Data-Driven and Computational Techniques
for Applications in Materials Modeling and Medical Imaging Problems
ME-EM Research Seminar Speaker Series
Dr. Susanta Ghosh
Research Assistant Professor and Instructor
Michigan Technological University
Abstract: Data-driven computational frameworks can integrate physics-based models and machine learning (ML) algorithms. In this talk, I will summarize our efforts in this direction in multi-scale material modeling and ultrasound elastography.
Two-dimensional (2D) materials have the potential to greatly contribute to nano-scale devices, nano-materials, and nano-composites. The technological exploitation of these systems requires predictive models and simulations to access realistic length-scales. I have developed an atomistic-continuum model for 2D layered crystalline materials. Presently we are incorporating ML algorithms to increase the efficiency of our model. Carbon nanotube (CNT) composites are considered as a promising candidate for space flight materials. We are integrating ML algorithms with molecular dynamics to optimize shear strength of CNT-polymer interfaces. Along a similar line, we have developed Neural Network surrogate models for performance optimization of micro-architectured materials to resolve the “curse of dimensionality” of the design space.
Ultrasound elastography is very prominent in modern medicine as mechanical properties of soft tissues can detect disease, such as breast cancer in a non-invasive manner. Currently, I am combining ML algorithms with partial differential equation -constrained optimization models to improve its predictive ability in presence of measurement noise.
Bio: Susanta Ghosh is currently a Research Assistant Professor and Instructor in the Department of Mechanical Engineering-Engineering Mechanics at the Michigan Technological University.
Earlier he has worked as an Associate in Research at the Duke University. Prior to that he was a postdoctoral scholar at the University of Michigan, Ann Arbor and at the Technical University Of Catalunya, Barcelona. He has received his M.S. and Ph.D. degrees from the Indian Institute of Science (IISc), Bangalore, and BS degree from Indian Institute of Engineering Science and Technology, Shibpur.
His research focuses on Data-driven Modeling, Predictive Material Modeling, Inverse Problems for Ultrasound Elastography, Computational Mechanics, Probabilistic Risk Modeling, and High-Performance-Computing (HPC).
Thursday, March 7 at 3:00 pm
R. L. Smith Mechanical Engineering-Engineering Mechanics Building (MEEM), 406
1400 Townsend Drive, Houghton, MI 49931