Neetu Sood
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Neetu Sood has a total of 16 co-authors. Here are the top co-authors by number of shared publications:
- Indu Saini
NIT Jalandhar
7 shared publications - Mayank Kumar Singh
Department of Electronics and Communication Engineering, Dr. B. R. Ambedkar National Institute of Technology, Jalandhar, India
2 shared publications Ajay Kumar Sharma
Department of Computer Science and Engineering, I.K. Gujral Punjab Technical University, Jalandhar, India
1 shared publication- Ajay Kumar Sharma
Department of Computer Science and Engineering, National Institute of Technology Delhi, New Delhi, India
1 shared publication - Amandeep Kaur Sohal
Department of Computer Science and Engineering, Amandeep Kaur Sohal, Guru Nanak Dev Engineering College, Ludhiana, India
1 shared publication - Anukul Pandey
Department of Electronics and Communication Engineering, Delhi Technological University, Delhi, India
1 shared publication
Latest Publications
- 2 May 2024
- Published by Springer Nature in Signal, Image and Video Processing
- Vol. 18 (6-7) , 1-13
Sub-RENet: a wavelet-based network for super resolution of diagnostic ultrasound
- 2 May 2024
- Published by Springer Nature in Signal, Image and Video Processing
- Vol. 18 (6-7) , 1-13
Sub-RENet: a wavelet-based network for super resolution of diagnostic ultrasound
- 16 February 2024
- Published by National Taiwan University in Biomedical Engineering: Applications, Basis and Communications
- Vol. 36 (02)
MULTI-LABEL CLASSIFICATION OF THORACIC DISEASES USING STRUCTURED DEEP LEARNING FRAMEWORK
The current medical scenario indicates that thoracic diseases are the primary cause of illnesses of human beings worldwide. COVID-19 outbreak results in the generation of tremendous Chest X-Ray (CXR) and Computed Tomography (CT) image data archives because CXR and CT imaging are the foremost effective screening tools that characterize patterns and features of pathologies present in the lungs. After careful observation, these chest radiographs show that images are often labeled with more than one pathology, that extends disease diagnosis problem to tedious multi-label classification task. So, this paper proposes a modified mish activation function and compound loss-based Separable Convolution Neural Network (SCNN) model that accomplishes image-level detection of multiple pathologies and COVID-19 from multi-modality chest radiographs. We investigate the power of separable convolution in the CNN network that interprets spatial and depth-wise dimensions of each pixel in CXR and CT images. A modified mish called as Swmish, a non-monotonic activation function, is introduced that preserves negative gradient flow of input and capture fine-grained detail of multiple regions. Moreover, a compound loss function of dice coefficient and binary cross entropy is applied that strongly optimizes the model during training. The generality of the proposed SCNN model is confirmed after conducting comprehensive experiments on multi-label CXR and COVID19-CT datasets. The proposed model gains an AUC score of 0.91 and loss of 0.13 on the multi-label CXR dataset and achieves a testing accuracy of 0.97, AUC score of 0.99 and a loss of 0.11 on the COVID19-CT dataset. These results show that the proposed model achieves highest evaluation metrics in classifying binary and multi-label thoracic abnormalities from multi-modality chest radiographs.
- 16 February 2024
- Published by National Taiwan University in Biomedical Engineering: Applications, Basis and Communications
- Vol. 36 (02)
MULTI-LABEL CLASSIFICATION OF THORACIC DISEASES USING STRUCTURED DEEP LEARNING FRAMEWORK
The current medical scenario indicates that thoracic diseases are the primary cause of illnesses of human beings worldwide. COVID-19 outbreak results in the generation of tremendous Chest X-Ray (CXR) and Computed Tomography (CT) image data archives because CXR and CT imaging are the foremost effective screening tools that characterize patterns and features of pathologies present in the lungs. After careful observation, these chest radiographs show that images are often labeled with more than one pathology, that extends disease diagnosis problem to tedious multi-label classification task. So, this paper proposes a modified mish activation function and compound loss-based Separable Convolution Neural Network (SCNN) model that accomplishes image-level detection of multiple pathologies and COVID-19 from multi-modality chest radiographs. We investigate the power of separable convolution in the CNN network that interprets spatial and depth-wise dimensions of each pixel in CXR and CT images. A modified mish called as Swmish, a non-monotonic activation function, is introduced that preserves negative gradient flow of input and capture fine-grained detail of multiple regions. Moreover, a compound loss function of dice coefficient and binary cross entropy is applied that strongly optimizes the model during training. The generality of the proposed SCNN model is confirmed after conducting comprehensive experiments on multi-label CXR and COVID19-CT datasets. The proposed model gains an AUC score of 0.91 and loss of 0.13 on the multi-label CXR dataset and achieves a testing accuracy of 0.97, AUC score of 0.99 and a loss of 0.11 on the COVID19-CT dataset. These results show that the proposed model achieves highest evaluation metrics in classifying binary and multi-label thoracic abnormalities from multi-modality chest radiographs.
