Dr.Deepika S

Assistant Professor

deepikas@rvu.edu.in

About
Publication & Works
Research Summary
Awards & Achievements

I am Dr. Deepika S, currently working as an Assistant Professor at RV University, Bengaluru. I am a dedicated academician and passionate researcher in the field of Artificial Intelligence and Deep Learning, with over 9 years of teaching experience and 5.5 years of research experience. I hold a Ph.D. in Computer Science and Engineering from Vellore Institute of Technology (VIT), Vellore, where my doctoral research focused on advanced deep learning techniques for heart disease classification using echocardiogram videos. I completed my M.Tech in Computer Science and Engineering from Dayananda Sagar College of Engineering, affiliated with Visvesvaraya Technological University (VTU), and earned my B.E. in Computer Science and Engineering from VTU as well.

My core areas of expertise include machine learning, convolutional and recurrent neural networks, transfer learning, and medical image and video analysis, particularly using echocardiogram sequences. I have published several research papers in reputed SCIE and Scopus-indexed journals, including IEEE Access, The Scientific Temper, and The Open Biomedical Engineering Journal. Notably, I was honored with the Raman Research Award twice and the Abdul Kalam Award from Vellore Institute of Technology for publishing in IEEE-indexed journals. I have participated in and conducted numerous workshops, webinars, and seminars, and have delivered invited talks at various engineering colleges on the applications of Artificial Intelligence.

Beyond my research, I actively contribute to syllabus development, lab course design, and examination coordination, while also mentoring undergraduate and postgraduate students in AI-driven interdisciplinary projects. My approach to education is guided by creativity, curiosity, and a strong commitment to bridging academic theory with real-world impact.

My current research is directed towards expanding into the broader area of medical imaging, including modalities such as MRI, CT, and ultrasound, to detect novel diseases using intelligent, scalable, and clinically deployable AI models. I aim to develop advanced diagnostic tools that offer effective and early solutions for complex medical challenges.

Apart from academics, I have a keen interest in digital storytelling, generative AI, and design tools. I enjoy playing chess and badminton, building websites, and continuously exploring innovations at the intersection of technology and creativity. I believe education is not merely the transmission of knowledge but a transformative journey that empowers learners to think critically, innovate responsibly, and contribute meaningfully to society.

Hard Road Leads to a beautiful Destination....!!!

A Novel Approach to Heart Disease Classification Using Echocardiogram Videos with Transfer Learning Architecture and MVCNN Integration

Write-up: This publication proposes a novel transfer learning framework integrating Multi-View Convolutional Neural Networks (MVCNN) for accurate classification of heart diseases from echocardiogram videos. The approach enhances feature extraction by leveraging both spatial and temporal patterns across echocardiographic views, focusing on diagnostic regions with high precision. The study emphasizes the effectiveness of deep learning integration in improving heart disease detection while maintaining clinical interpretability. Journal: The Scientific Temper DOI: 10.58414/scientifictemper.2024.15.4.33 ISSN: 2231-6396, 0976-8653

Echocardiogram Videos Focusing on the Heart Valves Region

Write-up: This work explores heart valve classification using echocardiogram video data to detect abnormalities such as stenosis and regurgitation. The study highlights a regional focus on valve motion and morphology using advanced deep learning techniques. It plays a crucial role in early identification of valve-related heart diseases and offers a valuable reference for further research in region-specific echocardiographic diagnosis. Journal: African Journal of Biomedical Research DOI: 10.53555/ajbr.v27i3.3208 Publication Date: 2024-09-20

Enhanced Heart Disease Classification Using Dual Attention Mechanisms and 3D-Echo Fusion Algorithm in Echocardiogram Videos

Write-up: This IEEE Access article introduces a hybrid architecture combining dual attention mechanisms and a 3D-Echo fusion network for improved heart disease classification from echocardiogram videos. The model effectively processes spatial and temporal features and achieves significant accuracy in detecting myocardial infarction, arrhythmias, and valve defects. Journal: IEEE Access DOI: 10.1109/ACCESS.2024.3522996 IEEE Xplore: https://ieeexplore.ieee.org/document/10494316 Impact Factor: 3.9

