Konguvel Elango
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Konguvel Elango has a total of 26 co-authors. Here are the top co-authors by number of shared publications:
- Goldwyn Sudhakar Jebaraj
Vellore Institute of Technology
3 shared publications M. Kannan
Anna University,Department of ElectronicsEngineering, MIT Campus,Chennai
3 shared publicationsKannan Muniandi
Anna University
2 shared publicationsAbhey S. M. Aryaan
Vellore Institute of Technology,School of Electronics Engineering (SENSE),Vellore,India
1 shared publicationAditi Sinha
School of Electronics Engineering, Vellore Institute of Technology, Vellore, India
1 shared publicationAyushi Yadav
Vellore Institute of Technology,School of Electronics Engineering (SENSE),Vellore,India
1 shared publication
Latest Publications
- Open Access
- 11 August 2026
- Published by Springer Nature in Scientific Reports
- Vol. 16 (1)
Composite EEG biomarker modeling and energy–accuracy trade-off analysis for multi-patient seizure detection
Reliable cross-patient seizure detection remains challenging because EEG characteristics vary considerably between patients. To address this problem, we proposed a compact composite feature termed the Seizure Intensity Index (SII) together with an extended representation incorporating additional theta and alpha band descriptors. The proposed feature representations were evaluated using Logistic Regression, Artificial Neural Network (ANN), and Spiking Neural Network (SNN) on the CHB-MIT and TUH EEG Epilepsy Corpus datasets. Spectral parameterization analysis showed that the proposed features primarily reflected localized oscillatory activity rather than global spectral shifts. The ANN achieved accuracies of 90.40% under random-split validation and 56.58% under CHB-MIT Leave-One-Patient-Out (LOPO) validation, indicating reduced cross-patient generalization. Logistic Regression achieved CHB-MIT LOPO accuracies of 72.83% using the baseline SII representation and 74.50% using the extended representation, while the SNN improved from 64.34% to 68.01%. During external validation on the TUH EEG Epilepsy Corpus, the SNN achieved the highest accuracy under the extended feature representation (90.78%). Although the extended representation improved performance, substantial cross-patient variability remained.
- 11 August 2026
- Published by Springer Nature in Scientific Reports
- Vol. 16 (1)
Open AccessComposite EEG biomarker modeling and energy–accuracy trade-off analysis for multi-patient seizure detection
Reliable cross-patient seizure detection remains challenging because EEG characteristics vary considerably between patients. To address this problem, we proposed a compact composite feature termed the Seizure Intensity Index (SII) together with an extended representation incorporating additional theta and alpha band descriptors. The proposed feature representations were evaluated using Logistic Regression, Artificial Neural Network (ANN), and Spiking Neural Network (SNN) on the CHB-MIT and TUH EEG Epilepsy Corpus datasets. Spectral parameterization analysis showed that the proposed features primarily reflected localized oscillatory activity rather than global spectral shifts. The ANN achieved accuracies of 90.40% under random-split validation and 56.58% under CHB-MIT Leave-One-Patient-Out (LOPO) validation, indicating reduced cross-patient generalization. Logistic Regression achieved CHB-MIT LOPO accuracies of 72.83% using the baseline SII representation and 74.50% using the extended representation, while the SNN improved from 64.34% to 68.01%. During external validation on the TUH EEG Epilepsy Corpus, the SNN achieved the highest accuracy under the extended feature representation (90.78%). Although the extended representation improved performance, substantial cross-patient variability remained.
- 9 February 2026
- Published by Elsevier in Renewable Energy Focus
- Vol. 57
A fast-recovery buck converter with differential current control and ripple suppression for energy constrained applications
- 9 February 2026
- Published by Elsevier in Renewable Energy Focus
- Vol. 57
A fast-recovery buck converter with differential current control and ripple suppression for energy constrained applications
- Open Access
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 57674-57683
Chaos-Modulated Spiking Neural Networks for Seizure Detection: A Dynamical Systems Approach to Clinical EEG
Automated seizure detection from EEG remains challenging due to signal complexity and the need for patient-independent generalization. While deep learning offers high accuracy, it lacks interpretability regarding neural dynamics. Spiking Neural Networks (SNNs) provide a biomimetic alternative but often struggle with limited dynamical states. We propose a Chaos-Modulated Spiking Neural Network (Chaos-SNN) operating at the ‘‘edge of chaos.’’ By incorporating controlled chaotic modulation into a recurrent spiking reservoir, we increase dynamical sensitivity while maintaining bounded stability and energy efficiency. The model was evaluated on controlled tasks and the clinical CHB-MIT dataset using leave-one-patient-out (LOPO) cross-validation. Our findings indicate that chaotic modulation reorganizes spike-train structures, creating interpretable distinctions between ictal and interictal patterns. While chaos did not significantly increase absolute accuracy in clinical settings, it substantially reduced cross-patient variability and stabilized internal dynamics. We conclude that chaos serves as a dynamical prior that balances sensitivity and stability, offering a robust, energy-efficient neuromorphic solution for clinical EEG monitoring.
