Sudhir Pathak
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Sudhir Pathak has a total of 26 co-authors. Here are the top co-authors by number of shared publications:
Walter Schneider
University of Pittsburgh Pittsburgh, Pennsylvania USA
4 shared publications- B. V. Rathish Kumar
Department of Mathematics and Statistics, Indian Institute of Technology Kanpur, Kanpur, India
2 shared publications - Aditya Nigam
Indian Institute of Technology Mandi
1 shared publication - Ajay Niranjan
Center for Image-Guided Neurosurgery, Department of Neurological Surgery, University of Pittsburgh Medical Center, Pittsburgh, USA
1 shared publication - Anthony Zuccolotto
Psychology Software Tools Pittsburgh, Pennsylvania USA
1 shared publication - Arnav Bhavsar
Indian Institute of Technology Mandi
1 shared publication
Latest Publications
- 24 October 2026
- Published by Wiley in Magnetic Resonance in Medicine
HARP: HARmonizing In‐Vivo Single‐Shell Diffusion MRI Using Phantom‐Only Training
Purpose: Combining multi‐site diffusion MRI (dMRI) data is hindered by inter‐scanner variability, which confounds subsequent analysis. Previous harmonization methods require large, matched or traveling human subjects from multiple sites, which are impractical to acquire in many situations. This study aims to develop a deep learning‐based dMRI harmonization framework that eliminates the reliance on multi‐site in‐vivo traveling human data for training. Methods: HARP employs a voxel‐wise 1D neural network trained on an easily transportable diffusion phantom. The model learns relationships between spherical harmonics coefficients of different sites without memorizing spatial structures. Results: HARP reduced inter‐scanner variability levels significantly in various measures. Quantitatively, it decreased inter‐scanner variability as measured by standard error in FA (12%), MD (10%), and GFA (30%) with scan‐rescan standard error as the baseline, while preserving fiber orientations and tractography after harmonization. Conclusion: We believe that HARP represents an important first step toward dMRI harmonization using only phantom data, thereby obviating the need for complex, matched in vivo multi‐site cohorts. This phantom‐only strategy substantially enhances the feasibility and scalability of quantitative dMRI for large‐scale clinical studies.
- 24 October 2026
- Published by Wiley in Magnetic Resonance in Medicine
HARP: HARmonizing In‐Vivo Single‐Shell Diffusion MRI Using Phantom‐Only Training
Purpose: Combining multi‐site diffusion MRI (dMRI) data is hindered by inter‐scanner variability, which confounds subsequent analysis. Previous harmonization methods require large, matched or traveling human subjects from multiple sites, which are impractical to acquire in many situations. This study aims to develop a deep learning‐based dMRI harmonization framework that eliminates the reliance on multi‐site in‐vivo traveling human data for training. Methods: HARP employs a voxel‐wise 1D neural network trained on an easily transportable diffusion phantom. The model learns relationships between spherical harmonics coefficients of different sites without memorizing spatial structures. Results: HARP reduced inter‐scanner variability levels significantly in various measures. Quantitatively, it decreased inter‐scanner variability as measured by standard error in FA (12%), MD (10%), and GFA (30%) with scan‐rescan standard error as the baseline, while preserving fiber orientations and tractography after harmonization. Conclusion: We believe that HARP represents an important first step toward dMRI harmonization using only phantom data, thereby obviating the need for complex, matched in vivo multi‐site cohorts. This phantom‐only strategy substantially enhances the feasibility and scalability of quantitative dMRI for large‐scale clinical studies.
- Open Access
- 5 September 2026
- Published by Springer Nature in Scientific Reports
- Vol. 16 (1)
Adaptive multi-scale attention-enhanced V-NET for accurate and efficient multi-shell diffusion MRI estimation
- 5 September 2026
- Published by Springer Nature in Scientific Reports
- Vol. 16 (1)
Open AccessAdaptive multi-scale attention-enhanced V-NET for accurate and efficient multi-shell diffusion MRI estimation
- 28 March 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-5
LFANET: Transforming 3T Single-Shell to 7T Multi-Shell DMRI Using Deep Learning Based Leapfrog and Attention
HARDI-based diffusion MRI acquisition technique is a relatively recent modality of interest as it can yield more accurate fiber tracts. Besides, HARDI at higher magnetic strength is more sensitive to tissue changes and accurately estimate anatomical details in the human brain. However, a higher magnetic strength scanner is costly and not available in most clinical settings. Furthermore, due to signal-to-noise ratio issues and severe imaging artefacts, most existing 3T dMRI scanners with low gradient-strengths generally acquire single-shell up to b = 1000s/mm 2 . Hence, in this work, we consider the task of transforming the 3T single-shell HARDI signal (at b = 1000s/mm 2 ) to a 7T multi-shell HARDI signal utilizing the proposed deep learning model LF ANet. The proposed model consists of modules based on a Leapfrog method and an attention module. In addition, we have included suitable loss functions such as L 1 and total variation loss. Several quantitative and qualitative results have been presented to show the effectiveness of the proposed method.
- 28 March 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-5
LFANET: Transforming 3T Single-Shell to 7T Multi-Shell DMRI Using Deep Learning Based Leapfrog and Attention
HARDI-based diffusion MRI acquisition technique is a relatively recent modality of interest as it can yield more accurate fiber tracts. Besides, HARDI at higher magnetic strength is more sensitive to tissue changes and accurately estimate anatomical details in the human brain. However, a higher magnetic strength scanner is costly and not available in most clinical settings. Furthermore, due to signal-to-noise ratio issues and severe imaging artefacts, most existing 3T dMRI scanners with low gradient-strengths generally acquire single-shell up to b = 1000s/mm 2 . Hence, in this work, we consider the task of transforming the 3T single-shell HARDI signal (at b = 1000s/mm 2 ) to a 7T multi-shell HARDI signal utilizing the proposed deep learning model LF ANet. The proposed model consists of modules based on a Leapfrog method and an attention module. In addition, we have included suitable loss functions such as L 1 and total variation loss. Several quantitative and qualitative results have been presented to show the effectiveness of the proposed method.