- 9 August 2023
- Published by IOP Publishing in Biomedical Physics & Engineering Express
- Vol. 9 (5) , 055022
A framework based on the information domain to measure coupling changes in electrophysiological signals
Objectives. In this paper, the features of physiological signals of healthy dataset are extracted using the linear and non-linear techniques, and a comparison has been made on healthy young and old subjects to study the aging and gender-related changes in the contribution of Heart Rate (HR), Blood Pressure (BP), and Respiration (RESP). Methods. To quantify the coupling changes in cardiovascular, cardiorespiratory, and vasculorespiratory complexity, an information domain approach based on compensated transfer entropy (cTE) is proposed. Result. The results show that there is a substantial decrease in the flow of information from BP tro the time interval between successive R-peaks (RR) and from RR to BP. There is also a significant decrease in the flow of information from RESP to BP and RESP to RR but there is no significant change in the information flow from BP to RESP and RR to RESP. Conclusion. We have done linear and non-linear analysis on the healthy datasets of young and old subjects. As already existed techniques lacks in studying complex behaviours of electrophysiological signals so to overcome these limitations, we have proposed compensated transfer entropy (cTE). We conducted an investigation to determine the degree to which recordings of RESP, BP, and HR can be utilized to predict changes in the other parameters. Specifically, the proposed analysis examined the relationship between these variables and assessed their consistency across different age groups and genders. By analyzing the data, we aimed to gain insights into the interdependencies and predictive potential of these physiological measures in relation to each other.
- 9 August 2023
- Published by IOP Publishing in Biomedical Physics & Engineering Express
- Vol. 9 (5) , 055022
A framework based on the information domain to measure coupling changes in electrophysiological signals
Objectives. In this paper, the features of physiological signals of healthy dataset are extracted using the linear and non-linear techniques, and a comparison has been made on healthy young and old subjects to study the aging and gender-related changes in the contribution of Heart Rate (HR), Blood Pressure (BP), and Respiration (RESP). Methods. To quantify the coupling changes in cardiovascular, cardiorespiratory, and vasculorespiratory complexity, an information domain approach based on compensated transfer entropy (cTE) is proposed. Result. The results show that there is a substantial decrease in the flow of information from BP tro the time interval between successive R-peaks (RR) and from RR to BP. There is also a significant decrease in the flow of information from RESP to BP and RESP to RR but there is no significant change in the information flow from BP to RESP and RR to RESP. Conclusion. We have done linear and non-linear analysis on the healthy datasets of young and old subjects. As already existed techniques lacks in studying complex behaviours of electrophysiological signals so to overcome these limitations, we have proposed compensated transfer entropy (cTE). We conducted an investigation to determine the degree to which recordings of RESP, BP, and HR can be utilized to predict changes in the other parameters. Specifically, the proposed analysis examined the relationship between these variables and assessed their consistency across different age groups and genders. By analyzing the data, we aimed to gain insights into the interdependencies and predictive potential of these physiological measures in relation to each other.
- 1 January 2023
- Published by Inderscience Publishers in International Journal of Medical Engineering and Informatics
- Vol. 15 (4) , 311-322
Analysis of chest X-ray images using deep learning approaches
- 1 January 2023
- Published by Inderscience Publishers in International Journal of Medical Engineering and Informatics
- Vol. 15 (4) , 311-322
Analysis of chest X-ray images using deep learning approaches
- 11 August 2022
- Published by Springer Nature in The Visual Computer
- Vol. 39 (10) , 4901-4914
Log exponential shrinkage: a denoising technique for breast ultrasound images
- 11 August 2022
- Published by Springer Nature in The Visual Computer
- Vol. 39 (10) , 4901-4914
Log exponential shrinkage: a denoising technique for breast ultrasound images
- 12 July 2021
- Published by Bentham Science Publishers Ltd. in Recent Advances in Computer Science and Communications
- Vol. 14 (4) , 1086-1096
An Energy Efficient Routing Approach to Enhance Coverage for Application-Specific Wireless Sensor Networks Using Genetic Algorithm
Background: An information gathering is a typical and important task in agriculture monitoring and military surveillance. In these applications, minimization of energy consumption and maximization of network lifetime have prime importance for green computing. As wireless sensor networks comprise of a large number of sensors with limited battery power and deployed at remote geographical locations for monitoring physical events, therefore it is imperative to have minimum consumption of energy during network coverage. The WSNs help in accurate monitoring of remote environment by collecting data intelligently from the individual sensors. Objective: The paper is motivated from green computing aspect of wireless sensor network and an Energy-efficient Weight-based Coverage Enhancing protocol using Genetic Algorithm (WCEGA) is presented. The WCEGA is designed to achieve continuously monitoring of remote areas for a longer time with least power consumption. Methods: The cluster-based algorithm consists two phases: cluster formation and data transmission. In cluster formation, selection of cluster heads and cluster members areas based on energy and coverage efficient parameters. The governing parameters are residual energy, overlapping degree, node density and neighbor’s degree. The data transmission between CHs and sink is based on well-known evolution search algorithm i.e. Genetic Algorithm. Results: The results of WCEGA are compared with other established protocols and shows significant improvement of full coverage and lifetime approximately 40% and 45% respectively. Conclusion: This paper proposes an evolutionary method to improve an energy-efficient clustering protocol for longer full coverage.