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 57674-57683
Open AccessChaos-Modulated Spiking Neural Networks for Seizure Detection: A Dynamical Systems Approach to Clinical EEG
Automated seizure detection from EEG remains challenging due to signal complexity and the need for patient-independent generalization. While deep learning offers high accuracy, it lacks interpretability regarding neural dynamics. Spiking Neural Networks (SNNs) provide a biomimetic alternative but often struggle with limited dynamical states. We propose a Chaos-Modulated Spiking Neural Network (Chaos-SNN) operating at the ‘‘edge of chaos.’’ By incorporating controlled chaotic modulation into a recurrent spiking reservoir, we increase dynamical sensitivity while maintaining bounded stability and energy efficiency. The model was evaluated on controlled tasks and the clinical CHB-MIT dataset using leave-one-patient-out (LOPO) cross-validation. Our findings indicate that chaotic modulation reorganizes spike-train structures, creating interpretable distinctions between ictal and interictal patterns. While chaos did not significantly increase absolute accuracy in clinical settings, it substantially reduced cross-patient variability and stabilized internal dynamics. We conclude that chaos serves as a dynamical prior that balances sensitivity and stability, offering a robust, energy-efficient neuromorphic solution for clinical EEG monitoring.
- Open Access
- 23 November 2025
- Published by Springer Nature in Journal of Engineering and Applied Science
- Vol. 72 (1)
EEG seizure classification with temporal spiking neural networks and mutual information-based feature selection
Epilepsy affects over 50 million individuals worldwide, necessitating accurate and energy-efficient seizure detection systems. While Electroencephalography (EEG) is the clinical gold standard, automated classification faces challenges from high-dimensional, noisy, and non-stationary data. Traditional deep learning solutions often possess excessive computational demands, limiting their deployment in real-time, resource-constrained devices. This work proposes TemporalSNN, a lightweight spiking neural network designed for robust EEG seizure classification, optimized for neuromorphic deployment. The methodology utilizes a biologically inspired pipeline involving comprehensive feature extraction and Mutual Information (MI)-based feature selection to reduce input dimensionality. TemporalSNN employs Leaky Integrate-and-Fire dynamics to capture temporal dependencies, with performance evaluated on the Bonn, HAUZ, and Panwar benchmark datasets. Results demonstrate that MI-based feature selection significantly enhances model robustness and accuracy. On the challenging HAUZ dataset, binary accuracy improved dramatically from $$77.2\%$$ to $$92.5\%$$ (at 75% feature reduction), while on the Bonn dataset, binary accuracy increased from $$97.1\%$$ to $$98.2\%$$ with a substantial decrease in performance variance. These findings confirm that integrating MI-driven feature reduction with spiking computation yields efficient, low-complexity, and reliable seizure detection models suitable for practical, real-time clinical applications.- 23 November 2025
- Published by Springer Nature in Journal of Engineering and Applied Science
- Vol. 72 (1)
Open AccessEEG seizure classification with temporal spiking neural networks and mutual information-based feature selection
Epilepsy affects over 50 million individuals worldwide, necessitating accurate and energy-efficient seizure detection systems. While Electroencephalography (EEG) is the clinical gold standard, automated classification faces challenges from high-dimensional, noisy, and non-stationary data. Traditional deep learning solutions often possess excessive computational demands, limiting their deployment in real-time, resource-constrained devices. This work proposes TemporalSNN, a lightweight spiking neural network designed for robust EEG seizure classification, optimized for neuromorphic deployment. The methodology utilizes a biologically inspired pipeline involving comprehensive feature extraction and Mutual Information (MI)-based feature selection to reduce input dimensionality. TemporalSNN employs Leaky Integrate-and-Fire dynamics to capture temporal dependencies, with performance evaluated on the Bonn, HAUZ, and Panwar benchmark datasets. Results demonstrate that MI-based feature selection significantly enhances model robustness and accuracy. On the challenging HAUZ dataset, binary accuracy improved dramatically from $$77.2\%$$ to $$92.5\%$$ (at 75% feature reduction), while on the Bonn dataset, binary accuracy increased from $$97.1\%$$ to $$98.2\%$$ with a substantial decrease in performance variance. These findings confirm that integrating MI-driven feature reduction with spiking computation yields efficient, low-complexity, and reliable seizure detection models suitable for practical, real-time clinical applications. - Open Access
- 3 October 2025
- Published by Springer Nature in Scientific Reports
- Vol. 15 (1) , 1-27
Optimized NCFET-based approximate multiplier for energy-aware image applications
