Date: Thursday, October 26, 2023

Time: 1:00 - 2:30 PM EST

Location: GTMI Conference Room 114
 

Virtual Link: Microsoft Teams

Bharat Kanwar

Robotics PhD Student

School of Mechanical Engineering

Georgia Institute of Technology

Committee:

Dr. Anirban Mazumdar (Advisor) – School of Mechanical Engineering, Georgia Institute of Technology

Dr. Stephen Balakirsky (Advisor) – Georgia Tech Research Institute

Dr. Levi Wood – School of Mechanical Engineering, Georgia Institute of Technology

Dr. Brandon Dixon – School of Mechanical Engineering, Georgia Institute of Technology

Dr. Annie Bowles-Welch – Marcus Center for Cell Characterization and Manufacturing

Abstract: 

Immune cell therapy is a rapidly growing field with immense clinical potential for several indications, including regenerating tissue, immunomodulation, and engineered cells for disease removal. As a nascent industry, biomanufacturing of these cell therapies involves lengthy manual protocols which leads to increased risk of failed or inconsistent cell product. This work proposes a framework for designing digital-twin models for bioreactor platforms that integrate novel sensors, imaging, process controls, and perfusion. This framework consists of a modular digital twin that can model the relevant fluid dynamics of convection, diffusion, osmosis, and cell fluxes of the bioreactor platform. Given the sterility requirements for living cell expansion, measurement of important parameters during the process is often untenable. This work proposes methods to compute unmeasured states and parameters from measured ones with an Extended Kalman Filter and predictive models to explore the domain of critical process parameters to control and measure. This framework then proposes an optimal-cost Linear Quadratic Regulator control architecture to regulate nutrients and cell output of the bioreactor process. Bioreactor process control is then demonstrated in a 3-donor hMSC expansion study showing improved expansion in a hollow fiber bioreactor, and a proposed T cell expansion study to investigate phenotype control.