Closing the Bio-Silicon Loop for Cellular Neural Prosthesis using FPGA-based Iono-Neuromorphic Models
01:42 Share

Link to bioRxiv paper: http://biorxiv.org/cgi/content/short/2022.12.27.522047v1?rss=1

Authors: luo, j.

Abstract: Neural prosthetic devices offer the ability to develop novel treatments for previously incurable diseases and ailments, such as deafness, blindness and tetraplegia. There is the potential to extend this concept to incorporate cognitive prosthetics, whereby damaged individual neuron cells or larger brain regions are substituted by silicon neurons, in order to overcome conditions such as stroke or epilepsy. The development of such applications relies heavily upon efficient, scalable and powerful technological platforms, particularly systems capable of running large-scale neural models. The advancements in field-programmable gate array (FPGA) technology provides an excellent foundation for the development of these neural models with the same cost of software-based architectures, but with the performance of close to a dedicated hardware system. This paper illustrates the design of a programmable FPGA-based neural model, which is capable of simulating a large range of ion-channel dynamics and delivering biologically realistic network models. Through comparisons with alternative implementations the proposed model is determined to be more scalable and more computationally efficient. We implemented a hybrid bio-silicon system to demonstrate the ability of silicon devices to provide cellular rehabilitation, restoring the functionality of a damaged biological network.

Copy rights belong to original authors. Visit the link for more info

Podcast created by Paper Player, LLC