ARCHANA P

@aitckm.edu.in

Assistant Professor, Department of Computer Science and Engineering
Adichunchanagiri Institute of Technology

ARCHANA P


                     Download Compact PDF  
https://researchid.co/archanap

RESEARCH, TEACHING, or OTHER INTERESTS

Computer Science, Engineering, Computer Engineering
14

Scopus Publications

Scopus Publications

  • A Hybrid Deep Learning Heart Disease Prediction Framework Utilizing Multi-Modal Medical Imaging and Novel Feature Fusion Techniques
    P. Archana, S. V. Shashikala
    Engineering Technology and Applied Science Research, 2025
    Heart diseases require advanced diagnostic techniques for early and accurate detection. This paper combines multi-modal data sources, such as CT images, MRI scans, and ECG signals to provide a hybrid deep learning architecture for accurate cardiac disease identification. The system uses specific feature extraction methods, such as 3D-UNet for 3D MRI and CT images and Temporal Convolutional Graph Neural Networks (TC-GNN) for ECG, and then uses genetic algorithms to optimize the features. Autoencoders, which are 1D for ECG and 3D for MRI and CT, are employed for non-linear dimensionality reduction in order to handle the high dimensionality of fused information. A Convolutional Neural Network (CNN) processes the fused compact features for the final classification. The proposed model achieved a 97.1% accuracy, outperforming known models. Accuracy, recall, F1-score, and ROC-AUC scores support its generalizability and robustness. This multi-modal and feature-aware approach significantly increases classification accuracy, reduces false positives and false negatives, and provides a scalable clinical decision support solution for cardiovascular diagnostics.
  • Improving the Efficiency of Predicting the Heart Diseases Using Optimized Feature Selection and Ensemble Machine Learning Techniques
    Archana P., Shashikala S. V.
    International Journal of Online and Biomedical Engineering, 2025
    Millions of people worldwide suffer from heart failure, a chronic illness that makes an effective machine learning (ML)-based approach for early detection and treatment necessary. Although medication is still the mainstay of care, exercise is becoming recognized as a useful adjunctive therapy for the management of heart failure. In this work, we used patient health parameter data to design a ML-based method to enhance heart failure detection. Improving the early detection of heart failure is our goal in an effort to save lives. To find the most important features for enhancing performance, we conducted a comparative analysis of ten distinct ML algorithms and applied feature engineering methodologies. By developing a novel new feature set, we improved our strategy and obtained the best accuracy ratings. The proposed system works on the statistical dataset and CT scan images. Numerous experiments were carried out to assess the efficacy of different algorithms, and our suggested approach outperformed other cutting-edge models, attaining impressive accuracy. Cross-validation approaches were employed to validate all applied procedures. On the CT scan dataset, AdaBoost (AB) achieved 100% accuracy, while gradient boosting (GB) led with 96% on the statistical dataset. Accuracy improved with random or synthetic data. Notably, applying a soft voting ensemble of all models further boosted accuracies to 98% and 95% on the respective datasets. Our study advances heart failure early detection techniques, which make important scientific contributions to the medical world.
  • Prediction of Stock Market Trends Using Sentiment Analysis with LSTM
    C N Arpitha, Disha Vijaykumar, J K Navya Prakash, R Shree Raksha, T M Kavya, P Archana
    Proceedings of 3rd IEEE International Conference on Knowledge Engineering and Communication Systems Ickecs 2025, 2025
    Stock market prices are influenced not only by digital data, such as economic benefits and indicators, but also by the sense of investors, analysts and the public. This study examines the use of the senses to predict the trend of the stock market, analyzing the textual data of financial news, social networks platforms and other online sources. The system uses natural language process techniques for pre-processing and analyzing unstructured text data and assigning a sentiment score based on the overall positive, negative, or neutral tone of each source. These sentiment scores are then combined with historical stock market data to create a comprehensive dataset. Machine learning models such as Long Short Term Memory, Support Vector Machine and Random Forest are used to identify patterns between sentiment changes and stock price movements. A predictive model is evaluated on its ability to predict short-term market trends.