- 1 March 2022
- Published by AIP Publishing in Physics of Fluids
- Vol. 34 (3) , 031909
Modeling and simulation of the potential indoor airborne transmission of SARS-CoV-2 virus through respiratory droplets
Respiratory viruses are transported from an infected person to other neighboring people through respiratory droplets. These small droplets are easily advected by air currents in a room and can potentially infect others. In this work, the spread of droplets released during coughing, talking, and normal breathing is numerically analyzed in a typical conference room setting. The room space is occupied by ten people, with eight people sitting around a conference table and two people standing. Four different scenarios are considered, with the air-conditioning turned on/off and people wearing/not-wearing masks, to understand the spread of respiratory droplets inside the room. The flow in the room is simulated using a multiphase mixture model with properties computed for the inhaled and exhaled air using fundamental gas relations. The transport of respiratory droplets is analyzed using the discrete phase model with a range of droplet sizes fitted to data from previous experimental studies. The mask is modeled as porous media with the properties of a woven fabric computed using a newly developed model for multilayered homemade masks. The human inhalation and exhalation are modeled using analytical functions to mimic the biological flow patterns during breathing, coughing, and talking. Important observations about the air flow and dispersion of respiratory droplets in the conference room are presented based on the numerical analysis. Animations of all the results are included to provide insight into flow physics of the various dynamic conditions occurring in the room during an ongoing meeting. Although this study is conducted for a typical conference room, the newly developed models and techniques can be applied to other confined environments.
- 1 March 2022
- Published by AIP Publishing in Physics of Fluids
- Vol. 34 (3) , 031909
Modeling and simulation of the potential indoor airborne transmission of SARS-CoV-2 virus through respiratory droplets
Respiratory viruses are transported from an infected person to other neighboring people through respiratory droplets. These small droplets are easily advected by air currents in a room and can potentially infect others. In this work, the spread of droplets released during coughing, talking, and normal breathing is numerically analyzed in a typical conference room setting. The room space is occupied by ten people, with eight people sitting around a conference table and two people standing. Four different scenarios are considered, with the air-conditioning turned on/off and people wearing/not-wearing masks, to understand the spread of respiratory droplets inside the room. The flow in the room is simulated using a multiphase mixture model with properties computed for the inhaled and exhaled air using fundamental gas relations. The transport of respiratory droplets is analyzed using the discrete phase model with a range of droplet sizes fitted to data from previous experimental studies. The mask is modeled as porous media with the properties of a woven fabric computed using a newly developed model for multilayered homemade masks. The human inhalation and exhalation are modeled using analytical functions to mimic the biological flow patterns during breathing, coughing, and talking. Important observations about the air flow and dispersion of respiratory droplets in the conference room are presented based on the numerical analysis. Animations of all the results are included to provide insight into flow physics of the various dynamic conditions occurring in the room during an ongoing meeting. Although this study is conducted for a typical conference room, the newly developed models and techniques can be applied to other confined environments.
- 16 May 2019
- Published by S. Karger AG
- Vol. 34, 28-39
Imaging Techniques for Leksell Radiosurgery
Stereotactic image acquisition is an important aspect of radiosurgery dose planning. Treatment planning is usually performed with magnetic resonance imaging (MRI) and computed tomography (CT). While CT is considered most accurate, MRI provides superior soft tissue contrast needed to delineate the tumor and the organs at risk. MRI is the most common imaging modality used for Leksell radiosurgery. Cerebral angiography is used in addition to MRI for arteriovenous malformation dose planning. Users should be aware of the possibility of MRI distortions while acquiring stereotactic images. The development of protocols for imaging of individual brain lesions and accuracy checks of stereotactic images are the important aspects of Leksell radiosurgery. In special cases, PET, MRS, MEG, etc. can be used to supplement stereotactic MRI. These complementary images can be integrated into treatment planning by stereotactic co-registration of these non-stereotactic images.
- 16 May 2019
- Published by S. Karger AG
- Vol. 34, 28-39
Imaging Techniques for Leksell Radiosurgery
Stereotactic image acquisition is an important aspect of radiosurgery dose planning. Treatment planning is usually performed with magnetic resonance imaging (MRI) and computed tomography (CT). While CT is considered most accurate, MRI provides superior soft tissue contrast needed to delineate the tumor and the organs at risk. MRI is the most common imaging modality used for Leksell radiosurgery. Cerebral angiography is used in addition to MRI for arteriovenous malformation dose planning. Users should be aware of the possibility of MRI distortions while acquiring stereotactic images. The development of protocols for imaging of individual brain lesions and accuracy checks of stereotactic images are the important aspects of Leksell radiosurgery. In special cases, PET, MRS, MEG, etc. can be used to supplement stereotactic MRI. These complementary images can be integrated into treatment planning by stereotactic co-registration of these non-stereotactic images.
- RetractedOpen Access
- 8 April 2019
- Published by Public Library of Science (PLoS) in PLOS Biology
- Vol. 17 (4) , e3000063
Response to: “Questioning the evidence for BCI-based communication in the complete locked-in state”
- 8 April 2019
- Published by Public Library of Science (PLoS) in PLOS Biology
- Vol. 17 (4) , e3000063
RetractedOpen AccessResponse to: “Questioning the evidence for BCI-based communication in the complete locked-in state”