Sitikantha Roy
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
Research Subjects
Sitikantha Roy has a total of 78 co-authors. Here are the top co-authors by number of shared publications:
- Lalan Kumar
Indian Institute of Technology Delhi
6 shared publications - Suriya Prakash Muthukrishnan
Department of Physiology, All India Institute of Medical Sciences
6 shared publications - Saswath Ghosh
Department of Applied Mechanics, Indian Institute of Technology Delhi, New Delhi 110016, India
5 shared publications Umesh Gautam
Department of Applied Mechanics, Indian Institute of Technology Delhi
4 shared publications- Abhilash Awasthi
Indian Institute of Technology Delhi
3 shared publications - Anant Jain
Indian Institute of Technology Delhi
3 shared publications
Latest Publications
- Open Access
- 19 October 2026
- Published by Springer Nature in Scientific Reports
Mechanical load modulates premovement cortical activity during voluntary upper limb movement
Examining how scalp electroencephalography (EEG) activity changes under increasing motor demands may provide insights into how the brain prepares for movement. In this exploratory study, premovement EEG activity was examined in ten healthy male participants (n = 10) performing a unilateral bicep curl task. Cue-locked EEG signals preceding kinematically defined movement onset were analyzed using complementary event-related potential analysis (ERP), time–frequency analysis (TF), and cross-frequency coupling analyses to examine load-dependent EEG responses. As mechanical load increased, premovement EEG activity exhibited temporal, frequency-specific, and spatially distributed changes. Mixed-effects analyses indicated that the relationship between graded mechanical load and premovement EEG activity was better described by a non-linear than a linear model. These differences were pronounced at higher load levels and exhibited non-linear and frequency-specific patterns across the frontal, central, and contralateral sensorimotor scalp regions. These findings provide preliminary insights into how increasing mechanical demand is associated with frequency-specific, spatially distributed, and non-linear changes in premovement EEG activity before movement onset.
- 19 October 2026
- Published by Springer Nature in Scientific Reports
Open AccessMechanical load modulates premovement cortical activity during voluntary upper limb movement
Examining how scalp electroencephalography (EEG) activity changes under increasing motor demands may provide insights into how the brain prepares for movement. In this exploratory study, premovement EEG activity was examined in ten healthy male participants (n = 10) performing a unilateral bicep curl task. Cue-locked EEG signals preceding kinematically defined movement onset were analyzed using complementary event-related potential analysis (ERP), time–frequency analysis (TF), and cross-frequency coupling analyses to examine load-dependent EEG responses. As mechanical load increased, premovement EEG activity exhibited temporal, frequency-specific, and spatially distributed changes. Mixed-effects analyses indicated that the relationship between graded mechanical load and premovement EEG activity was better described by a non-linear than a linear model. These differences were pronounced at higher load levels and exhibited non-linear and frequency-specific patterns across the frontal, central, and contralateral sensorimotor scalp regions. These findings provide preliminary insights into how increasing mechanical demand is associated with frequency-specific, spatially distributed, and non-linear changes in premovement EEG activity before movement onset.
- 4 June 2026
- Published by Springer Nature
- p. 290-301
EEG-Based 90-Degree Turn Intention Detection for Brain-Computer Interface
- 4 June 2026
- Published by Springer Nature
- p. 290-301
EEG-Based 90-Degree Turn Intention Detection for Brain-Computer Interface
- 2 January 2026
- Published by Wiley in Journal of Biomedical Materials Research Part A
- Vol. 114 (1) , e70017
Biomechanical Characterization of Epileptic Brain for Surgical Planning of Intractable Epilepsy
Neurological disorders, including epilepsy, often manifest with altered brain stiffness, particularly in affected regions. The complex relationship between the biomechanical and microstructural characteristics of epileptic brain (EB) is poorly understood and warrants comprehensive research. This study investigates the in vitro viscoelastic properties of surgically excised EB tissues ( S = 20) and marginal normal brain (NB) ( S = 10) from the same individuals diagnosed with varying epileptogenic substrates. The microstructural characterization including neuron density, myelin and collagen content was also performed. Additionally, in vivo magnetic resonance elastography (MRE) was conducted on one subject to complement the in vitro findings as a pilot investigation. EB exhibited significantly higher stiffness than NB (storage modulus : 6.49 3.83 kPa vs. 1.97 0.40 kPa; loss modulus : 1.53 0.93 kPa vs. 0.61 0.31 kPa; p = 0.001). Among pathological subtypes, mesial temporal sclerosis (MTS) tissues were the stiffest (: 8.42 4.05 kPa and : 1.95 1.03 kPa), while focal cortical dysplasia (FCD) tissues were the softest (: 2.56 0.45 kPa and : 0.83 0.41 kPa). Other etiologies fell between these extremes. Microstructural correlations revealed a strong positive relationship between stiffness and neuronal density ( r = 0.81), a moderate negative correlation with myelin content ( r = −0.52), and no significant association with collagen content ( r = 0.15), indicating that cellular composition, rather than extracellular matrix components, predominantly contributes tissue mechanics. The in vivo MRE findings in an FCD lesion (: 2.65 0.30 kPa; : 0.91 0.25 kPa) showed concordance with the in vitro measurement of specimen from same subject (: 2.50 0.41 kPa; : 0.47 0.36 kPa). A deeper understanding of the mechanical differences between EB and NB has implications for personalized surgical planning, the development of high‐fidelity computational models, and improved elastography and non‐rigid image registration algorithms.
