SEARCH · Engineering Papers
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
The Susceptible-Infected-Recovered/Removed model is a standard model for epidemiological spread of disease through vulnerable populations. In this paper we show how SIR network dynamics can be implemented using spiking neurons.
A unipolar terminal-attractor based neural associative memory (TABAM) system with adaptive threshold for perfect convergence is presented. By adaptively setting the threshold values for the dynamic iteration for the unipolar binary neuron states with terminal-attractors for the purpose of reducing the spurious states in a Hopfield neural network for associative memory and using the inner product approach, perfect convergence and correct retrieval is achieved. Simulation is completed with a small number of stored states (M) and a small number of neurons (N) but a large M/N ratio. An experiment with optical exclusive-OR logic operation using LCTV SLMs shows the feasibility of optoelectronic implementation of the models. A complete inner-product TABAM is implemented using a PC for calculation of adaptive threshold values to achieve a unipolar TABAM (UIT) in the case where there is no crosstalk, and a crosstalk model (CRIT) in the case where crosstalk corrupts the desired state.
A unipolar terminal-attractor based neural associative memory (TABAM) system with adaptive threshold for perfect convergence is presented. By adaptively setting the threshold values for the dynamic iteration for the unipolar binary neuron states with terminal-attractors for the purpose of reducing the spurious states in a Hopfield neural network for associative memory and using the inner-product approach, perfect convergence and correct retrieval is achieved. Simulation is completed with a small number of stored states (M) and a small number of neurons (N) but a large M/N ratio. An experiment with optical exclusive-OR logic operation using LCTV SLMs shows the feasibility of optoelectronic implementation of the models. A complete inner-product TABAM is implemented using a PC for calculation of adaptive threshold values to achieve a unipolar TABAM (UIT) in the case where there is no crosstalk, and a crosstalk model (CRIT) in the case where crosstalk corrupts the desired state.
For the first time, a unipolar terminal-attractor based neural associative memory (TABAM)system with adaptive threshold and perfect convergence is presented. By adaptively setting thethreshold values for the dynamic iteration for the unipolar binary neuron states with terminal-attractors and inner-product approach, we demonstrate via computer simulation the achievement ofperfect convergence and correct retrieval. The simulation is completed with a small number of storedstates (M) and a small number of neurons (N) but a large M/N ratio. An experiment with exclusive-or logic operation using LCTV SLMs is used to show feasibility of the optoelectronic implementationof the models.
Single neuron models are fundamental for computational modeling of the brain's neuronal networks, and understanding how ion channel dynamics mediate neural function. A challenge in defining such models is determining biophysically realistic channel distributions. Here, we present an efficient, highly parallel evolutionary algorithm for developing such models, named NeuroGPU-EA. NeuroGPU-EA uses CPUs and GPUs concurrently to simulate and evaluate neuron membrane potentials with respect to multiple stimuli. We demonstrate a logarithmic cost for scaling the stimuli used in the fitting procedure. NeuroGPU-EA outperforms the typically used CPU based evolutionary algorithm by a factor of 10 on a series of scaling benchmarks. We report observed performance bottlenecks and propose mitigation strategies. Finally, we also discuss the potential of this method for efficient simulation and evaluation of electrophysiological waveforms.
This innovation is used to connect between synapse and neuron arrays using nanowire in quantum dot and metal in CMOS (complementary metal oxide semiconductor) technology to enable the density of a brain-like connection in hardware. The hardware implementation combines three technologies: 1. Quantum dot and nanowire-based compact synaptic cell (50x50 sq nm) with inherently low parasitic capacitance (hence, low dynamic power approx.l0(exp -11) watts/synapse), 2. Neuron and learning circuits implemented in 50-nm CMOS technology, to be integrated with quantum dot and nanowire synapse, and 3. 3D stacking approach to achieve the overall numbers of high density O(10(exp 12)) synapses and O(10(exp 8)) neurons in the overall system. In a 1-sq cm of quantum dot layer sitting on a 50-nm CMOS layer, innovators were able to pack a 10(exp 6)-neuron and 10(exp 10)-synapse array; however, the constraint for the connection scheme is that each neuron will receive a non-identical 10(exp 4)-synapse set, including itself, via its efficacy of the connection. This is not a fully connected system where the 100x100 synapse array only has a 100-input data bus and 100-output data bus. Due to the data bus sharing, it poses a great challenge to have a complete connected system, and its constraint within the quantum dot and silicon wafer layer. For an effective connection scheme, there are three conditions to be met: 1. Local connection. 2. The nanowire should be connected locally, not globally from which it helps to maximize the data flow by sharing the same wire space location. 3. Each synapse can have an alternate summation line if needed (this option is doable based on the simple mask creation). The 10(exp 3)x10(exp 3)-neuron array was partitioned into a 10-block, 10(exp 2)x10(exp 3)-neuron array. This building block can be completely mapped within itself (10,000 synapses to a neuron).
