Iron oxide nanoparticles (t-CA-Fe₂O₃ NPs) were synthesized in this study using a green preparation method with trans-cinnamaldehyde (t-CA) as a reducing agent. After observing the first sign of successful preparation, namely the color change, the properties were studied using X-ray diffraction (XRD), field emission scanning electron microscopy (FESEM), energy-dispersive X-ray spectroscopy (EDX), ultraviolet-visible spectroscopy (UV-Vis), and Fourier transform infrared spectroscopy (FTIR) techniques. All characterization results were satisfactory, indicating the success of the preparation process and the purity of the prepared nanoparticles (NPs). The molecular docking process of several proteins important in antibiotic resistance and virulence in target bacteria was also studied. The highest binding affinity was -7 kcal/mol with the A9N5G2 protein. When studying the inhibitory activity of t-CA-Fe₂O₃ NPs against the tested bacteria, they showed excellent inhibitory activity, with the highest activity recorded at a concentration of 100 mM, resulting in an inhibition rate of 25.60 ± 0.56 mm. The effects of the t-CA-Fe₂O₃ NPs on the WRL-68 and GBC-X1 cell lines were also evaluated: the viability of cancer cells decreased compared to normal cells, with an IC50 of 125.3 µg/mL for normal cell lines versus 42.11 µg/mL for cancer cell lines. The objective of this research is to rapidly produce t-CA-Fe₂O₃ NPs by an eco-friendly method and evaluate the therapeutic efficacy of t-CA-Fe₂O₃ NPs against S. enterica serovar Paratyphi B, isolated from gallbladders removed from patients with gallstones, as well as their anticancer effects against a gallbladder cancer cell line; their potential mechanism of action against a selected group of bacterial proteins was assessed using molecular docking analysis.
This study describes the biosynthesis of silver nanoparticles (AgNPs) using an aqueous Salvia rosmarinus L. (SRL) extract (Family: Lamiaceae) as a natural source of reducing and capping agents through an easy and green method assisted by blue diode laser irradiation (450 nm, 5 W, 5 min). The formation of rosemary-mediated AgNPs was verified by observing a color change in the solution and the presence of a surface plasmon resonance (SPR) band at 454 nm in the UV–Vis spectrum. XRD and FESEM revealed crystalline silver-containing nanostructures with an average estimated crystallite size of 15.53 ± 3.54 nm and particles with an average size of 34.10 ± 14.15 nm, respectively. The presence of silver was confirmed by EDX, while FTIR analysis suggested that functional groups derived from rosemary may have contributed to the reduction of silver ions (Ag⁺) and/or the capping and stabilization of the nanoparticles. The laser-irradiated AgNPs (IR-AgNPs) inhibited clinical isolates of Candida albicans, Candida lusitaniae and Staphylococcus aureus, with minimum inhibitory concentration (MIC) values of 0.125, 0.0625 and 0.25 mM and maximum inhibition zones of 20.5, 22.0 and 17.8 mm, respectively. Furthermore, MTT analysis revealed higher cytotoxicity against ME-180 cancer cells than against normal HdFn cells, with IC₅₀ values of 131.8 and 198.4 µg/mL, respectively (selectivity index ≈ 1.51). These in vitro results indicate that IR-AgNPs possess antimicrobial activity and moderate selective cytotoxicity; however, because non-irradiated particles and in vivo models were not examined, further studies are required to establish the contribution of laser irradiation and the potential of IR-AgNPs for biomedical applications.
Human operators frequently encounter eye fatigue that typically comes from workload and sleep deprivation. This could affect the efficiency and performance of vigilance significantly. In safety-sensitive applications such as traffic systems, healthcare, and security, this issue is essential. To address this issue, there is a need for situational awareness and maintaining continuous monitoring of the workload of the person. To this end, advanced artificial intelligence (AI) methods, denoted by a deep learning (DL) based approach, can be used. Such methods are also supported by computer vision (CV) to detect and identify obscured objects. Combining CV with AI can overcome the challenges related to image/video detection in realistic scenarios, such as degradation due to the environmental conditions denoted by rain attenuation, smoke, fire, fog, and other realistic conditions and high visual noise. Therefore, DL algorithms can be effectively used to enhance detection through multi-scale feature extraction techniques and transformer-based models. The present paper reviews recent advancements in DL approaches and discusses the related work in CV toward obtaining effective object identification and detection under the challenges of realistic conditions.
The problem of automated feature selection is one of the most important aspects of modern recommender systems, as overestimating the number of feature fields used, which can increase model complexity, inference latency, and performance degradation due to redundancy and noise. Although differentiable Automated Machine Learning (AutoML) based feature selection methods, including AutoField, are effective, they remain susceptible to initialization and local optima. Conversely, population-based metaheuristic algorithms are effective at global exploration, but are often inefficient in high-dimensional discrete spaces when randomly initialized. This paper presents AutoField-POA, a hybrid AutoML-metaheuristic framework that combines a differentiable controller with the Pelican Optimization Algorithm (POA) to solve automated feature selection problems. The controller guides population initialization toward promising areas of the search space, whereas POA focuses on feature subsets through a balanced exploration–exploitation search strategy. A composite fitness function is jointly optimized to achieve both predictive accuracy and subset compactness. Large-scale Click-Through Rate (CTR) experiments on the Criteo and Avazu datasets indicate that AutoField-POA consistently achieves better performance than AutoField and baseline methods in Area Under the ROC Curve (AUC), Logloss, efficiency, and stability, and exhibits high transferability across recommender architectures.
Breast cancer is one of the leading causes of mortality among women worldwide, and early and accurate mammographic assessment is important for timely diagnosis and improved patient survival. However, accurate breast lesion detection requires overcoming the issues of low lesion contrast, dense breast tissue, image noise, and variation in mammographic acquisition. In this study, we present a YOLOv8x-based computer-aided detection framework for automated breast lesion localization in full-field digital mammography (FFDM) and digital breast tomosynthesis (DBT) images acquired under heterogeneous clinical conditions. The framework was developed and evaluated using 501 clinical mammographic images obtained from multiple mammography systems, namely GE Digital Mammography for 2D FFDM and GE Senographe Pristina for DBT/3D mammography. The proposed approach integrated expert-guided radiological annotation, dedicated preprocessing, and high-resolution mammographic analysis at 1024 × 1024 pixels to improve lesion localization performance. Lesion regions were manually annotated by a specialist radiologist through clinically guided bounding-box procedures compatible with YOLO-based detection. Experimental results demonstrated accurate lesion localization performance, with precision of 98.87%, recall of 100%, mAP@0.5 of 99.50%, and mAP@0.5:0.95 of 99.16%. External validation further demonstrated promising cross-domain performance under heterogeneous imaging conditions. The findings demonstrate the potential value of clinically heterogeneous data, expert-guided annotation, and external validation for developing deployment-oriented computer-assisted breast lesion localization systems.