- 2 January 2026
- Published by Wiley in Journal of Biomedical Materials Research Part A
- Vol. 114 (1) , e70017
Biomechanical Characterization of Epileptic Brain for Surgical Planning of Intractable Epilepsy
Neurological disorders, including epilepsy, often manifest with altered brain stiffness, particularly in affected regions. The complex relationship between the biomechanical and microstructural characteristics of epileptic brain (EB) is poorly understood and warrants comprehensive research. This study investigates the in vitro viscoelastic properties of surgically excised EB tissues ( S = 20) and marginal normal brain (NB) ( S = 10) from the same individuals diagnosed with varying epileptogenic substrates. The microstructural characterization including neuron density, myelin and collagen content was also performed. Additionally, in vivo magnetic resonance elastography (MRE) was conducted on one subject to complement the in vitro findings as a pilot investigation. EB exhibited significantly higher stiffness than NB (storage modulus : 6.49 3.83 kPa vs. 1.97 0.40 kPa; loss modulus : 1.53 0.93 kPa vs. 0.61 0.31 kPa; p = 0.001). Among pathological subtypes, mesial temporal sclerosis (MTS) tissues were the stiffest (: 8.42 4.05 kPa and : 1.95 1.03 kPa), while focal cortical dysplasia (FCD) tissues were the softest (: 2.56 0.45 kPa and : 0.83 0.41 kPa). Other etiologies fell between these extremes. Microstructural correlations revealed a strong positive relationship between stiffness and neuronal density ( r = 0.81), a moderate negative correlation with myelin content ( r = −0.52), and no significant association with collagen content ( r = 0.15), indicating that cellular composition, rather than extracellular matrix components, predominantly contributes tissue mechanics. The in vivo MRE findings in an FCD lesion (: 2.65 0.30 kPa; : 0.91 0.25 kPa) showed concordance with the in vitro measurement of specimen from same subject (: 2.50 0.41 kPa; : 0.47 0.36 kPa). A deeper understanding of the mechanical differences between EB and NB has implications for personalized surgical planning, the development of high‐fidelity computational models, and improved elastography and non‐rigid image registration algorithms.
- Open Access
Load-Aware Calibration of EMG-Driven Musculoskeletal Models for Accurate and Generalizable Joint Torque Estimation
Accurate EMG-driven musculoskeletal (MSK) modeling is critical for biomechanics, rehabilitation, and assistive technology. However, most models calibrate parameters under a single load, ignoring the fact that tasks with similar kinematics may differ in mechanical demand. This study introduces a load-aware calibration framework to improve joint torque prediction accuracy and generalizability. Surface EMG and joint kinematics were recorded from eleven participants during elbow flexion-extension under 0, 2, and 4\,kg loads. We evaluated three calibration strategies (load-specific, global, cross-load) and three optimization frameworks (simulated annealing (SA), particle swarm optimization (PSO), and hybrid PSO-pattern search (PSO-PS)). Results indicate that load-specific calibration significantly improves performance, with lower RMSE and higher correlation ($r > 0.75$). Parameters related to muscle force, fiber length, and activation dynamics showed high load sensitivity. PSO-based methods yielded more consistent and physiologically plausible estimates than simulated annealing. The proposed framework enables MSK models to distinguish between visually similar but mechanically distinct tasks, supporting robust subject-specific modeling for clinical and real-world applications.
Open AccessLoad-Aware Calibration of EMG-Driven Musculoskeletal Models for Accurate and Generalizable Joint Torque Estimation
Accurate EMG-driven musculoskeletal (MSK) modeling is critical for biomechanics, rehabilitation, and assistive technology. However, most models calibrate parameters under a single load, ignoring the fact that tasks with similar kinematics may differ in mechanical demand. This study introduces a load-aware calibration framework to improve joint torque prediction accuracy and generalizability. Surface EMG and joint kinematics were recorded from eleven participants during elbow flexion-extension under 0, 2, and 4\,kg loads. We evaluated three calibration strategies (load-specific, global, cross-load) and three optimization frameworks (simulated annealing (SA), particle swarm optimization (PSO), and hybrid PSO-pattern search (PSO-PS)). Results indicate that load-specific calibration significantly improves performance, with lower RMSE and higher correlation ($r > 0.75$). Parameters related to muscle force, fiber length, and activation dynamics showed high load sensitivity. PSO-based methods yielded more consistent and physiologically plausible estimates than simulated annealing. The proposed framework enables MSK models to distinguish between visually similar but mechanically distinct tasks, supporting robust subject-specific modeling for clinical and real-world applications.