  • Feature-Based Melanoma Detection and Classification Using Support Vector Machines
    Bhavana D E, Chaman K B, Pradhan S Gowda, Yashwanth Gowda N R, Bhanuprakash P L, Archana P
    2nd International Conference on Intelligent Algorithms for Computational Intelligence Systems Iacis 2025, 2025
    Melanoma, a highly lethal skin cancer, is challenging to diagnose due to its visual similarity to benign lesions, often delaying detection in resource-limited settings. This study proposes a Support Vector Machine (SVM)-based framework for classifying dermoscopic images as benign or malignant, utilizing a pipeline of preprocessing, Otsu's thresholding segmentation, feature extraction with Sobel edge detection and Gray-Level Co-occurrence Matrix (GLCM), and SVM classification with a Radial Basis Function (RBF) kernel. A Kaggle dataset of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{1 0, 0 0 0}$</tex> images was used, achieving a testing accuracy of 91.1 % for binary classification. The framework also supports subtype classification for malignant cases into Superficial Spreading Melanoma, Nodular Melanoma, Lentigo Maligna Melanoma, and Acral Lentiginous Melanoma, though evaluation focuses on binary classification. This lightweight, non-invasive approach enhances early melanoma detection, offering potential for clinical and telemedicine applications in resource-constrained environments.
  • Scaling CodeSourcerer: Cloud-Native Test Generation with Kubernetes and Helm
    Manikanta Prasad J, Puneeth Yogeesha, Archana P, Anser Pasha C A, Keerthishree B T, Shruthi G K
    2025 2nd International Conference on New Frontiers in Communication Automation Management and Security Iccams 2025, 2025
    The increasing complexity of software systems demands efficient and scalable solutions for automated testing. CodeSourcerer, an AI-driven test generation system, addresses this need by integrating seamlessly into GitHub CI/CD pipelines. Unlike traditional approaches that focus solely on test quality, this paper emphasizes the deployment and operational scaling of CodeSourcerer. We detail its architecture, highlighting containerized deployment models, webhook-based GitHub integration, and retry-driven resiliency mechanisms. Our system ensures that generated tests are not only accurate but also efficiently produced and delivered within CI/CD workflows. We further explore strategies for concurrent handling of multiple pull requests, dynamic resource allocation, and system robustness under varying workloads. Experimental evaluation demonstrates that CodeSourcerer maintains high reliability and fast turnaround times even under load, making it a practical and scalable choice for modern DevOps pipelines. This paper provides critical insights into designing and deploying AI-based automation tools at scale in software development environments.
  • Enhancing Diagnosing Lung Infections Using X-ray Imaging through machine learning optimized with Genetic Algorithm.
    P Archana, D.R Sangeetha, B.N Srusti, S Harshitha, D.P Palguna, T.M Kavya
    Proceedings of 3rd IEEE International Conference on Knowledge Engineering and Communication Systems Ickecs 2025, 2025
    Convolutional Neural Networks (CNNs) play a crucial role in deep learning for medical imaging, offering highly accurate detection of various lung diseases. This study strives to diagnose pulmonary conditions by identifying four preschool pulmonary conditions using chest X-ray visual stimuli, pinpoint lung cancer, tuberculosis, COVID-19, and pneumonia. A CNN-based model is developed to automatically extract features and classify these diseases, providing a fast, reliable, and non-invasive diagnostic tool. The goal is to enhance early detection and diagnosis, facilitating timely medical intervention and alleviating the strain on healthcare systems. To achieve superior classification accuracy, the CNN model is trained using an enormous set of chest X-ray depictions. Additionally, the scalability of CNNs allows them to improve as they process more data, making them highly effective for real-world applications. notwithstanding, the ability of the model to succeed is very relying on the calibre of the initial data set, and it persists and is laborious to decipher how it utilises recommendations. Despite these limitations, CNNs hold significant promise in advancing medical diagnostics and strengthening healthcare support systems.
  • Exploring Diverse AI Architectures for Accurate Breast Tumor Classification Using the WDBC Dataset
    Shruthi G K, Sahana Ganesh Siddoji, Akshata. Chitme, Arpitha C, Thanushree NB, Archana P
    International Conference on Communication Computer and Information Technology Ic3it 2025, 2025
  • Enhanced Machine Learning Framework For Early Sepsis Prediction
