A5002598947- Neural dynamics and brain function
- Functional Brain Connectivity Studies
- EEG and Brain-Computer Interfaces
- Blind Source Separation Techniques
- Fault Detection and Control Systems
- Neuroscience and Music Perception
- Statistical Methods and Inference
- Anesthesia and Pain Management
- Hearing Loss and Rehabilitation
- Advanced MRI Techniques and Applications
- Cardiac, Anesthesia and Surgical Outcomes
- Sparse and Compressive Sensing Techniques
- Gaussian Processes and Bayesian Inference
- Spectroscopy and Chemometric Analyses
- Optical Imaging and Spectroscopy Techniques
- Control Systems and Identification
- Anesthesia and Neurotoxicity Research
- Ecosystem dynamics and resilience
- Explainable Artificial Intelligence (XAI)
- Simulation Techniques and Applications
- Pain Management and Opioid Use
- Intensive Care Unit Cognitive Disorders
- Sleep and Wakefulness Research
- Cancer, Stress, Anesthesia, and Immune Response
- Heart Rate Variability and Autonomic Control
Massachusetts General Hospital
2021-2024
Stanford Medicine
2024
Stanford University
2023-2024
Harvard University
2022-2024
Palo Alto University
2023
Athinoula A. Martinos Center for Biomedical Imaging
2023
Harvard University Press
2021
University of Maryland, College Park
2017-2020
Indian Institute of Technology Kharagpur
2016
Opioids administered to treat postsurgical pain are a major contributor the opioid crisis, leading chronic use in considerable proportion of patients. Initiatives promoting opioid-free or opioid-sparing modalities perioperative management have led reduced administration operating room, but this reduction could unforeseen detrimental effects terms postoperative outcomes, as relationship between intraoperative usage and later requirements is not well understood.To characterize association...
Even though human experience unfolds continuously in time, it is not strictly linear; instead, entails cascading processes building hierarchical cognitive structures. For instance, during speech perception, humans transform a varying acoustic signal into phonemes, words, and meaning, these levels all have distinct but interdependent temporal Time-lagged regression using response functions (TRFs) has recently emerged as promising tool for disentangling electrophysiological brain responses...
1 Abstract Even though human experience unfolds continuously in time, it is not strictly linear; instead, entails cascading processes building hierarchical cognitive structures. For instance, during speech perception, humans transform a varying acoustic signal into phonemes, words, and meaning, these levels all have distinct but interdependent temporal Time-lagged regression using response functions (TRFs) has recently emerged as promising tool for disentangling electrophysiological brain...
Linear parametric state-space models are a ubiquitous tool for analyzing neural time series data, providing way to characterize the underlying brain dynamics with much greater statistical efficiency than non-parametric data analysis approaches. However, frequently time-varying, exhibiting rapid changes in dynamics, transient activity that is often key feature of interest data. Stationary methods can be adapted time-varying scenarios by employing fixed-duration windows under an assumption...
Abstract Preoperative knowledge of expected postoperative pain can help guide perioperative management and focus interventions on patients with the greatest risk acute pain. However, current methods for predicting require patient clinician input or laborious manual chart review often do not achieve sufficient performance. We use routinely collected electronic health record data from a multicenter dataset 234,274 adult non-cardiac surgical to develop machine learning method which predicts...
Identifying the directed connectivity that underlie networked activity between different cortical areas is critical for understanding neural mechanisms behind sensory processing. Granger causality (GC) widely used this purpose in functional magnetic resonance imaging analysis, but there temporal resolution low, making it difficult to capture millisecond-scale interactions underlying Magnetoencephalography (MEG) has millisecond resolution, only provides low-dimensional sensor-level linear...
Modern neurophysiological recordings are performed using multichannel sensor arrays that able to record activity in an increasingly high number of channels numbering the 100s 1000s. Often, underlying lower-dimensional patterns responsible for observed dynamics, but these representations difficult reliably identify existing methods attempt summarize multivariate relationships a post hoc manner from univariate analyses or current blind source separation methods. While such can reveal appealing...
