Large-scale M1 microcircuit model with plastic input connections from biological PMd neurons used for prosthetic arm control
© Dura-Bernal et al. 2015
Published: 18 December 2015
Pyramidal tract-projecting neurons in layer 5B were connected to a descending spinal cord neural population, which provided excitation to the muscles of a realistic virtual musculoskeletal arm. Proprioceptive feedback from the arm was encoded in an ascending spinal cord population which then projected to M1 layer 2/3. The virtual arm movements were also followed by a robotic arm .
An additional population, which reproduced the spiking patterns recorded from 92 neurons of macaque dorsal premotor cortex (PMd) during a center-out reaching task, was connected to M1 layer 5A providing a modulatory input .
The network was trained to drive the virtual arm to reach multiple targets, by combining arm exploratory movements with reinforcement learning and spike-timing dependent plasticity (STDP). Synaptic plasticity occurred between multiple populations, including between the PMd inputs and layer 5A neurons. Tuning of learning metaparameters was achieved by employing parallel evolutionary algorithms in a high-performance computing cluster.
This work moves towards a new generation of neuroprosthetic systems where biological brain circuits interact directly with biomimetic cortical models, and employ co-adaptation and learning to achieve a functional task. Such a multiscale approach, ranging from the cellular to the behavioral level, will provide deeper insights into brain dynamics and have applications for the diagnosis and restoration of brain disorders.
- Dura-Bernal S, Chadderdon GL, Neymotin SA, Francis JT, Lytton WW: Towards a real-time interface between a biometic model of sensorimotor cortex and a robotic arm. Pattern Recognition Letters. 2014, 36 (15): 204-212.PubMedPubMed CentralView ArticleGoogle Scholar
- Lee G, Matsunaga A, Dura-Bernal S, Zhang W, Lytton WW, Francis JT, Fortes JAB: Towards real-time communication between in vivo neurophysiological data sources and simulator-based brain biomimetic models. Journal of Computational Surgery. 2014, 3: 12-Google Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.