Abstract
We present a biologically inspired and scalable model of the basal ganglia (BG) simulated on the spiking neural network architecture (SpiNNaker) machine, a biologically inspired low-power hardware platform allowing parallel, asynchronous computing. Our BG model consists of six cell populations, where the neuro-computational unit is a conductance-based Izhikevich spiking neuron; the number of neurons in each population is proportional to that reported in anatomical literature. This model is treated as a single-channel of action-selection in the BG, and is scaled-up to three channels with lateral cross-channel connections. When tested with two competing inputs, this three-channel model demonstrates action-selection behavior. The SpiNNaker-based model is mapped exactly on to SpineML running on a conventional computer; both model responses show functional and qualitative similarity, thus validating the usability of SpiNNaker for simulating biologically plausible networks. Furthermore, the SpiNNaker-based model simulates in real time for time-steps $\geq {1}$ ms; power dissipated during model execution is $\approx {1.8}$ W.
Highlights
T HE AIM of this paper is to build a biologically inspired, scalable, spiking neural network model of the Basal Ganglia (BG) on the spiking neural network architecture (SpiNNaker) machine [1]
We have presented a biologically plausible and scalable model of the BG circuit, designed to run on the SpiNNaker machine—a biologically inspired architecture built with lowpower ARM processors, allowing inherent asynchronous, parallel computation, and in real time for time-steps ≥ 1 ms
A single neuro-computational unit in our BG model is simulated with a conductance-based Izhikevich neuron model, facilitated by the underlying SpiNNaker software toolchain, sPyNNaker, which in turn is based on PyNN, a python-based neural network application interface
Summary
T HE AIM of this paper is to build a biologically inspired, scalable, spiking neural network model of the Basal Ganglia (BG) on the spiking neural network architecture (SpiNNaker) machine [1]. Output from the BG is the specific action that is decided upon, referred to as “action-selection” [3], and is relayed to the motor pathway for execution via the thalamus, cortex, and other subcortical structures. Seminal modeling work by Gurney et al [3], [6] introduced the concept of “selection-control” pathways in the BG, a deviation from the more common nomenclature of “direct–indirect” pathways associated with how dopamine controls and executes the action-selection mechanism. The model was demonstrated as a computational tool to study brain disorders [7]–[9], as well as to form a conceptual understanding of the action-selection mechanism adopted by the BG and implemented in robots [10], [11]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: IEEE Transactions on Cognitive and Developmental Systems
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
