Sanjiban Sekhar Roy

Ishrat Jahan, Muhammad E. H Chowdhury, Murugappan Murugappan et al.

Glaucoma is a progressive eye disease that can lead to irreversible vision loss if not detected at an early stage. Conventional diagnostic procedures are often time-consuming and rely heavily on expert interpretation, limiting their scalability for large-scale screening. In this study, glaucoma detection is investigated under two evaluation settings: sample-wise, where individual samples are analyzed independently, and patient-wise, where data from each patient are aggregated for final prediction. An automated multimodal framework is proposed that integrates fundus images with clinical data. Under the sample-wise setting, detection is performed using fundus images and clinical features individually, as well as through their multimodal combination. Under the patient-wise setting, predictions are obtained by aggregating multiple fundus image representations with corresponding clinical information for each patient. Deep visual features are extracted using a Vision Transformer (ViT) architecture and classified using classical machine-learning models, with a stacking-based ensemble of the three best-performing classifiers employed to optimize performance. Experiments conducted on the publicly available PAPILA dataset demonstrate strong diagnostic performance, achieving 97.47% accuracy and a 97.50% F1-score for sample-wise multimodal classification, and 98.97% accuracy and F1-score for subject-wise detection. The proposed framework is further deployed as an end-to-end web-based platform to support automated glaucoma screening and clinical decision support.

Md. Sohanur Rahman, Muhammad E. H. Chowdhury, Hasib Ryan Rahman et al.

In this study, we propose a novel and robust framework, Self-DenseMobileNet, designed to enhance the classification of nodules and non-nodules in chest radiographs (CXRs). Our approach integrates advanced image standardization and enhancement techniques to optimize the input quality, thereby improving classification accuracy. To enhance predictive accuracy and leverage the strengths of multiple models, the prediction probabilities from Self-DenseMobileNet were transformed into tabular data and used to train eight classical machine learning (ML) models; the top three performers were then combined via a stacking algorithm, creating a robust meta-classifier that integrates their collective insights for superior classification performance. To enhance the interpretability of our results, we employed class activation mapping (CAM) to visualize the decision-making process of the best-performing model. Our proposed framework demonstrated remarkable performance on internal validation data, achieving an accuracy of 99.28\% using a Meta-Random Forest Classifier. When tested on an external dataset, the framework maintained strong generalizability with an accuracy of 89.40\%. These results highlight a significant improvement in the classification of CXRs with lung nodules.

Sarvesh SS Rawat, Dheeraj Dilip Mor, Anugrah Kumar et al.

High-speed, accuracy, meticulousness and quick response are notion of the vital necessities for modern digital world. An efficient electronic circuit unswervingly affects the maneuver of the whole system. Different tools are required to unravel different types of engineering tribulations. Improving the efficiency, accuracy and low power consumption in an electronic circuit is always been a bottle neck problem. So the need of circuit miniaturization is always there. It saves a lot of time and power that is wasted in switching of gates, the wiring-crises is reduced, cross-sectional area of chip is reduced, the number of transistors that can implemented in chip is multiplied many folds. Therefore to trounce with this problem we have proposed an Artificial intelligence (AI) based approach that make use of Rough Set Theory for its implementation. Theory of rough set has been proposed by Z Pawlak in the year 1982. Rough set theory is a new mathematical tool which deals with uncertainty and vagueness. Decisions can be generated using rough set theory by reducing the unwanted and superfluous data. We have condensed the number of gates without upsetting the productivity of the given circuit. This paper proposes an approach with the help of rough set theory which basically lessens the number of gates in the circuit, based on decision rules.

Anugrah Kumar, Sanjiban Shekar Roy, Sarvesh SS Rawat et al.

Phishing is a common online weapon, used against users, by Phishers for acquiring a confidential information through deception. Since the inception of internet, nearly everything, ranging from money transaction to sharing information, is done online in most parts of the world. This has also given rise to malicious activities such as Phishing. Detecting Phishing is an intricate process due to complexity, ambiguity and copious amount of possibilities of factors responsible for phishing . Rough sets can be a powerful tool, when working on such kind of Applications containing vague or imprecise data. This paper proposes an approach towards Phishing Detection Using Rough Set Theory. The Thirteen basic factors, directly responsible towards Phishing, are grouped into four Strata. Reliability Factor is determined on the basis of the outcome of these strata, using Rough Set Theory . Reliability Factor determines the possibility of a suspected site to be Valid or Fake. Using Rough set Theory most and the least influential factors towards Phishing are also determined.