EAGER: Implementation of Computational Circuits Based on Neural Dynamical Systems

San Jose State University Foundation, San Jose CA

Investigators

Abstract

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5). The objective of this project is to realize silicon neurons as physical circuits. A key requirement is to match the internal dynamics of the circuit, including timing, with the biological system, as needed for hybrid hardware-biological systems. The approach builds on existing work by the investigator that uses dynamical system theory. This project extends this model to include physiologically relevant parameters that are extracted from functioning biological cells. The project will also develop extraction routines and tools for reverse engineering of small biological neural networks. This offers the potential of enabling the realization of a wider range of neural parameters and a wider range of neural computational circuits. With respect to intellectual merit, the research is motivated by the ability of silicon neurons to implement hybrid electronic-biological systems and to provide a fundamental understanding of the neurobiological systems that are emulated. The specific goal of a physical circuit implementation, versus a software implementation, is motivated by the ability to enable real-time operation, ease interfacing to partial biological systems, and enable integration into robotics and control systems. Expected outcomes include physiologically relevant data to aid simulation, a parameter extraction scheme, and reproducible circuits. With respect to broader impact, the research has the potential to transform the understanding of and the application of hybrid electronic-biological systems by coupling circuit design, dynamical systems, and neuro-physiology. The research engages undergraduates, including those from underrepresented groups, through capstone design projects. Prototype circuits, documentation of the extraction method, and parametric values extracted from identified cells will be made available to other researchers.