Ph D in College of Engineering Anna University,Chennai-25
RESEARCH, TEACHING, or OTHER INTERESTS
Engineering
15
Scopus Publications
Scopus Publications
Designing an empirical grid-connected PV system based on FLC-MPPT approach for local community use S. Sathea Sree, M. Muthalakshmi, Prasath Alias Surendhar S, A. Satya Sai Kumar, Bhawani Sankar Panigrahi, et al. International Journal of Information Technology Singapore, 2025 Photovoltaic (PV) systems play a vital role in mitigating renewable energy issues ranging from the oil crisis to environmental concerns. The given paper proposes a grid-connected PV power system with high voltage gain (VG) and a high-speed multiphase buck-boost converter. With this converter, PV panels can be integrated in any fashion as per varying climatic conditions without affecting the switching stress. Also, the proposed system makes use of maximum power point tracking (MPPT) concept with fuzzy logic control (FLC) algorithm in order to reduce losses and complexities associated with the system. For validation of the system, an annual dataset related to global solar irradiance across three locations in Tamil Nadu is taken into consideration. The proposed system is validated and compared with traditional MPPT methods based on power output (W) and energy efficiency (%). It includes the computation of predicted mean (PM) values and comparing them with actual mean (AM) values of solar radiation for each of the locations. The results show a good correlation between the predicted and actual values and higher efficiency, thereby making the proposed system suitable for forecasting solar irradiance.
Revolutionizing brain tumor diagnosis with adaptive CNN models R. Kishore Kanna, Priyanka Singh, S. Raju, Ayodeji Olalekan Salau, Archibald Danquah Danquah-Amoah, et al. Integrative Machine Learning and Optimization Algorithms for Disease Prediction, 2025 Brain tumors affect millions of people worldwide and can be life-threatening if not detected early. These cancerous growths disrupt normal brain function and can spread to surrounding tissues. However, diagnosing brain tumors is challenging because they vary greatly in shape, size, and appearance, making manual detection difficult and time-consuming for doctors. Our research team developed an intelligent computer system using artificial intelligence to automatically identify and classify brain tumors from MRI scans. We tested three different AI models - ResNet-152, MobileNet, and DenseNet-121 - using a technique called transfer learning, which allows computers to build on existing medical knowledge.The results were impressive. The ResNet-152 model performed best, correctly identifying brain tumors 98.7% of the time with a 99.8% reliability score. DenseNet-121 achieved 96.5% accuracy with 98.6% reliability, while MobileNet reached 87.2% accuracy with 98.7% reliability. Even the smallest model performed well enough for practical use in hospitals with limited computing resources.
Deep Learning-Based Segmentation System for Renal Tumors Gomalavalli R, Veera Boopathy E, Hema J International Research Journal of Multidisciplinary Scope, 2025 This research set out to conduct a comprehensive analysis of methods currently used for segmenting renal tumors from CT images. Renal tumor (RT) remains maximum prevalent tumor for all globally, and it is one of the diseases that have greatly impacted our culture. In comparison to the time-consuming and labor-intensive method of conventional analysis, the automated recognition procedures of deep learning (DL) shall speed up analysis, tweak test precision, decrease expenses, besides relieve strain on radiologists. Here, detection models proposed which can be used to identify RTs in CT scans. Investigators in the area of medical imaging segmentation utilizes DL techniques for tackling difficulties in tumor delineation, cell delineation, and organ segmentation all at once. For radiation and therapeutic purposes, semantic tumors segmentation is essential. Automated recognition algorithms based on predictive modeling might speed up the diagnostic process, improve test precision, and reduce expenses contrast to lengthy, prolonged traditional methods. The hybrid V-Net method determines the renal segmentation of 0.977 and tumor segmentation of 0.865. A 300CT datasets are utilized to obtain the 91-99% of accuracy in modified CNN and 3 cross folds. Renal tumors are among the deadliest types of tumors, and previous research has demonstrated that deep learning can aid detection, segmentation, and categorization of this disease. Modern developments in DL-based segmentation systems for renal tumors are discussed in this article. Here, the components of renal tumor segmentation outlined, including the numerous medical picture types and segmentation algorithms, as well as the assessment criteria for segmentation outcomes.
