Human- AI Synergy in Oncology and Cybersecurity: Ethical Clinical Intelligence for Patient Care and Organ Trafficking Prevention R. Deepa, Shoba L. K., R. Swathy, B. Yasotha, P. Thilakavathy, Bennet Prabhu A. Organ Trafficking Prevention in the Healthcare Sector Examining the Co Creation of Cybersecurity Value, 2026 Artificial Intelligence (AI) has the potential to revolutionize oncology by improving cancer diagnosis and treatment planning through the analysis of medical images, genomics, and patient history. Despite its benefits, challenges such as misdiagnosis, clinician skill variation, treatment delays, interpretability issues, model bias, and ethical concerns prevent complete automation in oncology. To address these issues, this research proposes an Ethical AI-driven Cancer Diagnosis & Treatment system uses TCGA data, CNNs, Grad-CAM, and human-in-the-loop.y. A human-in-the-loop strategy allows oncologists to validate AI-generated insights, reinforcing trust and clinical reliability. Results confirm that CNN-based models improve diagnostic accuracy, reduce misdiagnosis, and support precision medicine while upholding ethical standards, data privacy, and patient-centered care.
Advanced image processing techniques for precision agriculture: Enhancing crop monitoring and disease detection through deep learning models R. Deepa, P Varun, L. K. Shoba, R. Swathy, B. V. Balaji Prabhu Innovations and Developments in Unmanned Aerial Vehicles, 2025 Precision agriculture uses technology to boost production and efficiency. When combined with deep learning models, innovative image processing can transform crop disease detection and monitoring. Advanced Crop Precision Disease Segmentation for Deep-Learning Classification (ACPDS-DLC) addresses these issues and improves crop precision disease segmentation for deep learning classification. This technology improves crop disease detection and segmentation using innovative image processing and deep learning. ACPDS-DLC enhances accuracy and resilience by combining multi-scale CNNs with enhanced photo pre-processing. The model can accurately detect many diseases and handle many cropped images. The precision agriculture applications for ACPDS-DLC are apparent, a few applications include automated disease diagnosis, yield prediction, and real-time crop monitoring. The deep learning models' precision, recall, and F1-score remained impressive, suggesting they could be effective in agriculture.
Smart Groundwater Recharge Management using Cloud Computing and Gradient Boosting Machines R. Swathy, Pramod Pandey, G Geethamahalakshmi, V. Dhayalan, S. Murugan Proceedings 2025 5th International Conference on Expert Clouds and Applications Icoeca 2025, 2025 This research presents a new method for managing groundwater recharge that uses gradient boosting machines (GBM) and cloud computing. In dry and semi-arid areas where water is scarce, groundwater recharge is vital for maintaining water supplies. Adaptability and real-time reactivity to changing environmental circumstances are sometimes lacking in traditional recharge management systems. To improve groundwater recharging techniques, provide a cloud-enabled system that combines GBM models with real-time data streams. The GBM framework is used to make recharge choices and anticipate groundwater levels. It is famous for its strong prediction powers. Faster model iteration and fast processing of massive data sets are made possible by the scalability improvements made possible by cloud computing infrastructure. Compared with more traditional methods, the proposed system's recharge management efficiency and flexibility are noticeably higher. The technique has proven useful in improving groundwater recharge under different environmental situations, as shown in case studies and simulations. This research addresses inefficient groundwater recharge prediction utilizing cloud computing and GBM, resulting in increased accuracy, real-time monitoring, and optimized sustainable water management practices.
Predictive Models for Recruiting Talent in Autonomous Vehicle Safety Development N. R. Shandy, R. Swathy, L. K. Shoba, R. Deepa, Indranil Debgupta, G. Manikandan AI S Role in Enhanced Automotive Safety, 2025 The rapid advancement of autonomous vehicle (AV) technology necessitates innovative approaches to recruiting talent capable of ensuring safety in AV systems. This study explores the application of advanced predictive modeling for identifying ideal candidates in autonomous vehicle safety development. Utilizing a deep learning-based natural language processing (NLP) approach, specifically BERT (Bidirectional Encoder Representations from Transformers), we analyze candidate profiles, resumes, and technical assessments to predict role suitability. The implementation of this model is achieved through TensorFlow, an open-source deep learning framework. By leveraging BERT's contextual understanding of language and TensorFlow's scalable architecture, the proposed solution evaluates candidates not only on technical proficiency but also on contextual experience and domain-specific knowledge. The results demonstrate significant improvements in recruitment efficiency and accuracy, providing a transformative approach to building high-caliber teams for AV safety.
