AI-Powered Patient Flow Optimization in Emergency Rooms
Authors:-Kumar S

Abstract-Emergency Rooms (ERs) are high-pressure environments characterized by unpredictability, time-sensitive decisions, and often overcrowding. These conditions, when not optimally managed, can lead to prolonged wait times, increased medical errors, clinician burnout, and compromised patient outcomes. As healthcare systems strive to deliver efficient, equitable, and timely emergency care, Artificial Intelligence (AI) has emerged as a transformative force. AI-powered patient flow optimization employs machine learning, predictive analytics, and intelligent decision support systems to streamline triage, resource allocation, and care coordination. This paper explores how AI is revolutionizing emergency room operations by enhancing real-time decision-making, reducing bottlenecks, forecasting demand, and personalizing patient care pathways. It also examines the integration of AI tools into clinical workflows, the ethical and infrastructural challenges of implementation, and the future of AI-driven operational excellence in emergency healthcare settings.

DOI: 10.61137/ijsret.vol.11.issue2.378

Published by: vikaspatanker

Integrating AI into Pediatric Health Management

Authors:-Varsha

Abstract-The application of Artificial Intelligence (AI) in healthcare has shown tremendous potential in various domains, yet one of its most impactful and delicate arenas is pediatric health management. Children are not merely miniature adults; their physiological, psychological, and developmental needs are distinct and require tailored approaches in clinical care. Pediatric health management is particularly complex, involving routine checkups, vaccinations, developmental monitoring, chronic disease management, and acute care—all while ensuring minimal invasiveness and maximum safety. Integrating AI into this domain promises transformative improvements in diagnosis, treatment planning, patient monitoring, early detection of developmental disorders, and personalized health interventions. This paper explores the significant role of AI in pediatric healthcare, examining current applications, challenges, ethical considerations, and future possibilities. By analyzing the technological advancements and real-world implementations of AI in pediatrics, this research underscores the importance of intelligent systems in ensuring the long-term health and well-being of children.

DOI: 10.61137/ijsret.vol.11.issue2.377

Published by: vikaspatanker

AI-Enhanced Decision Support for Radiology Technicians
Authors:-Pavan T.K

Abstract-The exponential rise in diagnostic imaging demands has outpaced the capacity of radiologists and radiology technicians worldwide, creating a bottleneck in timely and accurate diagnosis. Artificial Intelligence (AI) has emerged as a revolutionary tool in the field of radiology, particularly as a decision support system for radiology technicians. While much of the AI research in medical imaging focuses on automating radiologist tasks, the integration of AI tools into radiology technician workflows presents a valuable, underexplored frontier. This paper investigates the role of AI in assisting radiology technicians by enhancing image acquisition quality, automating repetitive tasks, supporting error detection, and optimizing workflow management. It also discusses AI’s contribution to patient safety, data annotation, training, and real-time support during imaging procedures. As AI technology evolves, radiology technicians are increasingly becoming empowered with tools that boost accuracy, improve efficiency, and reduce burnout. The paper also examines ethical, technical, and operational considerations in deploying AI systems in radiological environments, concluding with insights into the future of collaborative human-AI integration in medical imaging.

DOI: 10.61137/ijsret.vol.11.issue2.376

Published by: vikaspatanker

Interpretable AI for Intelligent Event Detection and Anomaly Classification in Healthcare Monitoring Systems
Authors:-Assistant Professor Mrs.K.S.R.Manjusha, D.Ashok Kumar, M.Harish, M.Hari Sathvik, M.Vinsy, A.Sri Sai Keerthi.

