Subodh Kumar
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
Subodh Kumar has a total of 20 co-authors. Here are the top co-authors by number of shared publications:
- Neeraj Kumar Sharma
University of Delhi
8 shared publications - Naveen Kumar
University of Delhi
5 shared publications Naveen Kumar
University of Delhi
4 shared publications- Ankit Rajpal
Department of Computer Science, Faculty of Mathematical Sciences, University of Delhi, Delhi, India
3 shared publications Neeraj Kumar Sharma
Ram Lal Anand College, University of Delhi
3 shared publicationsSanjeev Sharma
A.B.V.I.M.S and R.M.L Hospital
3 shared publications
Latest Publications
- 2 August 2026
- Published by Springer Nature
- p. 202-215
Cervical Cancer Detection Using a Machine Learning Framework with PSO-Assisted Feature Selection
- 2 August 2026
- Published by Springer Nature
- p. 202-215
Cervical Cancer Detection Using a Machine Learning Framework with PSO-Assisted Feature Selection
- 22 December 2025
- Published by Springer Nature in Circuits, Systems, and Signal Processing
- Vol. 45 (7) , 5655-5694
GARIAmark: A Geometric Attack-Resilient Image Adaptive Watermarking Scheme Using Quaternion Wavelet Transform and Nutcracker Optimizer Algorithm
- 22 December 2025
- Published by Springer Nature in Circuits, Systems, and Signal Processing
- Vol. 45 (7) , 5655-5694
GARIAmark: A Geometric Attack-Resilient Image Adaptive Watermarking Scheme Using Quaternion Wavelet Transform and Nutcracker Optimizer Algorithm
- 25 June 2025
- Published by Springer Nature
- p. 511-523
A DST-DCT-Based Adaptive Color Image Watermarking Scheme Using Coati Optimization Algorithm
- 25 June 2025
- Published by Springer Nature
- p. 511-523
A DST-DCT-Based Adaptive Color Image Watermarking Scheme Using Coati Optimization Algorithm
- 25 June 2025
- Published by Springer Nature
- p. 437-449
Exploring EEG Feature Extraction and Explainable AI for Accurate Depth of Anesthesia Prediction
- 25 June 2025
- Published by Springer Nature
- p. 437-449
Exploring EEG Feature Extraction and Explainable AI for Accurate Depth of Anesthesia Prediction
- 25 June 2025
- Published by Springer Nature
- p. 423-435
Assessing Depth of Anesthesia Using PPG Signals
- Open Access
- 1 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 119185-119206
XAI-VSDoA: An Explainable AI-Based Scheme Using Vital Signs to Assess Depth of Anesthesia
Administration of anesthesia is essential in surgical procedures, ensuring patient unconsciousness and safety. Traditional Depth of Anesthesia (DoA) assessment methods rely heavily on the clinical expertise of anesthesiologists and patient physiological responses, which can vary widely due to age, weight, and ethnicity. This variability poses significant challenges in maintaining appropriate anesthesia levels and making timely decisions in critical situations. To address these challenges, we propose XAI-VSDoA, an explainable AI model using vital signs designed to augment DoA assessment by providing accurate predictions and interpretable insights. Our study experimented with various machine learning classifiers, including XGBoost, CatBoost, LightGBM, Random Forest, ResNet, and Feed-forward Neural Networks. Among these, the XGBoost model achieved the highest accuracy, with 99.34% on the University of Queensland dataset and 93.07% on the VitalDB dataset. Statistical testing confirmed that XGBoost outperformed the other models.We employed explainable AI techniques such as LIME and SHAP to identify the top 10 features significantly influencing the model’s predictions, ensuring the model’s transparency and reliability. These methods consistently highlighted the same influential features, reinforcing the model’s interpretability. Our proposed scheme demonstrated exceptional performance using numeric vital signs, with XAI techniques validating the key features. This interpretability boosts confidence in the model, supporting critical reflection on anesthesia management. Our findings underscore the potential of XAI-VSDoA as a valuable tool for clinical use, enhancing patient safety and decision-making in anesthesia.
- 1 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 119185-119206
Open AccessXAI-VSDoA: An Explainable AI-Based Scheme Using Vital Signs to Assess Depth of Anesthesia
Administration of anesthesia is essential in surgical procedures, ensuring patient unconsciousness and safety. Traditional Depth of Anesthesia (DoA) assessment methods rely heavily on the clinical expertise of anesthesiologists and patient physiological responses, which can vary widely due to age, weight, and ethnicity. This variability poses significant challenges in maintaining appropriate anesthesia levels and making timely decisions in critical situations. To address these challenges, we propose XAI-VSDoA, an explainable AI model using vital signs designed to augment DoA assessment by providing accurate predictions and interpretable insights. Our study experimented with various machine learning classifiers, including XGBoost, CatBoost, LightGBM, Random Forest, ResNet, and Feed-forward Neural Networks. Among these, the XGBoost model achieved the highest accuracy, with 99.34% on the University of Queensland dataset and 93.07% on the VitalDB dataset. Statistical testing confirmed that XGBoost outperformed the other models.We employed explainable AI techniques such as LIME and SHAP to identify the top 10 features significantly influencing the model’s predictions, ensuring the model’s transparency and reliability. These methods consistently highlighted the same influential features, reinforcing the model’s interpretability. Our proposed scheme demonstrated exceptional performance using numeric vital signs, with XAI techniques validating the key features. This interpretability boosts confidence in the model, supporting critical reflection on anesthesia management. Our findings underscore the potential of XAI-VSDoA as a valuable tool for clinical use, enhancing patient safety and decision-making in anesthesia.
