Diabetic retinopathy (DR) is a leading cause of visual impairment worldwide. Early and accurate classification of disease severity using retinal fundus images is critical for timely diagnosis and prevention of vision loss. Automated systems can assist clinicians by improving screening efficiency and reducing diagnostic variability. This study proposes a novel feature-pitched classification model (FPCM) for automated classification of diabetic retinopathy stages using color fundus images, aiming to improve accuracy through enhanced feature extraction and spatial attention mechanisms. The proposed FPCM model extracts texture, intensity, and structural features from fundus images by identifying high-information regions referred to as pitch points. These features are processed using a concentric transformer learning (CTL) mechanism, which applies spatial attention across nested regions to capture both local and global patterns. The model was evaluated using the Kaggle Diabetic Retinopathy dataset. Performance metrics included accuracy, precision, and sensitivity. Statistical reliability was assessed using bootstrap-based confidence interval analysis, and results were compared with existing methods such as ERCN, EffNet-SVM, HPLBO_DMN, FCSAM, CLAHE TH, and MSTNet. The proposed model achieved an accuracy of 95.19
This study investigates the machining characteristics of Wire-Laser Metal Deposition (Wire-LMD) processed Inconel 718 using die-sinking Electrical Discharge Machining (EDM) and its performance in terms of surface roughness and overcut. Initially, the Inconel 718 wire was additively manufactured with a Meltio M450 system. The fabricated as-built Inconel 718 exhibited a density of 8.12 g/cm³ and a Vicker’s microhardness of 337.02 HV0.1 ± 12. The metallurgical examination exhibits a defect-free optical microscope and Field Emission Scanning Electron Microscope (FESEM) microstructure with fine columnar dendrites, interdendritic regions, minimal porosity, and Laves phase formation, confirming high-quality printability. An orthogonal array of Taguchi L9 was used to design experiments with the EDM by changing spark current (2–6 A), spark-on time (20–60 ms), and servo voltage (50–100 V) and then optimized by Analysis of Variance (ANOVA). ANOVA revealed that spark current contributed 85.55
The rapid evolution of Smart Grid (SG) technologies necessitates robust control and monitoring systems. This study introduces an effective SG energy monitoring system utilizing Type 2 Fuzzy Logic Controller (T2FLC), enhanced with the Mayfly Optimization Algorithm (MOA). The MOA plays a crucial role, drawing inspiration from the natural mating behavior of mayflies to perform an efficient search space exploration, thus optimizing the T2FLC parameters including membership functions and rule weights. The system utilizes a Wireless Sensor Network (WSN) at its core for real-time data acquisition of key electrical parameters. The system comprehensively manages both Photovoltaic (PV) systems and Wind Energy Conversion Systems (WECS), ensuring stable power output using the innovative Mayfly Algorithm optimized Type 2 Fuzzy Logic Controller (MF-T2FLC). Its primary function is to compute and generate reference power for the associated DC-DC and AC-DC converters connected to the PV system and WECS, respectively. The efficacy of the MF-T2FLC is thoroughly verified through both laboratory prototype implementations and MATLAB simulations, particularly under variable environmental conditions. The MF-T2FLC showcases remarkable performance through improvements in settling times and steady-state errors, with zero overshoot, outperforming conventional controllers. Moreover, the real-time energy monitoring system is instrumental in enhancing SG performance, contributing to power delivery optimization and resource utilization.
Aromatic heterocycles are the basis for numerous agrochemical and medicinal chemicals, as well as a variety of molecules found in life. In this study, cyanoacetyl hydrazone compounds were synthesized, characterized, and tested for their in vitro anti-stress, anti-obesity, and antimitotic properties. Thin layer chromatography method was used to verify the produced compound’s purity. IR, 13C and 1H NMR spectrum analyses were used to describe the produced substance. Stress is a feedback survival response that improves a person’s physical and mental health. An organism’s energy needs rise under stressful conditions, which leads to an increase in the production of free radicals. Oxidative stress is caused by the production of these free radicals. In the biomedical domain, malondialdehyde (MDA) is frequently employed as a biomarker for evaluating oxidative stress. In both in vitro and in vivo investigations, biomonitoring of MDA has been employed as a crucial biomarker for a number of disease patterns, including cancer, diabetes, atherosclerosis, heart failure and hypertension. The primary lipolytic enzyme produced and secreted by the pancreas, pancreatic lipase or triglycerolacyl hydrolase, is crucial for the effective digestion of triglycerides. Between 50 and 70 percent of all dietary lipids are hydrolyzed by pancreatic lipase. Long chain saturated and polyunsaturated fatty acids and β-monoglycerides are the lipolytic products that result from the removal of fatty acids from the α- addition of dietary triglycerides. An amazing possibility is presented by the bioactive heterocyclic molecules. One of the most extensively researched methods for assessing the possible effectiveness of anti-obesity medications is the inhibition of digestive enzymes. Chemical interactions in the human body can and do include most pharmaceutical drugs that imitate heterocyclic molecules. Allium cepa root meristematic cells, which have been widely employed in the screening of medications with antimitotic activity were used to screen antimitotic activity. The growth and mitotic activity of Allium cepa root meristems were assessed in order to compare the inhibitory effect of the produced compounds with that of the common anticancer medication methotrexate.
Medical image analysis is essential in modern healthcare, as it allows for the extraction of clinically valuable information from various imaging modalities such as X-rays, magnetic resonance imaging (MRI), computed tomography (CT), ultrasound, and positron emission tomography (PET). Each modality provides unique diagnostic insights for disease detection, monitoring, and treatment planning. Various Artificial intelligence approaches have been implemented for precise medical image analysis, yet they face key challenges, including overfitting, poor generalization to unseen data, and high computational demands. To address these limitations, this study proposes a novel deep learning framework that integrates Hybrid Henry Gas Solubility Optimization with Adaptive Boosting Stacked Gated Recurrent Unit-Recurrent Neural Network to improve medical image classification across multiple modalities. The Stacked Gated Recurrent Unit (GRU) Recurrent Neural Network is developed to improve the capacity of the model to capture complex features and patterns within medical images from different imaging modalities. By stacking multiple layers of Gated Recurrent Units, the network captures both low-level and high-level spatial-temporal dependencies within the image data. The proposed study introduces a Hybrid Henry Gas Solubility Optimization algorithm for optimization that incorporates multiple metaheuristic strategies, including the Sooty Tern Optimization Algorithm, Jaya, Owl Search Algorithm, and Butterfly Optimization Algorithm within a dynamic reward-penalty framework to improve global search capability and avoid local minima. Further, the Adaptive Boosting strategy is implemented to strengthen model robustness by aggregating weak learners and focusing on difficult-to-classify samples, mitigating class imbalance issues. The proposed framework is evaluated using medical image datasets from three imaging modalities, such as brain MRI, chest X-ray, and lung CT scans. The simulation results confirm that the proposed mechanism attains superior diagnostic performance across multiple imaging modalities and consistently outperforms existing methods in detecting critical diseases such as brain tumors, pneumonia, and lung cancer.