    T M Kavya, K N Anushree, H R Fizahath, S N Neha, Zoya Fathima, P Archana
    Proceedings of 3rd IEEE International Conference on Knowledge Engineering and Communication Systems Ickecs 2025, 2025
    Sepsis is a serious illness that can be quite dangerous, especially in intensive care unit (ICU) settings Reducing sepsis-related death rates requires prompt detection and care. Despite the truth that there might a lot of advancement in merous AI models, such as convolutional neural networks (CNNs), long-term short-term systems for clinical decision support that depend on power byAI for sepsis prediction There seems to be no flawless model. This article presents a sophisticated anticipating technique that incorporates machine learning with a Multitask Gaussian ProcessTemporal Convolutional Network (MGP-TCN). With a section beneath the precision-recall curve of 0.965 (0.710) and an area underneath the operating characteristic contour of the receiver of $0.994(0.924)$, the suggested model performed better than the state-of-the-art.
  • Enhancing Radio Signal Classification with Convolutional and Recurrent Neural Networks Using Deep Learning
    Archana P, A S Lakshmi Gowda, A Shakthi, Hruthik P, Praveen A K, Arpitha C N
    International Conference on Emerging Technologies in Electronics and Green Energy Iceteg 2025, 2025
    In this paper, Automatic Modulation Classification (AMC) represents a task which is fundamental enough for modern wireless communication because it enables management of the dynamic spectrum and because it monitors the real-time spectrum within noisy environments. Customary AMC methods often fail beneath low Signal-to-Noise Ratio (SNR) conditions. This failure happens since those methods rely on manual feature engineering. We propose a hybrid deep learning framework incorporating Convolutional Neural Networks (CNN) in order to achieve spatial feature extraction together with Recurrent Neural Networks (RNN) for temporal modeling. For denoising and low-SNR robustness, this framework uses advanced components like Transformers and Residual Networks (ResNet). The proposed model attains 95.9% accuracy at high SNR and 72.6% at low SNR classifying 24 modulation schemes while greatly exceeding state-of-the-art baselines. At 18 dB SNR, accuracy exceeds 90 % notably for BPSK coupled with OQPSK signals. These results go to show that CNN, RNN, ResNet, and Transformer modules can be combined in an effective way for strong AMC, so the model can be suitable for spectrum monitoring applications done in real time.
  • Neural Network-Based Prognostics for Dental Caries -A Deep Learning Paradigm
    P Archana, P C Suchi, A M Aishwarya, Savita Hagalambi, B S Nisarga, C N Arpitha
    Proceedings of 3rd IEEE International Conference on Knowledge Engineering and Communication Systems Ickecs 2025, 2025
    One of the most ubiquitous oral disorders impacting residents internationally is dental caries, which makes the development of comprehensive diagnostic techniques necessary. This paper presents an innovative technique to accomplish effective dental caries confirmation of reconnaissance that uses a deep multilayer neural network structure with an assortment of inputs that is score-based. Our model aspires for enhanced caries detection accuracy and aid medical professionals make well-informed treatment decisions via using a variety of input data, including as radiographic images and clinical characteristics. A substantial data set was used to validate the recommended technique, and it fared better than typical approaches. As demonstrated by the results, blending plenty input modalities may substantially improve dental practice’s evaluations. It additionally opens the door for potential advances in caries prevention and therapy besides to accomplish an invaluable contribution to the arena of dental diagnostics.
  • A Deep Learning Method for Identifying and Categorizing Fish Diseases in Aquaculture Systems
    Kavya T M, Hitha B M, Radhika S L, Medhavi L M, Neha N, Archana P
    2025 IEEE 4th International Conference for Advancement in Technology Iconat 2025, 2025
  • An Efficient Prediction System for Heart Diseases Using Machine Learning Algorithms
    Archana P, Shashikala S V
    2nd IEEE International Conference on Advances in Information Technology Icait 2024 Proceedings, 2024
  • Detection of Thyroid Diseases Using Deep Learning
    P Archana, Suhana Anjum S, Keerthana S M, Srusti P D, Suhana Anjum S
    2nd IEEE International Conference on Advances in Information Technology Icait 2024 Proceedings, 2024
  • Fetal Health Prediction
    Archana P, Tanushree D, Vamshika M J, Varsha B C, Varshini G
    2nd IEEE International Conference on Advances in Information Technology Icait 2024 Proceedings, 2024