- 19 December 2025
- Published by Elsevier in Neural Networks
- Vol. 197, 108500
A physics informed neural network architecture for moving boundary problems in science and engineering
- 19 December 2025
- Published by Elsevier in Neural Networks
- Vol. 197, 108500
A physics informed neural network architecture for moving boundary problems in science and engineering
- Open Access
Deep Muscle EMG construction using A Physics-Integrated Deep Learning approach
Electromyography (EMG)--based computational musculoskeletal modeling is a non-invasive method for studying musculotendon function, human movement, and neuromuscular control, providing estimates of internal variables like muscle forces and joint torques. However, EMG signals from deeper muscles are often challenging to measure by placing the surface EMG electrodes and unfeasible to measure directly using invasive methods. The restriction to the access of EMG data from deeper muscles poses a considerable obstacle to the broad adoption of EMG-driven modeling techniques. A strategic alternative is to use an estimation algorithm to approximate the missing EMG signals from deeper muscle. A similar strategy is used in physics-informed deep learning, where the features of physical systems are learned without labeled data. In this work, we propose a hybrid deep learning algorithm, namely the neural musculoskeletal model (NMM), that integrates physics-informed and data-driven deep learning to approximate the EMG signals from the deeper muscles. While data-driven modeling is used to predict the missing EMG signals, physics-based modeling engraves the subject-specific information into the predictions. Experimental verifications on five test subjects are carried out to investigate the performance of the proposed hybrid framework. The proposed NMM is validated against the joint torque computed from 'OpenSim' software. The predicted deep EMG signals are also compared against the state-of-the-art muscle synergy extrapolation (MSE) approach, where the proposed NMM completely outperforms the existing MSE framework by a significant margin.
Open AccessDeep Muscle EMG construction using A Physics-Integrated Deep Learning approach
Electromyography (EMG)--based computational musculoskeletal modeling is a non-invasive method for studying musculotendon function, human movement, and neuromuscular control, providing estimates of internal variables like muscle forces and joint torques. However, EMG signals from deeper muscles are often challenging to measure by placing the surface EMG electrodes and unfeasible to measure directly using invasive methods. The restriction to the access of EMG data from deeper muscles poses a considerable obstacle to the broad adoption of EMG-driven modeling techniques. A strategic alternative is to use an estimation algorithm to approximate the missing EMG signals from deeper muscle. A similar strategy is used in physics-informed deep learning, where the features of physical systems are learned without labeled data. In this work, we propose a hybrid deep learning algorithm, namely the neural musculoskeletal model (NMM), that integrates physics-informed and data-driven deep learning to approximate the EMG signals from the deeper muscles. While data-driven modeling is used to predict the missing EMG signals, physics-based modeling engraves the subject-specific information into the predictions. Experimental verifications on five test subjects are carried out to investigate the performance of the proposed hybrid framework. The proposed NMM is validated against the joint torque computed from 'OpenSim' software. The predicted deep EMG signals are also compared against the state-of-the-art muscle synergy extrapolation (MSE) approach, where the proposed NMM completely outperforms the existing MSE framework by a significant margin.
- 28 February 2025
- Published by Elsevier in Journal of the Mechanical Behavior of Biomedical Materials
- Vol. 167, 106961
Simulation of a Free Boundary Cell Migration Model through Physics Informed Neural Networks
- 28 February 2025
- Published by Elsevier in Journal of the Mechanical Behavior of Biomedical Materials
- Vol. 167, 106961
Simulation of a Free Boundary Cell Migration Model through Physics Informed Neural Networks
- Open Access
- 16 October 2024
- Published by Springer Nature in Research Square
Comparative Biomechanical Studies of Goat and Human Brain Applicable for Surgical Training
Background Haptics-based neurosurgical simulators play a paramount role in surgical training. The reliability of these simulators depends on the accurate mechanical response of brain tissue. However, no study has yet provided a comprehensive mechanical map of the whole brain, which is crucial for an in-depth understanding of brain mechanics. Methods In our study, we elucidated the viscoelastic behaviour of the whole brain in 12 goat brains and one human brain using dynamic nanoindentation. This involved comparing the mechanical properties of corresponding goat brain regions to human brain regions. Results The cortex gray matter (GM) in the goat brain was approximately 30% stiffer than its white matter (WM), whereas deep GM in the thalamus was 23% stiffer than the deep WM in the corpus callosum. Interestingly, the human brain GM was up to 150% stiffer than the goat brain GM; however, the human brain WM stiffness showed the same range as that of goat brain WM. The stiffness of the midbrain, pons, and medulla oblongata of the brain stem in humans and goats showed no significant difference. The loss moduli did not vary significantly across the brain. Additionally, the formalin fixation doubled the brain tissue stiffness. Conclusion Our study revealed a significant regional heterogeneity in the mechanical characteristics of brain tissue. We found that the GM was consistently stiffer than WM in both goat and human brains and that the human brain GM was stiffer than the goat brain GM. These findings have important implications for understanding brain biomechanics and for developing reliable and comprehensive neurosurgical training platforms.
