Devaraj Somasundaram
Research Subjects
Devaraj Somasundaram has a total of 30 co-authors. Here are the top co-authors by number of shared publications:
- Nirmala Madian
Department of Biomedical Engineering, Dr. N.G.P. Institute of Technology, Coimbatore, India
3 shared publications - Gnanasaravanan Subramaniam
Department of Biomedical Engineering, Karunya Institute of Technology and Sciences, Coimbatore, India
2 shared publications S. Suresh
MediScan Systems, Chennai, India
2 shared publications- A. Kumaresan
School of Computer Science and Engineering, Vellore Institute of Technology, Vellore, Tamil Nadu, India
1 shared publication A. S. Sindhu
Department of Electronics and Communication Engineering, Rajas Institute of Technology, Nagercoil, Tamil Nadu, India
1 shared publicationAditya Chapparadahallimath
Department of Micro and Nano Electronics, School of Electronics Engineering, Vellore Institute of Technology, Vellore, Tamilnadu, India
1 shared publication
Latest Publications
- Open Access
- 12 December 2025
- Published by Asian Research Association in NanoNEXT
- Vol. 6 (4) , 48-57
Real Time Pesticide Contaminant Monitoring Using Hybrid Mxene Doped Graphene FET Biosensors in Surface and ground Water
In many parts of the world, pollution caused by chemical releases into water systems has put aquatic ecosystems and human health at risk. Storm water overflows, industrial, agricultural, and wastewater treatment facilities are some of the sources of chemical discharge. It is crucial to develop and implement quick, dependable, and sensitive analytical detection technologies in order to meet these limits. The development of electrochemical biosensors for environmental toxicants, such as pesticides and heavy metals, has advanced over the last few decades. In this paper, a novel Mxene doped graphene FET is proposed to detect the pesticide and contaminations in water. This sensor is electronically compactable and has high chemical sensitivity. Due to the lack of an efficient method for distinguishing defective devices from preselected uniform devices based solely on electronic properties, two-dimensional (2D) electronic sensors frequently experience device-to-device variations. This causes sensor inaccuracy and slows down the real-world applications of the sensors. Real-time spontaneous detection of pesticides such as atrazine and metolachlor in flowing tap water was made possible by our sensors. In order to maximise the potential of electronic sensors for tracking contaminants in flowing water, this work provides a dependable quality control procedure.
- 12 December 2025
- Published by Asian Research Association in NanoNEXT
- Vol. 6 (4) , 48-57
Open AccessReal Time Pesticide Contaminant Monitoring Using Hybrid Mxene Doped Graphene FET Biosensors in Surface and ground Water
In many parts of the world, pollution caused by chemical releases into water systems has put aquatic ecosystems and human health at risk. Storm water overflows, industrial, agricultural, and wastewater treatment facilities are some of the sources of chemical discharge. It is crucial to develop and implement quick, dependable, and sensitive analytical detection technologies in order to meet these limits. The development of electrochemical biosensors for environmental toxicants, such as pesticides and heavy metals, has advanced over the last few decades. In this paper, a novel Mxene doped graphene FET is proposed to detect the pesticide and contaminations in water. This sensor is electronically compactable and has high chemical sensitivity. Due to the lack of an efficient method for distinguishing defective devices from preselected uniform devices based solely on electronic properties, two-dimensional (2D) electronic sensors frequently experience device-to-device variations. This causes sensor inaccuracy and slows down the real-world applications of the sensors. Real-time spontaneous detection of pesticides such as atrazine and metolachlor in flowing tap water was made possible by our sensors. In order to maximise the potential of electronic sensors for tracking contaminants in flowing water, this work provides a dependable quality control procedure.
- Open Access
- 1 January 2025
- Published by Wiley in Journal of Sensors
- Vol. 2025 (1)
Network Security and Privacy Protection in Cyberattacks With Asynchronous Reinforcement Federated Learning With Task Offloading: Decentralized Real‐Time Iteration Approach
Medical healthcare has experienced a revolution through the Internet of Medical Things (IoMT) system, which needs crucial improvements for security measures while working towards better privacy and efficiency capabilities. This paper develops asynchronous reinforcement federated learning (FL) with task offloading (ARFL‐TO) as a combination of FL with reinforcement learning (RL) through dynamic TO systems to boost scalability alongside adaptability and security in heterogeneous environments, including autonomous systems, smart grids, and industrial Internet of Things (IIoT) settings. ARFL‐TO requires validation through different datasets, which encompass a range of disease groups together with multiple medical instruments. The analysis indicates that ARFL‐TO surpasses FL‐TO and federated reinforcement‐based fusion (FRF) in productivity by 42.22% and reduces power requirements by 79.22% while shortening processing time by 7.13% during real‐world operation, where networks become unstable and data transmission is affected. This framework achieves privacy protection through secure data pooling, and it makes models understandable for clinical support in addition to delivering enhanced energy efficiency for low‐power medical devices. Future investigation directs toward integrating the proposed system into real healthcare environments and developing performance improvements for severe operational scenarios, and introducing adaptable hyperparameter techniques suitable for healthcare systems requiring dynamic adjustments. ARFL‐TO establishes itself as an efficient framework for secure time‐sensitive decision‐making across decentralized medical networks and cross‐sectoral systems, maintaining privacy concerns.