Integrating Deep Learning into VLSI Technology: Challenges and Opportunities Veera Boopathy E, Sasikala C, Vigneash L, Satheesh S, Gomalavalli R, et al. International Research Journal of Multidisciplinary Scope, 2024 This paper conducts a comprehensive review and analysis of the difficulties and possibilities related to integrating deep learning algorithms into the future of VLSI design and technology. The area of integrated circuit design is becoming increasingly complex as transistors become smaller and the expectations for enhanced reliability and environmental sustainability increase. Analysts are looking into novel techniques that involve deep learning, as traditional techniques find it challenging to tackle these issues. In particular, deep neural networks possess the ability to improve various aspects of integrated circuit design, including timing assessment, layout enhancement, fault detection, and energy utilization minimization. Deep learning has become a viable solution for addressing a range of VLSI challenges, providing opportunities for automated processes, enhancement, and creativity at several phases of the development and fabrication cycle. The incorporation of deep learning into system acceleration, identifying defects, layout synthesis, and future repairs is investigated in this article. It also draws attention to the challenges and opportunities associated with incorporating neural networks into VLSI, highlighting the necessity of multidisciplinary cooperation and creativity to realize their maximum potential. By surmounting these challenges and capitalizing on the prospects presented by deep computing, the integrated circuit sector might unleash unprecedented heights of efficiency, productivity, and inventiveness in integrated circuit innovations.
Clinical Application of Neural Network for Cancer Detection Application R Kishore Kanna, R Ravindraiah, C Priya, R Gomalavalli, Nimmagadda Muralikrishna Eai Endorsed Transactions on Pervasive Health and Technology, 2024 
 INTRODUCTION: The field of medical diagnostics is currently confronted with a significant obstacle in the shape of cancer, a disease that tragically results in the loss of millions of lives each year. Ensuring the administration of appropriate treatment to cancer patients is of paramount significance for medical practitioners.
 OBJECTIVES: Hence, the accurate identification of cancer cells holds significant importance. The timely identification of a condition can facilitates prompt diagnosis and intervention. Numerous researchers have devised multiple methodologies for the early detection of cancer.
 METHODS: The accurate anticipation of cancer has consistently posed a significant and formidable undertaking for medical professionals and researchers. This article examines various neural network technologies utilised in the diagnosis of cancer.
 RESULTS: Neural networks have emerged as a prominent area of research within the medical science field, particularly in disciplines such as cardiology, radiology, and oncology, among others.
 CONCLUSION: The findings of this survey indicate that neural network technologies demonstrate a high level of efficacy in the diagnosis of cancer. A significant proportion of neural networks exhibit exceptional precision when it comes to categorizing tumours cells.
Efficient Power Management in Multicore Systems using Distributed On-Chip Switched-Capacitor Converters with DVFS Aiswarya. M, Veera Boopathy. E, Kiruba. S, Karthick. L.S, Kannadhasan. S, et al. 2nd International Conference on Emerging Research in Computational Science Icercs 2024, 2024 Efficient power management is paramount in modern multicore systems to balance performance and energy consumption. This paper explores the integration of Distributed On-Chip Switched-Capacitor Converters (DoS-DCCs) with Dynamic Voltage and Frequency Scaling (DVFS) support to address this challenge. DoS-DCCs offer a decentralized approach to power delivery, enabling localized conversion tailored to individual cores or clusters. Leveraging a switched-capacitor architecture, these converters provide benefits such as high efficiency and compatibility with standard CMOS processes. By integrating DVFS, DoS-DCCs empower cores to dynamically adjust their voltage and frequency, optimizing performance while minimizing energy usage. This paper presents the implementation of DoS-DCCs with DVFS support in multicore systems and discusses their potential to enhance power efficiency and performance scalability in modern computing environments. This paper presents a comprehensive exploration of the implementation and benefits of DoS-DCCs with DVFS support in multicore systems. Through simulation studies and experimental validation, the effectiveness of this approach is demonstrated in improving power efficiency and performance scalability. Furthermore, potential challenges and future research directions for advancing the integration of DoS-DCCs with DVFS in next-generation multicore architectures are outlined. Overall, the combination of Distributed On-Chip Switched-Capacitor Converters and Dynamic Voltage and Frequency Scaling offers a compelling solution for efficient power management in multicore systems, paving the way for enhanced performance and energy efficiency in future computing platforms.