Abstract-Artificial intelligence (AI) is transforming healthcare by automating the detection and classification of events and anomalies, enhancing patient monitoring and intervention. In this context, events refer to abnormalities caused by medical conditions such as seizures or falls, while anomalies are erroneous data resulting from sensor faults or malicious attacks. AI-based event and anomaly detection (EAD) enables early identification of critical issues, reducing false alarms and improving patient outcomes. The advancement of Medical Internet of Things (MIoT) devices has further facilitated real-time data collection, AI-driven processing, and transmission, enabling remote monitoring and personalized healthcare. However, ensuring the transparency and explainability of AI systems is crucial in medical applications to foster trust and understanding among healthcare professionals. This work presents an online EAD approach utilizing a lightweight autoencoder (AE) on MIoT devices to detect abnormalities in real time. The detected abnormalities are then explained using Kernel SHAP, a technique from explainable AI (XAI), and subsequently classified as either events or anomalies using an artificial neural network (ANN). Extensive simulations conducted on the Medical Information Mart for Intensive Care (MIMIC) dataset demonstrate the robustness of the proposed approach in accurately detecting and classifying events, regardless of the proportion of anomalies present.

DOI: 10.61137/ijsret.vol.11.issue2.292

Published by: vikaspatanker

DeepSpineNet: Advanced Deep Learning for Multi-Class Spine X-Ray Condition Classification
Authors:-Assistant Professor Mrs.K.S.R.Manjusha, D.Ashok Kumar, M.Harish, M.Hari Sathvik, M.Vinsy, A.Sri Sai Keerthi.

Abstract-Addressing the complexchallenges of automated spine X-rayanalysis, our research introduces Deep Spine, a deep learning model designed for the multi-class classification of diverse spine conditions. Utilizing Convolutional Neural Networks (CNNs), Deep Spine demonstrates exceptional proficiency in identifying a range of spinal abnormalities, including Scoliosis, Osteochondrosis, Osteoporosis, Spondylolisthesis, Vertebral Compression Fractures (VCFs),Disability, Other, and Healthy cases. Trained on a Kaggle dataset, Deep Spine achieves high accuracy and robustness, ensuring reliable performance in classifying spinal conditions. The incorporation of transfer learning techniques further enhances its generalization capability, enabling the model to adapt effectively across different datasets. This approach not only strengthens its diagnostic accuracy but also highlights its potential for automated diagnosis and decision support in musculoskeletal radiology. This research contributes to the evolving intersection of artificial intelligence and medical imaging, demonstrating the transformative potential of deep learning in spine X-ray analysis. By leveraging AI-driven advancements, Deep Spine offers a promising step toward enhancing clinical outcomes, improving diagnostic precision, and revolutionizing spinal healthcare.

DOI: 10.61137/ijsret.vol.11.issue2.291

Published by: vikaspatanker

AI-Powered Ransomware Defence: Cutting-Edge Machine Learning Techniques for Threat Detection
Authors:-Assistant Professor Mrs.P.Satyavathi, M.Naga Sai Ganesh, N.V.Gowtham Kumar, V.S.V.Satya Yaswanth, G.Satya Nandini, V.Giri Sathvika.

Abstract-The increasing frequency and sophistication of ransomware attacks, there is a growing need for dynamic and effective detection and mitigation strategies. Traditional signature-based approaches often fall short in identifying new and evolving ransomware variants. This paper explores the application of machine learning techniques for ransomware detection, aiming to enhance the accuracy and adaptability of detection mechanisms. It provides a comprehensive analysis of various machine learning methods and algorithms, evaluating their effectiveness in identifying ransomware patterns. The findings offer valuable insights into the advancement of cybersecurity solutions, emphasizing resilience and proactive defense against the ever-evolving ransomware threat landscape.

DOI: 10.61137/ijsret.vol.11.issue2.290

Published by: vikaspatanker

Rapid Depression Detection Using Extreme Learning Machine: An AI-Driven Approach
Authors:-Assistant Professor Mrs.G.V.Rajeswari, Ch.Harikiran, K.L.Rishitha, K.H.Venkat Ganesh, B.Raj Kumar, V.L.Apoorva.

Abstract-Depression is one of the most prevalent psychological and mental health disorders, affecting a significant number of people worldwide. In recent years, Extreme Learning Machine (ELM) techniques have gained preference for addressing various health-related disease detection and prediction challenges. ELM is a single hidden layer feed-forward neural network (SLFN) that offers significantly faster convergence compared to traditional machine learning (ML) methods while delivering promising results. Although numerous studies have explored the application of ML models for depression detection, limited research has focused on utilizing ELM for this purpose. This study implements Extreme Learning Machine (ELM) alongside other ML techniques for depression detection, comparing their performance. The results demonstrate that ELM outperforms other methods, achieving the highest accuracy of 91.73%.