Quantum Fields of Vision UAV-Driven Rice Growth Stage Mapping With Quantum-Inspired Algorithms R. Deepa, S. Pushpalatha, B. Yasotha, R. Swathy, K. L. Shoba, P. Thilakavathy Advancing Environmental Research Through Applied GIS and Remote Sensing, 2025 Food safety and precision farming need accurate plot-scale rice yield predictions, so we developed a method combining UAV‐derived vegetation indices (VIs) with brightness, greenness and moisture data from tasseled cap transformation (TCT). Eight nitrogen gradients of rice were used during the booting and heading stages to obtain ground truth and six-band UAV imagery. We propose a hybrid quantum learning model that uses Bi-LSTM for extracting temporal features and quantum circuits for quantum feature processing. These enhanced features are combined with Bi-LSTM outputs into an XGBoost regressor. Our Quantum BiLSTM + XGBoost approach outperformed traditional models by 7-10%, achieving ~94% accuracy.
Neural Network-Based Automated Soil Salinity Mapping and Remediation Using Wireless Sensor and Cloud Computing B. Shadaksharappa, S. Sriram, R. Swathy, Nilamadhab Mishra, R. Meenakshi, J Suganya Proceedings of the 2024 10th International Conference on Communication and Signal Processing Iccsp 2024, 2024 This paper uses neural networks, wireless sensor networks, and cloud computing to automate soil salinity mapping and treatment. Soil salinity affects agricultural output, but conventional techniques of monitoring and managing it are laborious. This system collects real-time soil salinity data from a specific agricultural region using a wireless sensor network. These sensor nodes provide data to a cloud computing platform for automatic soil salinity mapping using a neural network model. The neural network is trained on a large soil property and salinity dataset to accurately predict field salinity distribution. Farmers and land managers may make quick decisions using the cloud-based data storage, processing, and analysis solution. The suggested system incorporates remediation options to inform sustainable soil management. Through field testing, the methodology proves its accuracy and efficiency above older approaches. Precision agriculture producers may monitor soil salinity in real time and take targeted repair actions using the described method. Due to rising agricultural needs and environmental concerns, neural networks, wireless sensor networks, and cloud computing provide a solid foundation for sustainable and technology-driven soil management.
Smart Contact Lenses Integrating IoT and Cloud-based Neural Networks for Early Detection of Ocular Diseases Aaron Kevin Cameron Theoderaj, S. Vijayanand, S. Sumithra, R. Swathy, S. Srinivasan, S. Velmurugan 2024 10th International Conference on Smart Computing and Communication Icscc 2024, 2024 An innovative method for the early detection of ocular diseases is presented by the combination of cloud-based neural networks and the Internet of Things (IoT) in smart contact lenses. These lenses have sensors built right into them that track intraocular pressure, tear composition, and eye movement continuously. They send the collected data in real time to neural networks that are hosted on the cloud for analysis. The objective of this system is to improve the accuracy of early detection and offer prompt interventions for eye health. Data privacy is protected during transmission and storage by security measures like secure access controls and encryption protocols. Subsequent investigations will concentrate on enhancing sensor downsizing, energy conservation, and neural network precision, in addition to tackling moral and legal issues by means of strong privacy regulations. This system has the potential to completely change the way ocular diseases are managed by utilizing cutting-edge IoT and neural network technologies. It provides a non-invasive, effective, and precise means of ongoing monitoring of eye health and early detection to better patient outcomes and lower healthcare expenses.