Applications for multimedia at high speeds has driven the need for error-resilient circuits that trade-off with accuracy while enhancing performance. However, the challenges posed by CMOS scaling are significant, and they have steered VLSI engineers towards exploring beyond-CMOS technology, Negative Capacitance FETs (NCFETs) with a sense of urgency. Enhancing the performance of multipliers is essential, as they are key components in various applications, including image processing and advanced learning algorithms. With the benefits of approximate computing for reducing the energy consumption of applications where exactness is not a requirement while simplifying the design of the circuits. The work here develops highly efficient approximate 4:2 compressors and multiplier circuits, representing core elements in approximate computing systems. This study introduces an innovative 8*8 approximate multiplier based on a 45nm Verilog-A NCFET technology model. It features a cutting-edge 4:2 compressor engineered for exceptional energy efficiency. By operating at a remarkably low voltage of 0.5V, the multiplier significantly reduces power consumption and energy usage compared to existing architectures. Our simulation results are compelling: the 4:2 compressor consumes just 0.122aJ of energy with NCFET technology, far outperforming alternatives like 31.67aJ, 10.49aJ, 172.75aJ in 45nm CMOS technology at 1V, which the proposed work outperforms them despite operating at half the voltage. Moreover, we have successfully realised an 8*8 approximate multiplier incorporating these compressors for image multiplication using MATLAB. The proposed design shows substantial improvements in PSNR and MSSIM and surpasses leading approximate accuracy and energy efficiency multipliers.
- 3 October 2025
- Published by Springer Nature in Scientific Reports
- Vol. 15 (1) , 1-27
Open AccessOptimized NCFET-based approximate multiplier for energy-aware image applications
Applications for multimedia at high speeds has driven the need for error-resilient circuits that trade-off with accuracy while enhancing performance. However, the challenges posed by CMOS scaling are significant, and they have steered VLSI engineers towards exploring beyond-CMOS technology, Negative Capacitance FETs (NCFETs) with a sense of urgency. Enhancing the performance of multipliers is essential, as they are key components in various applications, including image processing and advanced learning algorithms. With the benefits of approximate computing for reducing the energy consumption of applications where exactness is not a requirement while simplifying the design of the circuits. The work here develops highly efficient approximate 4:2 compressors and multiplier circuits, representing core elements in approximate computing systems. This study introduces an innovative 8*8 approximate multiplier based on a 45nm Verilog-A NCFET technology model. It features a cutting-edge 4:2 compressor engineered for exceptional energy efficiency. By operating at a remarkably low voltage of 0.5V, the multiplier significantly reduces power consumption and energy usage compared to existing architectures. Our simulation results are compelling: the 4:2 compressor consumes just 0.122aJ of energy with NCFET technology, far outperforming alternatives like 31.67aJ, 10.49aJ, 172.75aJ in 45nm CMOS technology at 1V, which the proposed work outperforms them despite operating at half the voltage. Moreover, we have successfully realised an 8*8 approximate multiplier incorporating these compressors for image multiplication using MATLAB. The proposed design shows substantial improvements in PSNR and MSSIM and surpasses leading approximate accuracy and energy efficiency multipliers.
- Top Cited PapersOpen Access
- 12 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 103531-103564
A Comprehensive Review of EEG-Based Seizure Detection Techniques
The automated classification and prediction of epileptic seizures using electroencephalography (EEG) remains a dynamic and critical area of interdisciplinary research, situated at the intersection of neuroscience and computational intelligence. Despite considerable progress in EEG signal acquisition, preprocessing methodologies, and the application of machine learning algorithms, significant challenges persist due to the inherent complexity and inter-subject variability observed in seizure patterns. This review aims to provide guidance for new researchers in this field, starting with the fundamental aspects of epilepsy, encompassing its diverse clinical presentations (syndromes), associated health conditions (comorbidities), and characteristic EEG manifestations. Furthermore, it provides a detailed analysis of recent advancements in automated seizure classification and prediction systems. We examine crucial preprocessing techniques, feature extraction strategies, and the application of both established statistical models and contemporary deep learning architectures, including innovative approaches such as image-based representations of EEG data and transfer learning paradigms. A comprehensive compilation of EEG datasets, spanning neonatal, pediatric, adult, and animal studies, is systematically presented, with a focus on their annotation schemes, and the underlying research objectives. Additionally, we discuss post-processing methodologies aimed at refining prediction accuracy, considerations for real-time hardware implementation to facilitate practical application, and emerging trends such as explainable artificial intelligence (XAI) to enhance the interpretability of AI-driven decisions, federated learning to enable collaborative model development across institutions, and neuromorphic computing inspired by the brain’s architecture. By discussing the pros and cons of different methods, current limitations, and ethical considerations, this review seeks to help new researchers develop better EEG-based seizure classification and prediction systems with real-world impact for patients.
