Gautam Kumar's Talk

Mar 23, 2016 - 15:00

Mar 23, 2016 - 16:00

Room 118 Department of Chemical Engineering

Seminars

Gautam Kumar Washington University in St. Louis USA

Abstract: Brain-machine interfaces (BMIs) are broadly defined as artificial systems which establish direct communications between living brain tissue (neurons) and external devices such as robotic arm. By sensing and interpreting neuronal activities to actuate an external device BMI-based neuroprostheses hold great promise in rehabilitating motor impaired subjects such as amputees. Over the past two decades research efforts have mainly been directed towards designing BMIs which rely on the available visual feedback and cortical learning for the movement corrections and thus the design and incorporation of proprioceptive and tactile feedback pathways have completely been ignored. In the absence of these feedback pathways BMIs fail to differentiate visually similar textures and are unable to provide a complete cortical control over the robotic arm. Recently it has been recognized in BMIs community that the inclusion of sensory feedback from the robotic arm to the brain in BMIs is necessary to improve the versatility of BMI-based neuroprostheses. In this talk I will present a design of artificial sensory (proprioceptive) feedback in an optimal control framework towards developing next generation feedback-enabled closed-loop BMIs. In the first part I will discuss our model predictive control framework in designing artificial proprioceptive feedback for a single-joint movement task. I will emphasize the degradation in the BMI performance in the absence of natural proprioception and show the natural recovery of the BMI motor task within our developed theoretical framework. The second part of the talk will focus on quantitative assessment of the Izhikevich single neuron model the building block of cortical circuit models using (in-vivo like) experimental data. I will conclude the talk by outlining my future research directions.About the speaker: Gautam Kumar is currently conducting a postdoctoral research in the area of brain dynamics and control in the Department of Electrical and Systems Engineering at Washington University in St. Louis working with ShiNung Ching. He received B.Tech. in Chemical Engineering (2005) from the IIT Kanpur; M.S. (2008) and Ph.D. (2013) both in Chemical Engineering with a focus on emerging applications of optimal control theory in neural and small length scale dynamical systems from the Lehigh University under the supervision of Mayuresh Kothare . He was awarded the Rossin Doctoral Fellowship from the Lehigh University in 2011. His research interests include multi-scale modeling of biological and neurophysiological processes understanding neural circuits and mechanisms that give rise to specific cognitive processes and optimal control problems in stochastic systems.