- 1 January 2025
- Published by Wiley in Journal of Sensors
- Vol. 2025 (1)
Open AccessNetwork Security and Privacy Protection in Cyberattacks With Asynchronous Reinforcement Federated Learning With Task Offloading: Decentralized Real‐Time Iteration Approach
Medical healthcare has experienced a revolution through the Internet of Medical Things (IoMT) system, which needs crucial improvements for security measures while working towards better privacy and efficiency capabilities. This paper develops asynchronous reinforcement federated learning (FL) with task offloading (ARFL‐TO) as a combination of FL with reinforcement learning (RL) through dynamic TO systems to boost scalability alongside adaptability and security in heterogeneous environments, including autonomous systems, smart grids, and industrial Internet of Things (IIoT) settings. ARFL‐TO requires validation through different datasets, which encompass a range of disease groups together with multiple medical instruments. The analysis indicates that ARFL‐TO surpasses FL‐TO and federated reinforcement‐based fusion (FRF) in productivity by 42.22% and reduces power requirements by 79.22% while shortening processing time by 7.13% during real‐world operation, where networks become unstable and data transmission is affected. This framework achieves privacy protection through secure data pooling, and it makes models understandable for clinical support in addition to delivering enhanced energy efficiency for low‐power medical devices. Future investigation directs toward integrating the proposed system into real healthcare environments and developing performance improvements for severe operational scenarios, and introducing adaptable hyperparameter techniques suitable for healthcare systems requiring dynamic adjustments. ARFL‐TO establishes itself as an efficient framework for secure time‐sensitive decision‐making across decentralized medical networks and cross‐sectoral systems, maintaining privacy concerns.
- 15 October 2024
- Published by Springer Nature in Knowledge and Information Systems
- Vol. 67 (2) , 977-1011
Chromosome segmentation and classification: an updated review
- 15 October 2024
- Published by Springer Nature in Knowledge and Information Systems
- Vol. 67 (2) , 977-1011
Chromosome segmentation and classification: an updated review
- 23 May 2024
- Published by Taylor & Francis in IETE Journal of Research
- Vol. 70 (10) , 7899-7910
Design of Approximate Tree Multiplier Using Approximate Compressor for Image Processing Applications
- 23 May 2024
- Published by Taylor & Francis in IETE Journal of Research
- Vol. 70 (10) , 7899-7910
Design of Approximate Tree Multiplier Using Approximate Compressor for Image Processing Applications
- Open Access
- 18 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 41042-41055
Optimizing Large-Scale RFID Networks With Energy-Efficient Dynamic Cluster Head Selection: A Performance Improvement Approach
This paper focuses on the use of radio frequency signals for non-contact tracking and localization utilising Radio Frequency Identification (RFID) technology. The clustering approach is quite useful when dealing with large RFID readers. This method stimulates the expansion of RFID nodes without compromising the overall network performance. The Cluster Head (CH) is the most critical node in a clustered RFID system. This research presents a method for dynamically selecting the cluster head that takes the connectivity and power of each RFID reader into account. In dynamic mode, selecting a new cluster head is based on Fuzzy Logic. Based on these data, the energy level of 0.443 and centrality of 0.809 are the thresholds at which a node has a 91.8 % chance of becoming a CH. The management of massive RFID networks is also handled, with cluster numbers increasing at different time intervals. The RFID network’s effectiveness is determined by measuring throughput, accuracy, delay, success, and error rates. The network may accommodate up to a thousand nodes with 13 node leaders for improved capacity. The results show a 97.8 % success rate, 0.22 % accuracy, a 2.64 % error rate, and a 36.91 second latency.