- Open Access
- 25 October 2023
- Published by Association for Computing Machinery (ACM)
- p. 1-8
Predicting Depth of Anesthesia using EEG Signals and Deep Convolution Network
Depth of Anesthesia is the degree of reversible unconsciousness achieved through anesthetic agents. An appropriate anesthetic depth must be maintained during surgical procedures to prevent perioperative complications. Electroencephalogram (EEG) signals allow us to capture the extent of brain activity in various stages of anesthesia. EEG signals thereby help anesthesiologists make informed decisions regarding the administration of anesthesia based on the extent of brain activity displayed by a patient. Bi-spectral Index (BIS) monitors use features extracted from EEG signal to display the depth of anesthesia as a value between 0 and 100. However, the algorithm behind BIS monitors’ functioning is unknown and these are costly machines. Under these circumstances, patient care or monitoring becomes entirely dependent on anesthesiologists’ experience. In this paper, we propose a model using deep Convolutional Neural Network (CNN) to predict the depth of anesthesia into three classes (Anesthesia Light, Anesthesia OK, Anesthesia Deep). The input of this network is a spectrogram obtained using Short-Time Fourier Transform (STFT) applied to EEG signals. This choice of using an STFT-derived spectrogram was made based on CNN’s superior performance over raw EEG waveforms-based CNN. The trained model achieved an accuracy of 90.8%. Experimental results reveal that the proposed scheme utilizing spectrograms achieves either superior or similar performance as compared to other works for accuracy and sensitivity even though it uses relatively fewer parameters. Further, beyond traditional evaluation metrics, this study employs F1-score and precision as additional criteria for assessing the proposed scheme, offering a comprehensive view of its efficacy.
- 25 October 2023
- Published by Association for Computing Machinery (ACM)
- p. 1-8
Open AccessPredicting Depth of Anesthesia using EEG Signals and Deep Convolution Network
Depth of Anesthesia is the degree of reversible unconsciousness achieved through anesthetic agents. An appropriate anesthetic depth must be maintained during surgical procedures to prevent perioperative complications. Electroencephalogram (EEG) signals allow us to capture the extent of brain activity in various stages of anesthesia. EEG signals thereby help anesthesiologists make informed decisions regarding the administration of anesthesia based on the extent of brain activity displayed by a patient. Bi-spectral Index (BIS) monitors use features extracted from EEG signal to display the depth of anesthesia as a value between 0 and 100. However, the algorithm behind BIS monitors’ functioning is unknown and these are costly machines. Under these circumstances, patient care or monitoring becomes entirely dependent on anesthesiologists’ experience. In this paper, we propose a model using deep Convolutional Neural Network (CNN) to predict the depth of anesthesia into three classes (Anesthesia Light, Anesthesia OK, Anesthesia Deep). The input of this network is a spectrogram obtained using Short-Time Fourier Transform (STFT) applied to EEG signals. This choice of using an STFT-derived spectrogram was made based on CNN’s superior performance over raw EEG waveforms-based CNN. The trained model achieved an accuracy of 90.8%. Experimental results reveal that the proposed scheme utilizing spectrograms achieves either superior or similar performance as compared to other works for accuracy and sensitivity even though it uses relatively fewer parameters. Further, beyond traditional evaluation metrics, this study employs F1-score and precision as additional criteria for assessing the proposed scheme, offering a comprehensive view of its efficacy.
- 5 September 2023
- Published by Springer Nature in Circuits, Systems, and Signal Processing
- Vol. 43 (2) , 965-993
CXRmark: A Watermarking Scheme for Chest X-Rays Using Online Sequential Reduced Kernel ELM
- 5 September 2023
- Published by Springer Nature in Circuits, Systems, and Signal Processing
- Vol. 43 (2) , 965-993
CXRmark: A Watermarking Scheme for Chest X-Rays Using Online Sequential Reduced Kernel ELM
- Open Access
- 24 July 2023
- Published by Institute of Mathematics and Informatics Bulgarian Academy of Sciences in Innovative STEM Education
- Vol. 5 (1) , 13-22
Detection and Correction of Repeated Restricted Burst Errors in Sub-Blocks
In this paper, we present bounds on check symbols required for the codes capable to detect as well as correct the restricted bursts errors that are repeating themselves in a single sub-block. For these codes the whole code length is considered to be subdivided into a certain number of mutually exclusive sub-blocks of equal length.
- 24 July 2023
- Published by Institute of Mathematics and Informatics Bulgarian Academy of Sciences in Innovative STEM Education
- Vol. 5 (1) , 13-22
Open AccessDetection and Correction of Repeated Restricted Burst Errors in Sub-Blocks
In this paper, we present bounds on check symbols required for the codes capable to detect as well as correct the restricted bursts errors that are repeating themselves in a single sub-block. For these codes the whole code length is considered to be subdivided into a certain number of mutually exclusive sub-blocks of equal length.
- 21 April 2023
- Published by Elsevier in Expert Systems with Applications
- Vol. 226
WSOmark: An adaptive dual-purpose color image watermarking using white shark optimizer and Levenberg–Marquardt BPNN
- 21 April 2023
- Published by Elsevier in Expert Systems with Applications
- Vol. 226
WSOmark: An adaptive dual-purpose color image watermarking using white shark optimizer and Levenberg–Marquardt BPNN