- 16 October 2024
- Published by Springer Nature in Research Square
Open AccessComparative Biomechanical Studies of Goat and Human Brain Applicable for Surgical Training
Background Haptics-based neurosurgical simulators play a paramount role in surgical training. The reliability of these simulators depends on the accurate mechanical response of brain tissue. However, no study has yet provided a comprehensive mechanical map of the whole brain, which is crucial for an in-depth understanding of brain mechanics. Methods In our study, we elucidated the viscoelastic behaviour of the whole brain in 12 goat brains and one human brain using dynamic nanoindentation. This involved comparing the mechanical properties of corresponding goat brain regions to human brain regions. Results The cortex gray matter (GM) in the goat brain was approximately 30% stiffer than its white matter (WM), whereas deep GM in the thalamus was 23% stiffer than the deep WM in the corpus callosum. Interestingly, the human brain GM was up to 150% stiffer than the goat brain GM; however, the human brain WM stiffness showed the same range as that of goat brain WM. The stiffness of the midbrain, pons, and medulla oblongata of the brain stem in humans and goats showed no significant difference. The loss moduli did not vary significantly across the brain. Additionally, the formalin fixation doubled the brain tissue stiffness. Conclusion Our study revealed a significant regional heterogeneity in the mechanical characteristics of brain tissue. We found that the GM was consistently stiffer than WM in both goat and human brains and that the human brain GM was stiffer than the goat brain GM. These findings have important implications for understanding brain biomechanics and for developing reliable and comprehensive neurosurgical training platforms.
- 23 August 2024
- Published by Springer Nature
- p. 363-370
A Study on Behavior of Functionally Graded Microbeam Under Static Deflection Using Physics-Informed Neural Network
- 23 August 2024
- Published by Springer Nature
- p. 363-370
A Study on Behavior of Functionally Graded Microbeam Under Static Deflection Using Physics-Informed Neural Network
- Open Access
Generative Design of Multimodal Soft Pneumatic Actuators
The recent advancements in machine learning techniques have steered us towards the data-driven design of products. Motivated by this objective, the present study proposes an automated design methodology that employs data-driven methods to generate new designs of soft actuators. One of the bottlenecks in the data-driven automated design process is having publicly available data to train the model. Due to its unavailability, a synthetic data set of soft pneumatic network (Pneu-net) actuators has been created. The parametric design data set for the training of the generative model is created using data augmentation. Next, the Gaussian mixture model has been applied to generate novel parametric designs of Pneu-net actuators. The distance-based metric defines the novelty and diversity of the generated designs. In addition, it is noteworthy that the model has the potential to generate a multimodal Pneu-net actuator that could perform in-plane bending and out-of-plane twisting. Later, the novel design is passed through finite element analysis to evaluate the quality of the generated design. Moreover, the trajectory of each category of Pneu-net actuators evaluates the performance of the generated Pneu-net actuators and emphasizes the necessity of multimodal actuation. The proposed model could accelerate the design of new soft robots by selecting a soft actuator from the developed novel pool of soft actuators.
Open AccessGenerative Design of Multimodal Soft Pneumatic Actuators
The recent advancements in machine learning techniques have steered us towards the data-driven design of products. Motivated by this objective, the present study proposes an automated design methodology that employs data-driven methods to generate new designs of soft actuators. One of the bottlenecks in the data-driven automated design process is having publicly available data to train the model. Due to its unavailability, a synthetic data set of soft pneumatic network (Pneu-net) actuators has been created. The parametric design data set for the training of the generative model is created using data augmentation. Next, the Gaussian mixture model has been applied to generate novel parametric designs of Pneu-net actuators. The distance-based metric defines the novelty and diversity of the generated designs. In addition, it is noteworthy that the model has the potential to generate a multimodal Pneu-net actuator that could perform in-plane bending and out-of-plane twisting. Later, the novel design is passed through finite element analysis to evaluate the quality of the generated design. Moreover, the trajectory of each category of Pneu-net actuators evaluates the performance of the generated Pneu-net actuators and emphasizes the necessity of multimodal actuation. The proposed model could accelerate the design of new soft robots by selecting a soft actuator from the developed novel pool of soft actuators.