- 18 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 41042-41055
Open AccessOptimizing Large-Scale RFID Networks With Energy-Efficient Dynamic Cluster Head Selection: A Performance Improvement Approach
This paper focuses on the use of radio frequency signals for non-contact tracking and localization utilising Radio Frequency Identification (RFID) technology. The clustering approach is quite useful when dealing with large RFID readers. This method stimulates the expansion of RFID nodes without compromising the overall network performance. The Cluster Head (CH) is the most critical node in a clustered RFID system. This research presents a method for dynamically selecting the cluster head that takes the connectivity and power of each RFID reader into account. In dynamic mode, selecting a new cluster head is based on Fuzzy Logic. Based on these data, the energy level of 0.443 and centrality of 0.809 are the thresholds at which a node has a 91.8 % chance of becoming a CH. The management of massive RFID networks is also handled, with cluster numbers increasing at different time intervals. The RFID network’s effectiveness is determined by measuring throughput, accuracy, delay, success, and error rates. The network may accommodate up to a thousand nodes with 13 node leaders for improved capacity. The results show a 97.8 % success rate, 0.22 % accuracy, a 2.64 % error rate, and a 36.91 second latency.
- Open Access
- 13 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of Nanotechnology
- Vol. 5, 1-8
Analysis of GAA Junction Less NS FET Towards Analog and RF Applications at 30 nm Regime
This research focuses on a quantum model created using an entirely novel nanosheet FET. The standard model describes the performance of a Gate-all-around (GAA) Junction-less (JL) nanosheet device with a gate dielectric of SiO 2 and HfO 2 , each having a thickness of 1 nm. The performance of both the classical and quantum models of the GAA nanosheet device is evaluated using the visual TCAD tool, which measures the ION , IOFF , ION/ IOFF , threshold voltage, DIBL, gain parameters (g m , g d , A v ), gate capacitance, and cut-off frequency ( fT ). The device is suited for applications needing rapid switching since it has a low gate capacitance of the order of 10 –18 , according to the simulation results. A transconductance (g m ) value of 21 μS and an impressive cut-off frequency of 9.03 GHz are displayed during device analysis. A detailed investigation has also been done into the P-type device response for the same device. Finally, the proposed GAA nanosheet device is used in the inverter model. The NSFET-based inverter, although having higher gate capacitance, has the shortest propagation latency.
- 13 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of Nanotechnology
- Vol. 5, 1-8
Open AccessAnalysis of GAA Junction Less NS FET Towards Analog and RF Applications at 30 nm Regime
This research focuses on a quantum model created using an entirely novel nanosheet FET. The standard model describes the performance of a Gate-all-around (GAA) Junction-less (JL) nanosheet device with a gate dielectric of SiO 2 and HfO 2 , each having a thickness of 1 nm. The performance of both the classical and quantum models of the GAA nanosheet device is evaluated using the visual TCAD tool, which measures the ION , IOFF , ION/ IOFF , threshold voltage, DIBL, gain parameters (g m , g d , A v ), gate capacitance, and cut-off frequency ( fT ). The device is suited for applications needing rapid switching since it has a low gate capacitance of the order of 10 –18 , according to the simulation results. A transconductance (g m ) value of 21 μS and an impressive cut-off frequency of 9.03 GHz are displayed during device analysis. A detailed investigation has also been done into the P-type device response for the same device. Finally, the proposed GAA nanosheet device is used in the inverter model. The NSFET-based inverter, although having higher gate capacitance, has the shortest propagation latency.
- Open Access
- 18 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 145516-145526
A Novel Automated Chromosome Analyzer Software Bundle for Karyotyping and Birth Defect Analysis
Karyotyping is a procedure to diagnose birth defects using chromosome pair. During the Karyotyping chromosomes arranged based on the length and each chromosome will be paired based on various parameters such as, chromosome length, banding pattern and Centromere position. Many methods are proposed to identify the above parameters to improve the Karyotyping accuracy. Since, it’s a challenging task for researchers to improve the accuracy of Karyotyping compared with clinical assays. In this paper, a novel computer geometry method is proposed for chromosome Karyotyping using inbuilt deep learning models with algorithms for chromosome segmentation, overlapped separation, banding pattern analysis and classification of chromosomes. Chromosome classification is carried out using various deep learning models and automated Karyotyping is carried out without manual intervention and model improved the Karyotyping accuracy over the existing methods. In this paper novel computer geometry method is proposed to automate the chromosome analysis in Matlab environment. The developed software provides the accuracy of 99.68% in classification and karyotyping.
