Debarshi Ghosh

Description

Dr. Debarshi Ghosh joined Long Island University in 2024 as a postdoctoral researcher at the Center of Excellence of Dassault Systems to be involved in the core modelling effort of the Living Liver Project in collaboration with clinicians from the Baylor College of Medicine, Houston, USA. He is currently involved in interdisciplinary projects that integrates clinical data to create patient specifics model for liver toxicity as well as for creating machine learning models for early detection of knee osteoarthritis.

Prior to joining Long Island University in 2024, He was assistant professor (research) from September 2022 to December 2023 at Chitkara University, Punjab, India. He is an experienced biomedical researcher with more than 6 years of research experience in computational modelling of biomedical phenomena spanning from drug delivery to digital medicine and AI-driven healthcare diagnostics. He is experienced in in-silico and in-vivo techniques for outcome-based analysis of transdermal drug delivery experiments. Previously, he was a member of the core research team, actively applying and guiding faculties for submission of government research grants. He holds extensive experience in extra-mural research grant writing for research projects and for the creation of interdisciplinary research centres of excellence. His research has won him 6 international and national awards for humanitarian work with technology.

Research

Cellular Model of Hepatic Drug Metabolism and its Impact on Drug Induced Liver Injury: Developed a multiscale framework, by integrating Physiologically Based Pharmacokinetic (PBPK) Model to the Virtual Liver Lobule model and mathematical model describing drug hepatic metabolism to predict drug induced liver injury. Further, the model is used to analyze the impact of liver metabolic zonation creating liver injury using quantitative and sensitivity analysis. The study paved the way for improved understanding of the effects of spatial heterogeneity in drug-induced liver injury, with significant potential for improving clinical practice and enhancing patient care outcomes.

Machine Learning in Knee Image Segmentation: Co-supervised two master's students on machine learning-based early detection of knee osteoarthritis. Our contribution to the project involves developing a YOLOv8 model with knee joint space width for enhanced automated detection of early-grade knee osteoarthritis, which often gets overlooked as the radiographic images are difficult to distinguish with the human eye. The other project involves developing a regression-based model using biomarkers for enhanced prognosis of knee osteoarthritis.

E-Pancreas: A Non-Invasive Insulin Delivery System: Developed a wearable device for non-invasive delivery of insulin without punching the skin. The device enables a pain free inclusive life for diabetic kids, young and old aged patients. Lead the computational simulation for non-invasive insulin delivery using electrical pulses of nanosecond (ns) and millisecond (ms) duration. Co-developed a high frequency electroporator circuit (Simulation and experimental) capable of delivering nanosecond (ns) and millisecond (ms) pulses with varying load conditions. As a part of team was involved in the pharmacokinetic and dynamic study of insulin permeation in Wistar rat skin model. Developed the molecular dynamics study to analyze the effect of electric field on the conformational integrity of insulin.

BhuGoal: A Smart Weather Monitoring System: Developed a device for ultra-local and precise weather prediction system using DTH satellite signal. Lead the technological design and development of data acquisition system (Texas Instrument Microprocessor). Successfully designed 4 MVP’s resulting in deployment of resource optimized final system.

Radio Frequency (RF) Based Sensor for Adulteration Detection of Alcoholic Beverages: Developed and fabricated a portable microfluid sensor for detection of adulteration in alcoholic beverages. The substrates of the proposed sensor have a running polysiloxane microfluidic channel where the samples of alcoholic beverages are applied. The ethanol concentration is detected with the help of a radio frequency spectroscopy. The results are validated with nuclear magnetic resonance (NMR) spectroscopy.

A Comparative Study of Biopsy Feature Based Breast Cancer Detection: Developed Fine needle aspiration (FNA) biopsy technique-based detection and classification of breast cancer. Classification techniques, including logistic regression function (LR), sequential minimal optimization (SMO), LAZY-LWL (Locally Weighted Learning), and simple logistic function, are employed for detection and classification. The study considers two different split ratios (80:20 and 60:40) and two cross-validation ratios (5 and 10). The results highlight that using the 80:20 split ratio resulted in higher accuracy across various algorithms. The Simple logistic function emerged as the most effective algorithm, consistently achieving accuracy values in the range of 98%.