- 12 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 103531-103564
Top Cited PapersOpen AccessA Comprehensive Review of EEG-Based Seizure Detection Techniques
The automated classification and prediction of epileptic seizures using electroencephalography (EEG) remains a dynamic and critical area of interdisciplinary research, situated at the intersection of neuroscience and computational intelligence. Despite considerable progress in EEG signal acquisition, preprocessing methodologies, and the application of machine learning algorithms, significant challenges persist due to the inherent complexity and inter-subject variability observed in seizure patterns. This review aims to provide guidance for new researchers in this field, starting with the fundamental aspects of epilepsy, encompassing its diverse clinical presentations (syndromes), associated health conditions (comorbidities), and characteristic EEG manifestations. Furthermore, it provides a detailed analysis of recent advancements in automated seizure classification and prediction systems. We examine crucial preprocessing techniques, feature extraction strategies, and the application of both established statistical models and contemporary deep learning architectures, including innovative approaches such as image-based representations of EEG data and transfer learning paradigms. A comprehensive compilation of EEG datasets, spanning neonatal, pediatric, adult, and animal studies, is systematically presented, with a focus on their annotation schemes, and the underlying research objectives. Additionally, we discuss post-processing methodologies aimed at refining prediction accuracy, considerations for real-time hardware implementation to facilitate practical application, and emerging trends such as explainable artificial intelligence (XAI) to enhance the interpretability of AI-driven decisions, federated learning to enable collaborative model development across institutions, and neuromorphic computing inspired by the brain’s architecture. By discussing the pros and cons of different methods, current limitations, and ethical considerations, this review seeks to help new researchers develop better EEG-based seizure classification and prediction systems with real-world impact for patients.
- Top Cited PapersOpen Access
- 16 April 2025
- Published by Springer Nature in Scientific Reports
- Vol. 15 (1) , 1-33
Federated learning with differential privacy for breast cancer diagnosis enabling secure data sharing and model integrity
In the digital age, privacy preservation is of paramount importance while processing health-related sensitive information. This paper explores the integration of Federated Learning (FL) and Differential Privacy (DP) for breast cancer detection, leveraging FL’s decentralized architecture to enable collaborative model training across healthcare organizations without exposing raw patient data. To enhance privacy, DP injects statistical noise into the updates made by the model. This mitigates adversarial attacks and prevents data leakage. The proposed work uses the Breast Cancer Wisconsin Diagnostic dataset to address critical challenges such as data heterogeneity, privacy-accuracy trade-offs, and computational overhead. From the experimental results, FL combined with DP achieves 96.1% accuracy with a privacy budget of ε = 1.9, ensuring strong privacy preservation with minimal performance trade-offs. In comparison, the traditional non-FL model achieved 96.0% accuracy, but at the cost of requiring centralized data storage, which poses significant privacy risks. These findings validate the feasibility of privacy-preserving artificial intelligence models in real-world clinical applications, effectively balancing data protection with reliable medical predictions.
- 16 April 2025
- Published by Springer Nature in Scientific Reports
- Vol. 15 (1) , 1-33
Top Cited PapersOpen AccessFederated learning with differential privacy for breast cancer diagnosis enabling secure data sharing and model integrity
In the digital age, privacy preservation is of paramount importance while processing health-related sensitive information. This paper explores the integration of Federated Learning (FL) and Differential Privacy (DP) for breast cancer detection, leveraging FL’s decentralized architecture to enable collaborative model training across healthcare organizations without exposing raw patient data. To enhance privacy, DP injects statistical noise into the updates made by the model. This mitigates adversarial attacks and prevents data leakage. The proposed work uses the Breast Cancer Wisconsin Diagnostic dataset to address critical challenges such as data heterogeneity, privacy-accuracy trade-offs, and computational overhead. From the experimental results, FL combined with DP achieves 96.1% accuracy with a privacy budget of ε = 1.9, ensuring strong privacy preservation with minimal performance trade-offs. In comparison, the traditional non-FL model achieved 96.0% accuracy, but at the cost of requiring centralized data storage, which poses significant privacy risks. These findings validate the feasibility of privacy-preserving artificial intelligence models in real-world clinical applications, effectively balancing data protection with reliable medical predictions.
