Dr. Ghosh's basic research interest is different practical applications of computer vision and pattern recognition techniques. His post-graduate research was on intelligent sensor fusion for online tool-condition monitoring in a metal-cutting process. In his doctoral research Dr. Ghosh applied video processing and Bayesian learning framework for 3D model building and video activity recognition in uncontrolled and unpredictable video sequences. During his post-doctoral research, he has been involved in developing computational techniques for medical imaging data analysis, multi-modality fusion and informatics. Dr. Ghosh developed a hierarchical region splitting (HRS; patented) method that could detect and quantify ischemic stroke and traumatic brain injury (TBI) in animal and clinical multi-modality magnetic resonance imaging and spectroscopy (MRI, MRS) data. These tools could also delineate injury topology – specifically ischemic core and penumbra in MRI that was further correlated with brain inflammation related cytokine-chemokine and tissue immunohistochemical information. Core-penumbra proportions were studied successfully for applicability in therapeutic monitoring for post-ischemic stem-cell implantation and hypothermia – individually and in combination. Dr. Ghosh has worked with automated anatomical brain parsing for region-specific medical information that leads to multi-modality fusion for on-going research on predictive models for clinical outcomes in stroke and TBI. One of his current research interest is to extend this injury topology concept to other pathologies and bringing in Bayesian and other machine learning techniques for intelligent and robust medical decision support systems. Dr. Ghosh is also contuning his research on video understanding and exploratory pattern recognition for human activity classification, 3D model building and robotic navogation.
-
Automated Core-Penumbra Quantification in Neonatal Ischemic Brain Injury by N. Ghosh, X. Yuan, C.I. Turenius, B. Tone, K. Ambadipudi, E.Y. Snyder, A. Obenaus, S. Ashwal J of Cerebral Blood Flow and Metabolism (JCBFM) Vol. 33(12) 2161-2170 (2012)
-
Incremental Unsupervised 3-D Vehicle Model Learning from Video by Ghosh N., Bhanu B. IEEE Trans. on Intelligent Transportation Systems (TITS) 11 423-440 (2010)
-
Hypothermia Modulates Cytokine Responses after Neonatal Rat Hypoxic-Ischemic Injury and Reduces Brain Damage by Ghosh N., Yuan X. , Mcfadden B. , Tone B. , Bellinger D. L., Obenaus A. , Ashwal S. ASN Neuro 6 1-15 (2014)
-
Automated Detection of Brain Abnormalities in Neonatal Hypoxia Ischemic Injury from MR Images by Ghosh N., Sun Y. , Bhanu B. , Ashwal S. , Obenaus A. Medical Image Analysis 18 1059-1069 (2014)
-
Combined Diffusion Tensor and Magnetic Resonance Spectroscopic Imaging Methodology for Automated Regional Brain Analysis: Application in a Normal Pediatric Population by Ghosh N., Holshouser B. , Oyoyo U. , Barnes S. , Tong K. , Ashwal S. Developmental Neuroscience 39 413-429 (2017)
-
A Biomarker for Predicting Responsiveness to Stem Cell Therapy Based on Mechanism-of-Action: Evidence from Cerebral Injury by Hartman R. E., Nathan N. H., Ghosh N. , Pernia C. D., Law J. , Nuryyev R. , Plaia A. , Yusof A. , Tone B. , Dulcich M. , Wakeman D. R., Dilmac N. , Niles W. D., Sidman R. L., Obenaus A. , Snyder E. Y., Ashwal S. Cell Reports 31 107622- (2020)
-
MCA-DN: Multi-path convolution leveraged attention deep network for salvageable tissue detection in ischemic stroke from multi-parametric MRI by Vupputuri A., Gupta A. , Ghosh N. Computers in Biology and Medicine 136 104724- (2021)
-
Detection of Phonocardiogram Event Patterns in Mitral Valve Prolapse: An Automated Clinically Relevant Explainable Diagnostic Framework by S R. B., Patra M. , Sinha A. , Sengupta A. , Ghosh N. IEEE Transactions on Instrumentation and Measurement 72 22596671- (2023)
-
Evolving Bayesian Graph for 3D Vehicle Model Building from Video by Ghosh N., Bhanu B. IEEE Trans. on Intelligent Transportation Systems (TITS) 15 563-578 (2014)
Principal Investigator
- Camera Firmware Video Management Stack and Application Development RASHMI MIND TECHNOLOGIES PRIVATE LIMITED
- Deep Curvelet-Net- A Generalized CNN architecture for Medical Image Analysis and Restoration using Curvelet Features Science and Engineering Research Board (SERB)
- Deep Curvelet-Net- A Generalized CNN architecture for Medical Image Analysis and Restoration using Curvelet Features Anusandhan National Research Foundation (ANRF)
- Generative Artificial Intelligence Models for Medical Imaging Software As A Service RASHMI MIND TECHNOLOGIES PRIVATE LIMITED
Co-Principal Investigator
- Development and Validation of an Immerssive Virtual Reality System Integrating Motor Imagery and Action Observation to Enhance Motor Recovery in Post-stroke rehabilitation: A Pragmatic Trial Indian Council of Medical Research, Dept. of Health Research, Ministry of Health and Family Welfare
- VIRTUAL LABS PROJECT(PHASE-III EXTENDED) MHRD, Department of Higher Education, NMEICT, New Delhi
Ph. D. Students
Abhra Majumder
Area of Research: Computer Vision for Advanced Driver Assist Systems
Aditya Kameswara Rao Nandula
Area of Research: Computer Vision and Pattern Recognition
Dhaladhuli Jahnavi
Area of Research: Neurocardiac Signal Analysis and Diagnostics
Dipdisha Bose
Area of Research: Audio Signal Processing
G Satya Prasad
Area of Research: Signal Processing
Ishita Maiti
Area of Research: Medical Image Analysis
Pritam Chakraborty
Area of Research: Transformer Diagnostics with Machine Learning
Souvik Datta
Area of Research: Biomedical Signal Analysis