Characterizing the neural dynamics underlying sensory processing is one of central areas investigation in systems and cognitive neuroscience. Neuroimaging techniques such as magnetoencephalography (MEG) Electroencephalography (EEG) have provided significant insights into continuous stimuli, speech, thanks to their high temporal resolution. Existing work context auditory suggests that certain features acoustic envelope, can be used reliable linear predictors response manifested M/EEG. The...
Abstract Cortical ischaemic strokes result in cognitive deficits depending on the area of affected brain. However, we have demonstrated that difficulties with attention and processing speed can occur even small subcortical infarcts. Symptoms appear independent lesion location, suggesting they arise from generalized disruption networks. Longitudinal studies evaluating directional measures functional connectivity this population are lacking. We evaluated six patients minor stroke exhibiting...
Granger causality is among the widely used data-driven approaches for causal analysis of time series data with applications in various areas including economics, molecular biology, and neuroscience. Two main challenges this methodology are: 1) over-fitting as a result limited duration, 2) correlated process noise confounding factor, both leading to errors identifying influences. Sparse estimation via LASSO has successfully addressed these parameter estimation. However, classical statistical...
Spectral analysis using overlapping sliding windows is among the most widely used techniques in analyzing non-stationary time series. Although window convenient to implement, resulting estimates are sensitive length and overlap size. In addition, it undermines dynamics of series as estimate associated each uses only data within. Finally, between consecutive hinders a precise statistical assessment. this paper, we address these shortcomings by explicitly modeling spectral through integrating...
Simultaneously recorded non-invasive multichannel time-series such as electroencephalogram (EEG) and magnetoen-cephalogram (MEG) comes with a very specific challenge due to volume conduction spatial mixing. This makes individual analysis of signal from each sensors, treating them independently, highly redundant often leading sub-optimal results that mischaracterizes dependencies. There exists filtering or blind source separation approaches try decompose EEG/MEG into small number dominant...
Abstract Modern neurophysiological recordings are performed using multichannel sensor arrays that able to record activity in an increasingly high number of channels numbering the 100’s 1000’s. Often, underlying lower-dimensional patterns responsible for observed dynamics, but these representations difficult reliably identify existing methods attempt summarize multivariate relationships a post-hoc manner from univariate analyses, or current blind source separation methods. While such can...
We consider the problem of determining Granger causal influences among sources that are indirectly observed through low-dimensional and noisy linear projections. Commonly used methods proceed in a two-stage fashion, by first solving an inverse to localize sources, then inferring from estimated sources. The inferred links thus inherit various biases source localization techniques, form spatiotemporal priors designed favor spatial localization. In addition, this approach does not account for...
The magnetoencephalography (MEG) response to continuous auditory stimuli, such as speech, is commonly described using a linear filter, the temporal function (TRF). Though components of sensor level TRFs have been well characterized, underlying neural sources responsible for these are not understood. In this work, we provide unified framework determining directly from MEG data, by integrating TRF and distributed forward source models into one, casting joint estimation task Bayesian...
Modern neurophysiological recordings are performed using multichannel sensor arrays that able to record activity in an increasingly high number of channels numbering the 100s 1000s. Often, underlying lower-dimensional patterns responsible for observed dynamics, but these representations difficult reliably identify existing methods attempt summarize multivariate relationships a post hoc manner from univariate analyses or current blind source separation methods. While such can reveal appealing...
Modern neurophysiological recordings are performed using multichannel sensor arrays that able to record activity in an increasingly high number of channels numbering the 100’s 1000’s. Often, underlying lower-dimensional patterns responsible for observed dynamics, but these representations difficult reliably identify existing methods attempt summarize multivariate relationships a post-hoc manner from univariate analyses, or current blind source separation methods. While such can reveal...
Modern neurophysiological recordings are performed using multichannel sensor arrays that able to record activity in an increasingly high number of channels numbering the 100’s 1000’s. Often, underlying lower-dimensional patterns responsible for observed dynamics, but these representations difficult reliably identify existing methods attempt summarize multivariate relationships a post-hoc manner from univariate analyses, or current blind source separation methods. While such can reveal...