Improving Iris Recognition Systems with Transfer Learning and Pretrained CNN Models S Lalitha, B. Padmavathy, Chaithra S, R Gomalavalli, P. V. Rajlakshmi, et al. 2024 1st International Conference on Advanced Computing and Emerging Technologies Acet 2024, 2024 Conventional iris identification systems have failed to manage environmental variables and changes in iris patterns, sometimes depending on handmade characteristics that may lack the delicate details required for reliable identification. The paper offers a revolutionary iris identification system based on deep learning, especially pre-trained Convolutional Neural Networks (CNNs) such as ResNet and VGG, which are fine-tuned utilizing transfer learning methods. The system is trained and verified using a variety of iris datasets, taking into account illumination fluctuations, occlusions, and other environmental conditions. Using these pre-trained CNN models, the proposed system intends to dramatically improve the accuracy and reliability of iris detection, even under difficult settings. The proposed system attains significant gains in recognition accuracy, even under challenging circumstances, thorough preprocessing, transfer learning, fine-tuning, and data augmentation. Testing and validation show how effective the proposed system is; it has better performance metrics than existing systems, with 98.5% accuracy, 97.8% precision, 98.9% recall, and 98.3% F1-score, which are all higher than those of existing systems. The methodology also guarantees scalability and computing efficiency, confirming the proposed system's potential for accurate and dependable iris detection in practical settings. The reliability and efficiency of iris recognition technology have been significantly improved by these advances.
Smart detection and removal of artifacts in cognitive signals using biomedical signal intelligence applications R. Kishore Kanna, K. Yamuna Devi, R. Gomalavalli, A. Ambikapathy Quantum Innovations at the Nexus of Biomedical Intelligence, 2023 A complete and detailed literature evaluation concentrating on the detection and elimination of artifacts from EEG data was described in the preceding chapter. Issue-wise solution suggestions and their limitations were also studied, which eventually led to finding the gaps in the recommended task and scope of the study activity. In this chapter, the complete explanation of system design and its implementation is addressed. The principal objective of the proposed research is to identify and eliminate the undesired signals known as artifacts from the collected EEG data. This chapter spoke about the design of the system and its implementation. In this chapter specifics of EEG acquisition methods have been discussed. The initial stage in EEG signal processing is recording EEG data from the individuals. It also looks into the categorization of EEG data by sort. The obtained EEG data was sorted into two categories: normal and epileptic.
Computational cognitive analysis for intelligent engineering using eeg applications R. Kishore Kanna, R. Gomalavalli, Yamuna Devi, A. Ambikapathy Intelligent Engineering Applications and Applied Sciences for Sustainability, 2023 NFT increases EEG's higher alpha band to improve working memory. Five sessions of visual cue feedback instructed patients. Single-channel EEGs collect EEG signals. Each participant's unique alpha frequency band calculated the Higher Alpha band. LabVIEW programme extracted the higher alpha band (10–13hz) signal. The patient was then encouraged to relax by watching the device's nature. Thus, higher alpha waves predominate Relaxation creates alpha waves. Thus, the NFT retrains the brain to make alpha waves on its own and boosts activity. Participants learned and increased alpha band amplitude. Neuro-feedback training using NFT enhanced cognitive processing speed. 60-65-year-olds were selected for this training. This research examined if training improves elderly people's cognitive processing speed. Visual input improves brain control and consistency. Brainwaves were rewarded with visual messages. The training uses Lab View. Finally, mental, physical, and emotional health improves Neuro-feedback system to assess healthy volunteers and the elderly's cognitive performance might be built.
Computing Model for Alzheimer Prediction Using Support Vector Machine Classifier R. Kishore Kanna, U. Mutheeswaran, V. Subha Ramya, R. Gomalavalli, L.K Hema, et al. Proceedings of 2022 IEEE International Conference on Current Development in Engineering and Technology Ccet 2022, 2022 The first stage of Alzheimer's disease is known as Mild Cognitive Impairment. Identification of MCI subjects who are at high risk of developing frame over time is crucial for successful therapy. In order to track the progress of numerous Alzheimer's forecasts over time, automated modelling was used in this work. Three separate longitudinal data systems are used to train models. Then, for each experimental investigation, these models are used to assess biomarker data. Finally, MCI patients at risk of acquiring in the future are identified using a typical support vector machine categorization. Both cognitive points and magnetic resonance image-based measurements are used to thoroughly assess the proposed models' prediction ability. Our suggested strategy produced the maximum AUC 88.93% (Accuracy = 84.29%) and 88.13% (Accuracy = 83.26%) in the five different split verification settings, respectively, 1 year and 2 years prior to conversion prediction of MCI data. Important findings of this research include: Clinical changes in magnetic resonance image- based therapies may be predicted more accurately than cognitive points in two ways: Multiple predictive models are more accurate in predicting change than single biomarker models and Neuropsychology programme by themselves may provide superior long-term change prediction. Enhancing the accuracy of Alzheimer occurrences prediction using SVM Classifier Modelling will be the ultimate goal of this research.
Smart Electronic Arm Module using Arduino Applications R. Kishore Kanna, L.K Hema, V. Subha Ramya, N. Kripa, R. Gomalavalli, et al. Proceedings of 2022 IEEE International Conference on Current Development in Engineering and Technology Ccet 2022, 2022