- 18 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 145516-145526
Open AccessA Novel Automated Chromosome Analyzer Software Bundle for Karyotyping and Birth Defect Analysis
Karyotyping is a procedure to diagnose birth defects using chromosome pair. During the Karyotyping chromosomes arranged based on the length and each chromosome will be paired based on various parameters such as, chromosome length, banding pattern and Centromere position. Many methods are proposed to identify the above parameters to improve the Karyotyping accuracy. Since, it’s a challenging task for researchers to improve the accuracy of Karyotyping compared with clinical assays. In this paper, a novel computer geometry method is proposed for chromosome Karyotyping using inbuilt deep learning models with algorithms for chromosome segmentation, overlapped separation, banding pattern analysis and classification of chromosomes. Chromosome classification is carried out using various deep learning models and automated Karyotyping is carried out without manual intervention and model improved the Karyotyping accuracy over the existing methods. In this paper novel computer geometry method is proposed to automate the chromosome analysis in Matlab environment. The developed software provides the accuracy of 99.68% in classification and karyotyping.
- Open Access
- 31 January 2023
- Published by Univ. of Malaya in Malaysian Journal of Computer Science
- Vol. 36 (1) , 14-39
DEEP PERONA–MALIK DIFFUSIVE MEAN SHIFT IMAGE CLASSIFICATION FOR EARLY GLAUCOMA AND STARGARDT DISEASE DETECTION
Glaucoma and Stargardt’s, an inherited disease predominantly affect the retinal portion of the eye. The diagnosis of Glaucoma in a fundus image is an arduous, time consuming process. There were many research works carried out to detect early stages of Glaucoma and Stargardt’s disease. However, the accuracy, diagnostic time and performance were not improved. To resolve the above said problems, a computational method called Deep Neural Perona–Malik Diffusive Mean Shift Mode Seeking Segmented Image Classification (DNP-MDMSMSIC) is introduced for the early detection of Glaucoma and Stargardt’s disease with retinal fundus images. The DNP-MDMSMSIC method comprises diverse types of layers that support to identify early detection of disease with improved accuracy and less time. Process as explained; initially, numerous qualified retinal images are given as input to the input layer. These input images are transmitted further to the hidden layer 1 to perform image pre-processing. In DNP-MDMSMSIC, Space-Variant Perona–Malik Diffusive Image Preprocessing is carried out to decrease the noise from input image without removing contents like edges, lines, etc., for image interpretation with a higher peak signal-to-noise ratio. This preprocessed image is further processed in the hidden layer 2 where the feature extraction process is performed to extract features like color, texture, and intensity with a higher degree of accuracy. Based on the extracted features, an input feature image gets segmented in hidden layer 3. Mean Shift Mode Seeking Segmentation algorithm is employed to segment the pixels in image space with corresponding feature space points. Then the segmented images are given to the output layer to perform retinal fundus image classification using Bregman Divergence Function. During the image classification, the distance between two segmented regions (i.e., testing image region of particular class and training image region) with convex is measured. In this way, the retinal fundus images get classified with higher accuracy. Experimental evaluation is performed by considering the metrics such as peak signal-to-noise, disease detection accuracy, disease detection time, and error rate corresponding to the number of retina fundus images and image size. DNP-MDMSMSIC method is designed to detect Glaucoma and Stargardt’s disease at an earlier stage with higher accuracy by 8% and less time by 20% with aid of ACRIMA database.