The ability of neural networks to provide ‘best in class’ approximation across a wide range of applications is well-documented. Nevertheless, the powerful expressivity of neural networks comes to naught if one is unable to effectively train (choose) the parameters defining the network. In general, neural networks are trained by gradient descent type optimization methods,a stochastic variant thereof. In practice, such methods result in the loss function decreases rapidly at the beginning of training but then, after a relatively small number of steps, significantly slow down. The loss may even appear to stagnate over the period of a large number of epochs, only to then suddenly start to decrease fast again for no apparent reason. This so-called plateau phenomenon manifests itself in many learning tasks. The present work aims to identify and quantify the root causes of plateau phenomenon.analysis is carried out in the setting of univariate ReLU networks. No assumptions are made on the number of neurons relative to the number of training data, and our results hold for both the lazy and adaptive regimes. Here, the main findings are: plateaux correspond to periods during which activation patterns remain constant, where activation pattern refers to the number of data points that activate a given neuron; quantification of convergence of the gradient flow dynamics; and, characterization stationary points in terms solutions of local least squares regression lines over subsets of the training data. Based on these conclusions, we propose a new iterative training method, the Active Neuron Least Squares (ANLS), characterised by the explicit adjustment of the activation pattern at each step, which is designed to enable a quick exit from a plateau. Illustrative numerical examples are included throughout.
This article is a review of work in three species, squirrel monkey, cat, and rat studying the inputs and outputs from the lateral vestibular nucleus (LVN). Different electrophysiological shock paradigms were used to determine the synaptic inputs derived from thick to thin diameter vestibular nerve afferents. Angular and linear mechanical stimulations were used to activate and study the combined and individual contribution of inner ear organs and neck afferents. The spatio-temporal properties of LVN neurons in the decerebrated rat were studied in response to dynamic acceleration inputs using sinusoidal linear translation in the horizontal head plane. Outputs were evaluated using antidromic identification techniques and identified LVN neurons were intracellularly injected with biocytin and their morphology studied.
Mathematical model of supervised learning by artificial neural network provides for simultaneous adjustments of both temperatures of neurons and synaptic weights, and includes feedback as well as feedforward synaptic connections. Extension of mathematical model described in "Adaptive Neurons For Artificial Neural Networks" (NPO-17803). Dynamics of neural network represented in new model by less-restrictive continuous formalism.
Anterograde labeling techniques were used to examine peripheral innervation patterns of vestibular efferent neurons in the crista ampullares of the gerbil. Vestibular efferent neurons were labeled by extracellular injections of biocytin or biotinylated dextran amine into the contralateral or ipsilateral dorsal subgroup of efferent cell bodies (group e) located dorsolateral to the facial nerve genu. Anterogradely labeled efferent terminal field varicosities consist mainly of boutons en passant with fewer of the terminal type. The bouton swellings are located predominately in apposition to the basolateral borders of the afferent calyces and type II hair cells, but several boutons were identified close to the hair cell apical border on both types. Three-dimensional reconstruction and morphological analysis of the terminal fields from these cells located in the sensory neuroepithelium of the anterior, horizontal, and posterior cristae were performed. We show that efferent neurons densely innervate each end organ in widespread terminal fields. Subepithelial bifurcations of parent axons were minimal, with extensive collateralization occurring after the axons penetrated the basement membrane of the neuroepithelium. Axonal branching ranged between the 6th and 27th orders and terminal field collecting area far exceeds that of the peripheral terminals of primary afferent neurons. The terminal fields of the efferent neurons display three morphologically heterogeneous types: central, peripheral, and planum. All cell types possess terminal fields displaying a high degree of anisotropy with orientations typically parallel to or within +/-45 degrees of the longitudinal axis if the crista. Terminal fields of the central and planum zones predominately project medially toward the transverse axis from the more laterally located penetration of the basement membrane by the parent axon. Peripheral zone terminal fields extend predominately toward the planum semilunatum. The innervation areas of efferent terminal fields display a trend from smallest to largest for the central, peripheral, and planum types, respectively. Neurons that innervate the central zone of the crista do not extend into the peripheral or planum regions. Conversely, those neurons with terminal fields in the peripheral or planum regions do not innervate the central zone of the sensory neuroepithelium. The central zone of the crista is innervated preferentially by efferent neurons with cell bodies located in the ipsilateral group e. The peripheral and planum zones of the crista are innervated preferentially by efferent neurons with cell bodies located in the contralateral group e. A model incorporating our anatomic observations is presented describing an ipsilateral closed-loop feedback between ipsilateral efferent neurons and the periphery and an open-loop feed-forward innervation from contralateral efferent neurons. A possible role for the vestibular efferent neurons in the modulation of semicircular canal afferent response dynamics is proposed.
