Hippocampal memory prosthesis is a Brain Machine Interface device developed for restoring or enhancing memory functions. It is designed to circumvent damaged hippocampal tissue by re-establishing the ensemble coding of spike trains performed by a normal population of hippocampal neurons. The objective is to restore long-term memory function in a stimulus-specific manner using a multi-input, multi-output (MIMO) nonlinear dynamical model.
The hippocampus is responsible for the formation of new long-term declarative memories: the formation of mnemonic labels that identify a unifying collection of features and form relations between multiple collections of features. It is the degeneration and malformation of hippocampal neurons that is the underlying cause of the memory disorders associated with stroke, epilepsy, dementia, and Alzheimer’s disease.
The hippocampal memory prosthesis is a multi-circuit, biomimetic system that consists of three components: (1) a multi-electrode array for recording the ensemble spike trains from an upstream hippocampal region, e.g., CA3, (2) a VLSI chip with a MIMO nonlinear dynamical model for predicting the output (e.g., CA1) ensemble spike trains based on the ongoing input (e.g., CA3) ensemble spike trains, and (3) an electrical stimulator for stimulating the downstream hippocampal region, e.g., CA1, with the predicted output spatio-temporal patterns of spikes.
The MIMO nonlinear dynamical model of input-output spike train transformation takes the form of the sparse generalized Laguerre-Volterra model. A MIMO model is a concatenation of a series of MISO models that each can be considered a spiking neuron model. Each MISO model consists of (a) a MISO Volterra kernel model transforming the input spike trains (x) to the synaptic potential u, (b) a Gaussian noise term capturing the stochastic properties of spike generation, (c) a threshold for generating output spikes, (e) an adder generating the pre-threshold membrane potential w, and (d) a single-input, single-output Volterra kernel model describing the output spike-triggered feedback nonlinear dynamics.
The hippocampal memory prosthesis has been successfully implemented in rodents, nonhuman primates, and human subjects performing memory-dependent behavioral tasks such as the delayed nonmatch-to-sample (DNMS) task and the delayed match-to-sample (DMS) task.
Selected peer-reviewed publications-
1. Hampson, R.E., Song, D., Robinson, B.S., Fetterhoff, D., Dakos, A.S., Roeder, B.M., Wicks, R.T., Witcher, M.R., Couture, D.E., Laxton, A.W., Munger-Clary, H., Popli, G., Sollman, M.J., Marmarelis, V.Z., Berger, T.W., and Deadwyler, S.A. A hippocampal neural prosthetic for restoration of human memory function. Journal of Neural Engineering, 2018, 15, 036014, DOI: 10.1088/1741-2552/aaaed7.
2. Song, D., Robinson, B.S., Hampson, R.E., Marmarelis, V.Z., Deadwyler, S.A., and Berger, T.W. Sparse large-scale nonlinear dynamical modeling of human hippocampus for memory prostheses. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2018, 26(2), 272-280, DOI: 10.1109/TNSRE.2016.2604423.
3. Robinson, B.S., Berger, T.W., and Song, D. Identification of stable spike-timing-dependent plasticity from spiking activity with generalized multilinear modeling. Neural Computation, 2016, 28:11, 2320-2351, DOI: 10.1162/NECO_a_00883.
4. Song, D., Chan, R.H.M., Robinson, B.S., Marmarelis, V.Z., Opris, I., Hampson, R.E., Deadwyler, S.A., and Berger, T.W. Identification of functional synaptic plasticity from spiking activities using nonlinear dynamical modeling. Journal of Neuroscience Methods, 2015, 244, 123-135, DOI: 10.1016/j.jneumeth.2014.09.023.
5. Song, D., Harway, M., Marmarelis, V.Z., Hampson, R.E., Deadwyler, S.A., and Berger, T.W. Extraction and restoration of hippocampal spatial memories with nonlinear dynamical modeling. Frontiers in Systems Neuroscience, 2014, DOI: 10.3389/fnsys.2014.00097.
6. Song, D., Wang, H., Tu, C.Y., Marmarelis, V.Z., Hampson, R.E., Deadwyler, S.A., and Berger, T.W. Identification of sparse neural functional connectivity using penalized likelihood estimation and basis functions. Journal of Computational Neuroscience, 2013, 35, 335-357.
7. Berger, T.W., Song, D., Chan, R.H.M., Marmarelis,V.Z., LaCoss, J., Wills, J., Hampson, R.E., Deadwyler, S.A., and Granacki, J.J. A hippocampal cognitive prosthesis: Multi-input, multi-output nonlinear modeling and VLSI implementation. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2012, 20(2): 198-211.
8. Song, D., Chan, R.H.M., Marmarelis, V.Z., Hampson, R.E., Deadwyler, S.A., and Berger, T.W. Nonlinear modeling of neural population dynamics for hippocampal prostheses. Neural Networks, 2009, 22, 1340-1351.
9. Song, D., Marmarelis, V.Z., and Berger, T.W. Parametric and non-parametric modeling of short-term synaptic plasticity. Part I: Computational study. Journal of Computational Neuroscience, 2009, 26, 1-19.
10. Song, D., Wang, Z., Marmarelis, V.Z., and Berger, T.W. Parametric and non-parametric modeling of short-term synaptic plasticity. Part II: Experimental study. Journal of Computational Neuroscience, 2009, 26, 21-37.
11. Song, D., Chan, R.H.M., Marmarelis, V.Z., Hampson, R.E., Deadwyler, S.A., and Berger, T.W. Nonlinear dynamic modeling of spike train transformations for hippocampal-cortical prostheses. IEEE Transactions on Biomedical Engineering, 2007, 54, 1053-1066.
12. Song, D., Wang, Z., and Berger, T.W. Contribution of T-type VDCCs to TEA-induced long-term synaptic modification in hippocampal CA1 and dentate gyrus. Hippocampus, 2002, 12, 689-697.