DOI: 10.61137/ijsret.vol.11.issue2.289

Published by: vikaspatanker

AI-Powered Fraud Detection: Secure Online Transaction Monitoring Using Machine Learning
Authors:-Mrs.G.Tejasri Devi, T.Sai Srinath, G.Naga Kastusi, V.Anshitha, G.Janitha Sree, G.Jashwitha

Abstract-Fraud detection remains one of the most critical challenges in financial transactions, driving on going research and the adoption of advanced technologies such as machine learning. Financial transaction fraud detection aims to explore and compare various machine learning approaches to assess their effectiveness, challenges, and potential future developments comprehensively.This paper begins by highlighting the importance of fraud detection in financial transactions, emphasizing the widespread impact of fraudulent activities on individuals, businesses, and the overall economy. While traditional fraud detection methods have been valuable, they often struggle to counter increasingly sophisticated and evolving fraudulent schemes. As a result, more advanced techniques are required to enhance detection accuracy.Machine learning-based approaches have emerged as a promising solution, enabling algorithms to analyse vast amounts of transactional data and identify patterns indicative of potential fraud. In particular, supervised learning techniques—such as logistic regression, decision trees, and support vector machines—have gained significant popularity in fraud detection due to their ability to classify transactions as legitimate or fraudulent based on historical data.

DOI: 10.61137/ijsret.vol.11.issue2.288

Published by: vikaspatanker

A Unified HAPS-LEO NTN Architecture for 6G: Enabling Hybrid RF-FSO Backhaul and Distributed Federated Learning
Authors:-Aakash Jain, Prakhar Vats, Priyanshu Singh, Shreya Tiwari. Mohammed Alim

Abstract-As communication systems evolve towards beyond-5G and 6G, the demand for high data rates, minimized latency, global connectivity, and distributed intelligence intensifies. Traditional terrestrial and backhaul networks face limitations in scalability, bandwidth, and coverage, particularly in challenging environments. This paper proposes a unified, multi-tier Non-Terrestrial Network (NTN) architecture integrating Low Earth Orbit (LEO) satellites and High Altitude Platform Stations (HAPS) to address these challenges. We explore a hybrid RF-Free-Space Optical (FSO) communication model to leverage the strengths of both technologies, enhancing backhaul efficiency and resilience against atmospheric disruptions. The architecture incorporates Contact Graph Routing (CGR) for optimized data routing in dynamic backhaul scenarios and a distributed Hierarchical Federated Learning (HFL) framework, utilizing HAPS as intermediate servers, to enable privacy-preserving, scalable machine learning across the network. This unified approach offers a versatile platform for future communication systems, supporting both high-performance backhaul and distributed intelligence. Simulated performance results, adapted from component studies, demonstrate the potential advantages of this integrated architecture in terms of latency, throughput, scalability, and learning accuracy.

DOI: 10.61137/ijsret.vol.11.issue2.287

Published by: vikaspatanker

Authors: Vanaja Kumari Degala

Abstract: The rapid evolution of smart manufacturing systems has intensified the adoption of cloud–edge computing architectures to support real-time data processing, resource sharing, and intelligent decision-making. However, the convergence of heterogeneous devices, distributed services, and cross-domain interactions introduces complex security challenges that traditional perimeter-based protection models fail to address effectively. This paper presents an adaptive security framework for cloud edge enabled smart manufacturing environments based on zero-trust principles. The proposed framework integrates identity-centric access control, continuous trust evaluation, intelligent anomaly detection, and distributed data protection mechanisms to ensure secure interactions across cloud, edge, and terminal layers. Unlike static security architectures, the proposed approach dynamically adjusts access privileges and protection policies based on contextual risk assessment. The framework enhances system resilience against unauthorized access, data leakage, and lateral movement attacks while supporting scalability and cross-domain collaboration. Conceptual analysis demonstrates that the proposed framework provides proactive and fine-grained security protection suitable for next-generation manufacturing ecosystems.

Published by: vikaspatanker