Review on Machine Learning and Deep Learning-based Heart Disease Classification and Prediction

Write-up: This review consolidates the advancements in machine learning and deep learning techniques for the prediction and classification of heart disease using various medical imaging and signal datasets. It evaluates model performance, datasets, feature extraction strategies, and provides future directions in smart healthcare applications. Journal: The Open Biomedical Engineering Journal DOI: Link to abstract Indexing: Scopus Indexed

Detecting and Classifying Myocardial Infarction in Echocardiogram Frames with an Enhanced CNN Algorithm and ECV-3D Network

Write-up: This study presents a framework using enhanced CNN with ECV-3D to detect and classify myocardial infarction from echocardiogram video frames. The architecture achieves high performance by capturing both frame-level and volume-based features, providing a robust clinical decision-support tool. Journal: IEEE Access DOI: 10.1109/ACCESS.2024.3522996 IEEE Xplore: ttps://ieeexplore.ieee.org/document/10494316

Research: Automated Heart Disease Classification using Early CNN Models

Write-up: An early experimental study focusing on classifying heart diseases using traditional CNN models. This unpublished work laid the foundation for subsequent advancements in echocardiographic analysis and provided baseline results for myocardial feature identification. DOI: 10.13140/RG.2.2.31652.19847 Status: Unpublished (Available on ResearchGate)

Dr. Deepika S specializes in the application of artificial intelligence, deep learning, and biomedical imaging for heart disease classification and prediction, with a strong focus on echocardiogram video analysis. Her research integrates advanced convolutional neural networks (CNN), transfer learning architectures, attention mechanisms, and 3D medical image processing to develop highly accurate and interpretable diagnostic models.
She has contributed to both review and empirical research on heart disease detection using machine learning and deep learning techniques. Notably, her work published in IEEE Access introduces a dual attention mechanism fused with a 3D echocardiogram-based framework, significantly enhancing the classification accuracy of myocardial infarction and other heart conditions. Another key innovation includes the integration of Multi-View Convolutional Neural Networks (MVCNN) with transfer learning strategies to classify complex echocardiographic data.
Her studies also explore region-specific analysis, particularly focusing on heart valve abnormalities such as stenosis and regurgitation, offering high-impact clinical insights for early diagnosis. Her research has been published in Scopus- and SCIE-indexed journals including The Open Biomedical Engineering Journal, African Journal of Biomedical Research, and The Scientific Temper.
Dr. Deepika’s ongoing work continues to bridge the gap between AI research and real-time clinical applications, making her a significant contributor to the field of biomedical signal and image analysis in cardiovascular diagnostics
Research Interests
Dr. Deepika S’s research is centered on the application of artificial intelligence and deep learning for heart disease classification and prediction using echocardiogram videos. Her work integrates advanced architectures such as enhanced Convolutional Neural Networks (CNN), 3D-Echo fusion networks, Multi-View CNNs (MVCNN), and dual attention mechanisms to improve diagnostic accuracy in detecting conditions like myocardial infarction and heart valve abnormalities, including stenosis and regurgitation. These contributions, published in reputed SCIE and Scopus-indexed journals, have laid a strong foundation in the area of cardiovascular image analysis. Building on this, her future research is directed toward exploring a broader spectrum of medical imaging techniques to detect novel and emerging diseases. She aims to work extensively with multimodal medical data—such as CT, MRI, and ultrasound—to develop intelligent, scalable, and clinically deployable deep learning models. This expansion will not only reinforce her existing work in heart disease diagnosis but also enable the creation of advanced AI systems capable of offering effective solutions for a wide range of complex health conditions. Her long-term vision is to contribute meaningfully to the development of next-generation medical diagnostic tools that are accurate, interpretable, and supportive of early intervention in real-world clinical settings.

Raman Research Award – Received twice from Vellore Institute of Technology (VIT) for excellence in research contributions.
Abdul Kalam Award – Conferred by VIT for publishing impactful research in IEEE-indexed journals.
Delivered invited talks and seminars on Artificial Intelligence at various engineering colleges.
Participated in and conducted numerous workshops, webinars, and academic seminars related to AI and deep learning.