- Open Access
- 13 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 68061-68108
Era of Sentinel Tech: Charting Hardware Security Landscapes Through Post-Silicon Innovation, Threat Mitigation and Future Trajectories
To meet the demanding requirements of VLSI design, including improved speed, reduced power consumption, and compact architectures, various IP cores from trusted and untrusted platforms are often integrated into a single System-on-Chip (SoC). However, this convergence poses a significant security challenge, as adversaries can exploit it to extract unauthorized information, compromise system performance, and obtain secret keys. Meanwhile, traditional CMOS features have limitations in addressing hardware vulnerabilities and security threats, so promising post-silicon technologies offer potential solutions. Beyond-CMOS technologies offer avenues to fortify hardware security through distinct physical properties and nontraditional computing paradigms. These advancements bolster authentication processes, enhance key generation mechanisms, ensure hardware integrity and fortify resilience against side-channel attacks, hardware Trojans and quantum-resistant cryptography in securing hardware systems. This article provides a detailed review of hardware security, encompassing the identification and mitigation of threats, the implementation of robust countermeasures, the utilization of innovative primitives, countermeasures, various methodologies and distinct features offered by emerging technologies to resist hardware threats. Moreover, strategies to address challenges, explore future directions, and outline plans for achieving further research outcomes have been put forth in this field.
- 13 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 68061-68108
Open AccessEra of Sentinel Tech: Charting Hardware Security Landscapes Through Post-Silicon Innovation, Threat Mitigation and Future Trajectories
To meet the demanding requirements of VLSI design, including improved speed, reduced power consumption, and compact architectures, various IP cores from trusted and untrusted platforms are often integrated into a single System-on-Chip (SoC). However, this convergence poses a significant security challenge, as adversaries can exploit it to extract unauthorized information, compromise system performance, and obtain secret keys. Meanwhile, traditional CMOS features have limitations in addressing hardware vulnerabilities and security threats, so promising post-silicon technologies offer potential solutions. Beyond-CMOS technologies offer avenues to fortify hardware security through distinct physical properties and nontraditional computing paradigms. These advancements bolster authentication processes, enhance key generation mechanisms, ensure hardware integrity and fortify resilience against side-channel attacks, hardware Trojans and quantum-resistant cryptography in securing hardware systems. This article provides a detailed review of hardware security, encompassing the identification and mitigation of threats, the implementation of robust countermeasures, the utilization of innovative primitives, countermeasures, various methodologies and distinct features offered by emerging technologies to resist hardware threats. Moreover, strategies to address challenges, explore future directions, and outline plans for achieving further research outcomes have been put forth in this field.
- Open Access
- 7 August 2023
- Published by Springer Nature in Research Square
Profile Verification and Secured Social Engineering with Machine Learning Models
Digital media as a platform is one of highest interactive yet sensitive arenas of information technology. With the rise of Online Social Networks, the idea of authentic identity is often seen on the brink of attack with the impact of cases of Social Engineering and other forms of Cyber Threats. The research here mainly aims at developing a Profile Verification Model with existing datasets from platforms such as Instagram. This includes understanding the behavioural aspects of interactions with masked identity elements such as bots and fake accounts Using the concepts of Supervised Machine Learning, its objective is to use existing algorithms such as Random Forest Classifier, K-Nearest Neighbour Classifier, and Linear SVM, using data pre-processing techniques to transform valid input training and testing formats. Finally, drawing comparative results using mathematical operations for accuracy and testing and depict the same through data visualization tools.
- 7 August 2023
- Published by Springer Nature in Research Square
Open AccessProfile Verification and Secured Social Engineering with Machine Learning Models
Digital media as a platform is one of highest interactive yet sensitive arenas of information technology. With the rise of Online Social Networks, the idea of authentic identity is often seen on the brink of attack with the impact of cases of Social Engineering and other forms of Cyber Threats. The research here mainly aims at developing a Profile Verification Model with existing datasets from platforms such as Instagram. This includes understanding the behavioural aspects of interactions with masked identity elements such as bots and fake accounts Using the concepts of Supervised Machine Learning, its objective is to use existing algorithms such as Random Forest Classifier, K-Nearest Neighbour Classifier, and Linear SVM, using data pre-processing techniques to transform valid input training and testing formats. Finally, drawing comparative results using mathematical operations for accuracy and testing and depict the same through data visualization tools.