Cloud-Powered Patient Interaction: Humanoid Robots in Hospital Reception using NLP N. Naveenkumar, S. Priyadarshini, R. Swathy, Nilamadhab Mishra, P. Thirumaraiselvan, S. Renukadevi Proceedings of the 2024 10th International Conference on Communication and Signal Processing Iccsp 2024, 2024 This research presents a novel method for improving patient engagement in hospital reception areas by using cloud computing and natural language processing (NLP) approaches to power humanoid robots. In order to improve the patient experience overall, the greeting procedure is being streamlined and personalized via the deployment of humanoid robots. To facilitate an effective connection between the hospital’s computer systems and the humanoid robots, cloud computing enables the processing and analysis of data in real-time. The ability to comprehend and react to patient questions, provide pertinent information, and aid with navigation inside the hospital grounds are all made possible by the robots’ natural language processing skills. Robots may access and update patient data using cloud resources, proving accurate and up-to-date information transmission. And since the system is cloud-based, it can learn and adapt on the go, so the robots may improve their communication abilities in response to human input. Results in decreasing wait times, improving information accuracy, and increasing overall patient happiness have been encouraging since the deployment of this cloud-powered patient contact system. To improve healthcare administration and patient care by integrating advanced technology into healthcare environments. Adding new features and improving the system in response to input from users (both patients and doctors) are potential directions for future growth.
Data-Driven Recycling Transformation for Enhancing Paper and Cardboard Bin Efficiency through IoT and Random Forest S. Srinivasan, R. Swathy, V. Srividhya, S. Murugan, C. Srinivasan, M. Muthulekshmi 2024 International Conference on Advances in Modern Age Technologies for Health and Engineering Science Amathe 2024, 2024 This study uses data to enhance paper and cardboard recycling bin efficiency for sustainable waste management. It uses Internet of Things (IoT) and Random Forest algorithms to dynamically optimize bin use to improve recycling. It starts by installing IoT sensors on paper and cardboard recycling bins to track fill levels and use. A Random Forest system trained on past data predicts bin fill levels from this continuous data stream. The predictive algorithm adapts collection schedules and resource distribution based on temporal trends, weather, and community events. IoT and Random Forest increase fill-level forecasts and enable data-driven recycling bin management. This reduces wasted collections, fuel use, and carbon emissions, making recycling more sustainable and cost-effective. The study also analyzes the system's real-world urban application, demonstrating its scalability and flexibility to varied waste management circumstances. Our empirical study and case studies show that the technique improves paper and cardboard recycling efficiency, contributing to data-driven sustainability programs.
Bayes Theorem Based Virtual Machine Scheduling for Optimal Energy Consumption R. Swathy, B. Vinayagasundaram Computer Systems Science and Engineering, 2022 This paper proposes an algorithm for scheduling Virtual Machines (VM) with energy saving strategies in the physical servers of cloud data centers. Energy saving strategy along with a solution for productive resource utilization for VM deployment in cloud data centers is modeled by a combination of “Virtual Machine Scheduling using Bayes Theorem” algorithm (VMSBT) and Virtual Machine Migration (VMMIG) algorithm. It is shown that the overall data center’s consumption of energy is minimized with a combination of VMSBT algorithm and Virtual Machine Migration (VMMIG) algorithm. Virtual machine migration between the active physical servers in the data center is carried out at periodical intervals as and when a physical server is identified to be under-utilized. In VM scheduling, the optimal data centers are clustered using Bayes Theorem and VMs are scheduled to appropriate data center using the selection policy that identifies the cluster with lesser energy consumption. Clustering using Bayes rule minimizes the number of server choices for the selection policy. Application of Bayes theorem in clustering has enabled the proposed VMSBT algorithm to schedule the virtual machines on to the physical server with minimal execution time. The proposed algorithm is compared with other energy aware VM allocations algorithms viz. “Ant-Colony” optimization-based (ACO) allocation scheme and “min-min” scheduling algorithm. The experimental simulation results prove that the proposed combination of ‘VMSBT’ and ‘VMMIG’ algorithm outperforms other two strategies and is highly effective in scheduling VMs with reduced energy consumption by utilizing the existing resources productively and by minimizing the number of active servers at any given point of time.