Distinctions & Awards

• Diabetes Innovation Challenge T1D Exchange, USA – Organized by the Organized by the T1D Community Exchange USA and JDRF Type 1 Diabetes Research Funding and Advocacy [Winner]
• Economic Times the Power of Ideas 2018 - Organized by the Economic Times, Department of Science and Technology (DST), and Facebook. [Winner and received grant of INR 5.00 Lakhs]
• Gandhian Young Technological Innovation (GYTI) 2018 - Organized by SRISTI, Techpedia, and the Department of Science and Technology (DST). [Winner]
• Indian Innovation Challenge Design (IICDC) 2018 - Organized by the Texas Instrument (TI) and Department of Science and Technology (DST). [Winner and received grant of INR 25 Lakhs]
• India Innovation Growth Programme (IIGP) 2.0 University Challenge - Organized by the Department of Science and Technology (DST), Lockheed Martin, and Tata Trust. [Winner and received grant of INR 10 Lakhs]

Selected Publications

• Ghosh, D., Saluja, N., & Singh, T. G. (2019). A critical analysis of electroporation in medical technology. Int J Pharm Sci Res, 10(1), 23-8.
• Haldiyan, A., Ghosh, D., Saluja, N., Ganeshan, S., & Singh, T. G. (2021). Comparison of nano-second and millisecond pulse generators for biological applications of electroporation. Research Journal of Pharmacy and Technology, 14(5), 2843-2851.
• Ganesan, S., Ghosh, D., Taneja, A., Saluja, N., Rani, S., Singh, A., ... & Noya, I. D. (2022). A modified marx generator circuit with enhanced tradeoff between voltage and pulse width for electroporation applications. Electronics, 11(13), 2013.
• Sharma, A. K., Debarshi Ghosh, D., Saluja, N. K., & Singh, T. G. (2021). A mathematical model to expedite electroporation-based vaccine development for COVID-19. Bioint. Res. App. Chem, 12, 1951-1961.
• Singh, N. P., Sarathy, P., Mantri, A., & Singh, G. (2021). Spatio-Temporal Pattern Representation from AI Inspired Brain Model in Spiking Neural Network. Biointerface Research in Applied Chemistry, 12.
• Ghosh, D., Saluja, N. K., & Singh, T. G. (2021). A FEM study of molecular transport through a single nanopore in a spherical cell. Biointerface Res. Appl. Chem., 12(3), 2958-2969.
• Camara Dit Pinto, S., Cherkaoui, J., Ghosh, D., Cazaubon, V., Benzeroual, K. E., Levine, S. M., ... & Gallo, N. R. (2024). A virtual scalable model of the Hepatic Lobule for acetaminophen hepatotoxicity prediction. NPJ Digital Medicine, 7(1), 340.
• Dhingra, N., Ghosh, D., Saluja, N., & Sabapathay, T. (2023). Radio Frequency Based Sensor for Adulteration Measurement in a Continuous Two Phase-Flow of Alcoholic Beverages. Sensing and Imaging, 24(1), 32.
• Ganesan, S., Taneja, A., Ghosh, D., & Saluja, N. (2023, October). High-Frequency Load Independent Electroporator Circuit Design for Nanosecond Pulse Generation. In 2023 First International Conference on Advances in Electrical, Electronics and Computational Intelligence (ICAEECI) (pp. 1-5). IEEE.
• Haldiyan, A., Ghosh, D., & Saluja, N. K. (2022, May). Analysis of six stage Marx electroporator circuit based on pulsed parameters. In AIP Conference Proceedings (Vol. 2357, No. 1). AIP Publishing.
• Dhingra, N., Saluja, N., Garg, R., Kanwar, V., & Ghosh, D. (2020). Radio frequency-based material estimation of old age 3D sculptures. Materials Today: Proceedings, 28, 2015-2020.

Patents

• Antitheft System for a Veichle, Patent Application Number: 202011027232, Patent Granted- IN 486773.
• Apparatus for Non-Invasive Delivery of Compound, Patent Application Number: 201711039928, Patent Granted – IN 507123
• An Automated System for Sciatica Pain Management, Patent Application Number: 202311060783, (Patent Published)
• Device and Method for Detecting and Quantifying Strabismus in an Eye, Patent Application No: 202311048191, (Patent Published)
• A Retrofit Machine Learning Based Milling Machine, Patent Application No: 202311043753, (Patent Published)
• System and Method for Augmented Reality-Based Online Shopping, Patent Application No: 202311035527, (Patent Published)
• Robust and Real-time Augmented Reality-Based Driving Assistive System, Patent Application No: 202111056565, (Patent Published)
• System and Device for Structural Monitoring of Sculptures, Patent Application Number: 202011033544, (Patent Published)
• Device for Detection of Purity of a Fluid, Patent Application Number: 202211002675, (Patent Published).
• Virtual Reality System to Monitor Behavior of Rodents, Patent Application Number: 202211046851, (Patent Published)
• A Finger Worn Device for Remote Operation of an External Articles, Patent Application Number: 202211046850, (Patent Published)
• Resonator Based System for Estimation of Adulteration, Patent Application Number:202211058126, (Patent Published)
• Apparatus And Method for Microwave Torrefaction of Biomass, Patent Application Number: 202211070435, (Patent Published)
• System For Metal Cladding Using Microwave Energy, Patent Application Number: 202211070925, (Patent Published)
• DNA Vaccine Delivery, Patent Application Number: 202311002072, (Patent Published)