowledge, males reported significantly higher stigma (B = 4.33 p < .001) compared to their female counterparts. Similarly, participants who completed the Arabic form had significantly lower knowledge, and higher stigma (B = 4.17, p < .001). Conclusion: HIV stigma was prevalent among medical students in Egypt, influenced by gender, culture, and knowledge. These findings call for multifaceted interventions and curriculum reform to reduce stigma among future clinicians.

Gastrointestinal angiodysplasia is frequently observed in patients with aortic stenosis and may present with bleeding and anemia. We conducted a systematic review and meta-analysis to estimate its prevalence in this population and to summarize outcomes after valve intervention. Following PRISMA 2020, we searched PubMed, Scopus, and Web of Science and registered the protocol in PROSPERO (CRD42024550839). Eligible observational studies reporting angiodysplasia, von Willebrand factor abnormalities, or both in aortic stenosis were appraised for quality and pooled using a random effects model; heterogeneity and publication bias were assessed with I² and Egger's test. Eleven studies were included. The pooled prevalence of gastrointestinal angiodysplasia among patients with aortic stenosis was 6.3% (95% confidence interval: 4.51-8.38, I² = 98.68, P < 0.0001), with no evidence of publication bias. Across studies that reported longitudinal outcomes, aortic valve replacement or transcatheter aortic valve implantation was associated with a reduction in lesion burden and lower rates of gastrointestinal bleeding, anemia, transfusion, and readmission, although early postprocedural bleeding could occur and typically declined over follow-up. These findings indicate that angiodysplasia is a clinically relevant comorbidity in aortic stenosis and support proactive gastrointestinal evaluation in patients with anemia or unexplained bleeding. Standardized diagnostic criteria and prospective studies are needed to clarify long-term outcomes after valve therapy and to define screening and management pathways.

Organisations today face an increasingly complex and rapidly changing business environment, driving them towards adopting advanced technological innovations to achieve sustainable success. This study focuses on the impact of adopting Internet of Things (IoT) technologies within Industry 5.0 on the sustainable performance of small and medium-sized enterprises (SMEs) in the manufacturing sector, with a particular emphasis on the mediating role of supply chain integration. The primary data were gathered from 424 professionals working in Egyptian manufacturing SMEs by using an online questionnaire. The collected data were analysed through the application of regression analysis, correlation analysis, and structural equation modelling to evaluate the study hypotheses. Results showed that Industry 5.0 (IoT) adoption has a partial significance on the sustainable performance of manufacturing SMEs. This is further supported by the fact that the Industry 5.0 characteristics positively empower supplier, internal, and customer integration. The results also indicated that although the supplier integration is not critical in influencing the environmental performance, it has a strong impact on economic and social performance. Internal and customer integration affects environmental and social performances but does not affect economic performance. The findings are consistent with the study hypotheses that supply chain integration indeed plays a critical mediating role in the relationship between IoT adoption and sustainable performance.

Purpose A comprehensive insight into bioheat transfer and the coupled thermomechanical response within tissues is vital for the precise application of thermal medical interventions. In the proposed model, the tissue surface is assumed to be traction-free and subjected to a time-dependent, exponentially diminishing heat flux. Design/methodology/approach This study explores the thermoelastic behavior of viscoelastic biological tissues exposed to thermal stimuli, employing the Green and Naghdi framework both with and without including energy dissipation mechanisms, as relevant to hyperthermia therapy. This is the first application of the eigenvalue approach in GN-III modeling of viscoelastic tissues. Findings To obtain exact solutions to the governing field equations, the generalized biothermoelastic formulation is addressed using Laplace transformation techniques in conjunction with an eigenvalue-based approach. Originality/value Graphical results for displacement, temperature and stress distributions are presented and a subsequent parametric investigation is carried out to identify optimal design variables that enhance the precision and therapeutic effectiveness of thermal treatments.