Hardware Implementation of the Cerebellar Neural Network with Conductance-based Models

Cerebellar cortex plays a vital role in motor learning. To investigate the functions of the cerebellar cortex, a hardware model simulating the cerebellar cortex is implemented on a FPGA chip with the dynamics of the functional cerebellum. The conductance-based models are used to realize the functions of the cerebellar cells and three types of synaptic current models are used to realize the interactions among these cells. The cerebellar models implemented on FPGA can respond the stimulus in 0.02ms, which makes the hardware model more flexible. The hardware model simulates different firing rates of Purkinje cells with different afferent stimulus, so that the action potential maps of five types of cerebellar cells are obtained on the oscilloscope. The results of the hardware implementation are consistent with the biological dynamics of the mammalian cerebellar, which means the functional cerebellum model is reliable and feasible.

[1]  Y. Saeki,et al.  Four distinct phases of basket/stellate cell migration after entering their final destination (the molecular layer) in the developing cerebellum. , 2009, Developmental biology.

[2]  F. Rossi,et al.  Handbook of the Cerebellum and Cerebellar Disorders , 2013, Springer Netherlands.

[3]  Yu Wang,et al.  Large scale recurrent neural network on GPU , 2014, 2014 International Joint Conference on Neural Networks (IJCNN).

[4]  Bertram E. Shi,et al.  Expandable Networks for Neuromorphic Chips , 2007, IEEE Transactions on Circuits and Systems I: Regular Papers.

[5]  E. De Schutter,et al.  Synchronization of golgi and granule cell firing in a detailed network model of the cerebellar granule cell layer. , 1998, Journal of neurophysiology.

[6]  Egidio D'Angelo,et al.  Computational Reconstruction of Pacemaking and Intrinsic Electroresponsiveness in Cerebellar Golgi Cells , 2007, Frontiers in cellular neuroscience.

[7]  Bin Deng,et al.  A multi-FPGA embedded system for the emulation of modular small-world network with real time dynamics , 2016, 2016 12th World Congress on Intelligent Control and Automation (WCICA).

[8]  R. Khazipov,et al.  GABA: a pioneer transmitter that excites immature neurons and generates primitive oscillations. , 2007, Physiological reviews.

[9]  C. Koch,et al.  Multiple channels and calcium dynamics , 1989 .

[10]  R. Hawkes,et al.  The modular cerebellum , 1991, Progress in Neurobiology.

[11]  G. L. Massonc,et al.  Hardware computation of conductance-based neuron models , 2004 .

[12]  S. Lisberger,et al.  The Cerebellum: A Neuronal Learning Machine? , 1996, Science.

[13]  Yutaka Maeda,et al.  Simultaneous perturbation learning rule for recurrent neural networks and its FPGA implementation , 2005, IEEE Transactions on Neural Networks.

[14]  Gordon M. Shepherd,et al.  Dendritic action potentials connect distributed dendrodendritic microcircuits , 2008, Journal of Computational Neuroscience.

[15]  Distribution and morphology of ghrelin-immunopositive cells in the cerebellum of the African ostrich. , 2012, Tissue & cell.

[16]  Lei Wang,et al.  Neural circuit and its functional roles in cerebellar cortex , 2011, Neuroscience bulletin.

[17]  Lei Wang 汪雷,et al.  Neural circuit and its functional roles in cerebellar cortex , 2011, Neuroscience Bulletin.

[18]  Walther Akemann,et al.  Interaction of Kv3 Potassium Channels and Resurgent Sodium Current Influences the Rate of Spontaneous Firing of Purkinje Neurons , 2006, The Journal of Neuroscience.

[19]  Giacomo Indiveri,et al.  An event-based VLSI network of integrate-and-fire neurons , 2004, 2004 IEEE International Symposium on Circuits and Systems (IEEE Cat. No.04CH37512).