.png){kind=icon}
.png){kind=icon}
Return
Advancing the neurocomputer
.png){kind=icon}
OA
AI
DOI：10.1016/j.neucom.2018.01.021.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Cultivating neural populations onto microelectrode arrays (MEAs) is a popular technique used by neuro-scientists to study cortical microcircuits in a more controlled setting than they appear in vivo. However, recent works show that these neural cultures may be more than just another analog to the brain-they can also be used to compute. Researchers in academia and industry may soon need the computing power of a true neuromorphic computer: a silicon computing device that utilizes the processing power of biological neurons. Such a device can be realized by cultivating biological neurons onto an MEA in vitro and carefully interfacing the MEA to a computer or robot. This new device can make use of short-and long-term memory mechanisms intrinsic to biological neurons to train neuronal cultures to perform unique computations. Research shows this type of computing device may be able to solve problems in fields like image recognition, temporal pattern classification, and others. In this paper, we review the methods of interfacing cultures with MEAs, learning mechanisms exhibited in neuronal cultures, the current applications of a neuron-silicon hybrid computing device in research, and potential challenges this field may face. We propose that the current research in neurocomputers has provided the foundations for a new era in computing, one in which the computational power of true biological neural networks may be exploited and eventually surpass the power of artificial neural networks. (C) 2018 Elsevier B.V. All rights reserved.
Keywords:
Neurocomputing
Microelectrode arrays
Silicon-neuron interfaces
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
6.5
Papers:
2.5W
Citations:
6.5W