- 31 January 2023
- Published by Univ. of Malaya in Malaysian Journal of Computer Science
- Vol. 36 (1) , 14-39
Open AccessDEEP PERONA–MALIK DIFFUSIVE MEAN SHIFT IMAGE CLASSIFICATION FOR EARLY GLAUCOMA AND STARGARDT DISEASE DETECTION
Glaucoma and Stargardt’s, an inherited disease predominantly affect the retinal portion of the eye. The diagnosis of Glaucoma in a fundus image is an arduous, time consuming process. There were many research works carried out to detect early stages of Glaucoma and Stargardt’s disease. However, the accuracy, diagnostic time and performance were not improved. To resolve the above said problems, a computational method called Deep Neural Perona–Malik Diffusive Mean Shift Mode Seeking Segmented Image Classification (DNP-MDMSMSIC) is introduced for the early detection of Glaucoma and Stargardt’s disease with retinal fundus images. The DNP-MDMSMSIC method comprises diverse types of layers that support to identify early detection of disease with improved accuracy and less time. Process as explained; initially, numerous qualified retinal images are given as input to the input layer. These input images are transmitted further to the hidden layer 1 to perform image pre-processing. In DNP-MDMSMSIC, Space-Variant Perona–Malik Diffusive Image Preprocessing is carried out to decrease the noise from input image without removing contents like edges, lines, etc., for image interpretation with a higher peak signal-to-noise ratio. This preprocessed image is further processed in the hidden layer 2 where the feature extraction process is performed to extract features like color, texture, and intensity with a higher degree of accuracy. Based on the extracted features, an input feature image gets segmented in hidden layer 3. Mean Shift Mode Seeking Segmentation algorithm is employed to segment the pixels in image space with corresponding feature space points. Then the segmented images are given to the output layer to perform retinal fundus image classification using Bregman Divergence Function. During the image classification, the distance between two segmented regions (i.e., testing image region of particular class and training image region) with convex is measured. In this way, the retinal fundus images get classified with higher accuracy. Experimental evaluation is performed by considering the metrics such as peak signal-to-noise, disease detection accuracy, disease detection time, and error rate corresponding to the number of retina fundus images and image size. DNP-MDMSMSIC method is designed to detect Glaucoma and Stargardt’s disease at an earlier stage with higher accuracy by 8% and less time by 20% with aid of ACRIMA database.
- Open Access
- 1 January 2023
- Published by Tech Science Press in Intelligent Automation and Soft Computing
- Vol. 36 (2) , 1283-1299
Early Detection Glaucoma and Stargardt’s Disease Using Deep Learning Techniques
Retinal fundus images are used to discover many diseases. Several Machine learning algorithms are designed to identify the Glaucoma disease. But the accuracy and time consumption performance were not improved. To address this problem Max Pool Convolution Neural Kuan Filtered Tobit Regressive Segmentation based Radial Basis Image Classifier (MPCNKFTRS-RBIC) Model is used for detecting the Glaucoma and Stargardt’s disease by early period using higher accuracy and minimal time. In MPCNKFTRS-RBIC Model, the retinal fundus image is considered as an input which is preprocessed in hidden layer 1 using weighted adaptive Kuan filter. Then, preprocessed retinal fundus is given for hidden layer 2 for extracting the features like color, intensity, texture with higher accuracy. After extracting these features, the Tobit Regressive Segmentation process is performed by hidden layer 3 for partitioning preprocessed image within more segments by analyzing the pixel with the extracted features of the fundus image. Then, the segmented image was given to output layer. The radial basis function analyzes the testing image region of a particular class as well as training image region with higher accuracy and minimum time consumption. Simulation is performed with retinal fundus image dataset with various performance metrics namely peak signal-to-noise ratio, accuracy and time, error rate concerning several retina fundus image and image size.
- 1 January 2023
- Published by Tech Science Press in Intelligent Automation and Soft Computing
- Vol. 36 (2) , 1283-1299
Open AccessEarly Detection Glaucoma and Stargardt’s Disease Using Deep Learning Techniques
Retinal fundus images are used to discover many diseases. Several Machine learning algorithms are designed to identify the Glaucoma disease. But the accuracy and time consumption performance were not improved. To address this problem Max Pool Convolution Neural Kuan Filtered Tobit Regressive Segmentation based Radial Basis Image Classifier (MPCNKFTRS-RBIC) Model is used for detecting the Glaucoma and Stargardt’s disease by early period using higher accuracy and minimal time. In MPCNKFTRS-RBIC Model, the retinal fundus image is considered as an input which is preprocessed in hidden layer 1 using weighted adaptive Kuan filter. Then, preprocessed retinal fundus is given for hidden layer 2 for extracting the features like color, intensity, texture with higher accuracy. After extracting these features, the Tobit Regressive Segmentation process is performed by hidden layer 3 for partitioning preprocessed image within more segments by analyzing the pixel with the extracted features of the fundus image. Then, the segmented image was given to output layer. The radial basis function analyzes the testing image region of a particular class as well as training image region with higher accuracy and minimum time consumption. Simulation is performed with retinal fundus image dataset with various performance metrics namely peak signal-to-noise ratio, accuracy and time, error rate concerning several retina fundus image and image size.
- 9 June 2022
- Published by Elsevier in Experimental Cell Research
- Vol. 418 (1) , 113251
Mathematical approach for segmenting chromosome clusters in metaspread images
- 9 June 2022
- Published by Elsevier in Experimental Cell Research
- Vol. 418 (1) , 113251
Mathematical approach for segmenting chromosome clusters in metaspread images