- 12 July 2021
- Published by Bentham Science Publishers Ltd. in Recent Advances in Computer Science and Communications
- Vol. 14 (4) , 1086-1096
An Energy Efficient Routing Approach to Enhance Coverage for Application-Specific Wireless Sensor Networks Using Genetic Algorithm
Background: An information gathering is a typical and important task in agriculture monitoring and military surveillance. In these applications, minimization of energy consumption and maximization of network lifetime have prime importance for green computing. As wireless sensor networks comprise of a large number of sensors with limited battery power and deployed at remote geographical locations for monitoring physical events, therefore it is imperative to have minimum consumption of energy during network coverage. The WSNs help in accurate monitoring of remote environment by collecting data intelligently from the individual sensors. Objective: The paper is motivated from green computing aspect of wireless sensor network and an Energy-efficient Weight-based Coverage Enhancing protocol using Genetic Algorithm (WCEGA) is presented. The WCEGA is designed to achieve continuously monitoring of remote areas for a longer time with least power consumption. Methods: The cluster-based algorithm consists two phases: cluster formation and data transmission. In cluster formation, selection of cluster heads and cluster members areas based on energy and coverage efficient parameters. The governing parameters are residual energy, overlapping degree, node density and neighbor’s degree. The data transmission between CHs and sink is based on well-known evolution search algorithm i.e. Genetic Algorithm. Results: The results of WCEGA are compared with other established protocols and shows significant improvement of full coverage and lifetime approximately 40% and 45% respectively. Conclusion: This paper proposes an evolutionary method to improve an energy-efficient clustering protocol for longer full coverage.
- 1 April 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 11, 326-330
Brain Computer Interface Based Smart Environment Control
The main advantage of a Brain Computer Interfaced (BCI) systems is that it enables direct communication between brain and computer. This study proposes an Electroencephalogram (EEG) based BCI system for smart environment control. Features from EEG data, ISRUC-Sleep was extracted. Extracted features from data were used for training a classifier for classification of the cognitive stage of the person (Alert, Relaxed and Sleep). The weighted k nearest neighbor (Wk-NN) algorithm based classifier was designed on MATLAB. And the environment was controlled based on the cognitive state of the person. The accuracy achieved for classification of the cognitive state was 92.5%. At last a prototype smart environment control was practiced.
- 1 April 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 11, 326-330
Brain Computer Interface Based Smart Environment Control
The main advantage of a Brain Computer Interfaced (BCI) systems is that it enables direct communication between brain and computer. This study proposes an Electroencephalogram (EEG) based BCI system for smart environment control. Features from EEG data, ISRUC-Sleep was extracted. Extracted features from data were used for training a classifier for classification of the cognitive stage of the person (Alert, Relaxed and Sleep). The weighted k nearest neighbor (Wk-NN) algorithm based classifier was designed on MATLAB. And the environment was controlled based on the cognitive state of the person. The accuracy achieved for classification of the cognitive state was 92.5%. At last a prototype smart environment control was practiced.
- 1 January 2017
- Published by IGI Global Scientific Publishing
- p. 1-21
Nonlinear Complexity Sorting Approach for 2D ECG Data Compression
Nonlinear Complexity Sorting Approach for 2D ECG Data Compression: 10.4018/978-1-5225-0660-7.ch001: Cardiovascular Disease (CVD) is globally acknowledged research problem. The continuous Electrocardiogram (ECG) monitoring can assist in tackling the problem
- 1 January 2017
- Published by IGI Global Scientific Publishing
- p. 1-21
Nonlinear Complexity Sorting Approach for 2D ECG Data Compression
Nonlinear Complexity Sorting Approach for 2D ECG Data Compression: 10.4018/978-1-5225-0660-7.ch001: Cardiovascular Disease (CVD) is globally acknowledged research problem. The continuous Electrocardiogram (ECG) monitoring can assist in tackling the problem
- 25 October 2015
- Published by Springer Nature
- p. 107-117
Application of Compressed Sensing in Cognitive Radio
- 25 October 2015
- Published by Springer Nature
- p. 107-117
Application of Compressed Sensing in Cognitive Radio