The cerebellum is a neural structure that is essential for agility in vertebrate movements. Its contribution to motor control appears to be due to a fundamental role in dynamical state estimation, which also underlies its role in various non-motor tasks. Single spikes in vestibular sensory neurons carry information about head state. We show how computations for optimal dynamical state estimation may be accomplished when signals are encoded in spikes. This provides a novel way to design dynamical state estimators, and a novel way to interpret the structure and function of the cerebellum.
Here, we analyze the properties of a quantum system composed of two coherently coupled quantum oscillators and show through simulations that it fulfills the two properties required for reservoir computing: non-linearity and fading memory. We first show that the basis states of this system apply a set of nonlinear transformations on the input signals and thus can implement neurons. We then show that the system exhibits a fading memory that can be controlled by its dissipation rates. Finally we show that a strong coupling between the oscillators is important in order to ensure complex dynamics and to populate a number of basis state neurons that is exponential in the number of physical devices.
A method for the rapid learning of nonlinear mappings and topological transformations using a dynamically reconfigurable artificial neural network is presented. This fully-recurrent Adaptive Neuron Model (ANM) network was applied to the highly degenerate inverse kinematics problem in robotics, and its performance evaluation is bench-marked. Once trained, the resulting neuromorphic architecture was implemented in custom analog neural network hardware and the parameters capturing the functional transformation downloaded onto the system. This neuroprocessor, capable of 10(exp 9) ops/sec, was interfaced directly to a three degree of freedom Heathkit robotic manipulator. Calculation of the hardware feed-forward pass for this mapping was benchmarked at approximately 10 microsec.
Abstract Cervical spondylotic myelopathy (CSM) is a common cause of disability with few treatments. Aberrant mitochondrial dynamics play a crucial role in the pathogenesis of various neurodegenerative diseases. Thus, regulation of mitochondrial dynamics may offer therapeutic benefit for the treatment of CSM. Muscone, the active ingredient of an odoriferous animal product, exhibits anti‐inflammatory and neuroprotective effects for which the underlying mechanisms remain obscure. We hypothesized that muscone might ameliorate inflammatory responses and neuronal damage by regulating mitochondrial dynamics. To this end, the effects of muscone on a rat model of chronic cervical cord compression, as well as activated BV2 cells and injured neurons, were assessed. The results showed that muscone intervention improved motor function compared with vehicle‐treated rats. Indeed, muscone attenuated pro‐inflammatory cytokine expression, neuronal‐apoptosis indicators in the lesion area, and activation of the nod‐like receptor family pyrin domain‐containing 3 inflammasome, nuclear transcription factor‐κB, and dynamin‐related protein 1 in Iba1‐ and βIII‐tubulin‐labeled cells. Compared with vehicle‐treated rats, compression sites of muscone‐treated animals exhibited elongated mitochondrial morphologies in individual cell types and reduced reactive oxygen species. In vitro results indicated that muscone suppressed microglial activation and neuronal damage by regulating related‐inflammatory or apoptotic molecules. Moreover, muscone inhibited dynamin‐related protein 1 activation in activated BV2 cells and injured neurons, whereby it rescued mitochondrial fragmentation and reactive oxygen species production, which regulate a wide range of inflammatory and apoptotic molecules. Our findings reveal that muscone attenuates neuroinflammation and neuronal damage in rats with chronic cervical cord compression by regulating mitochondrial fission events, suggesting its promise for CSM therapy. image
Report presents theoretical study of dynamics of neural network organizing own response in both phase space and in position space. Postulates several mathematical models of dynamics including spatial derivatives representing local interconnections among neurons. Shows how neural responses propagate via these interconnections and how spatial pattern of neural responses formed in homogeneous biological neural network.
Neurodegenerative diseases (ND) have been linked to the critical process in aging-cellular senescence. However, the temporal dynamics of cellular senescence in ND conditions is unresolved. Here, we show senescence features develop in human Huntington's disease (HD) neural stem cells (NSCs) and medium spiny neurons (MSNs), including the increase of p16 INK4a , a key inducer of cellular senescence. We found that HD NSCs reprogram the transcriptional targets of FOXO3, a major cell survival factor able to repress cell senescence, antagonizing p16 INK4a expression via the FOXO3 repression of the transcriptional modulator ETS2. Additionally, p16 INK4a promotes cellular senescence features in human HD NSCs and MSNs. These findings suggest that cellular senescence may develop during neuronal differentiation in HD and that the FOXO3-ETS2-p16 INK4a axis may be part of molecular responses aimed at mitigating this phenomenon. Our studies identify neuronal differentiation with accelerated aging of neural progenitors and neurons as an alteration that could be linked to NDs.
Non-von Neumann computing using neuromorphic systems based on analogue synaptic and neuronal elements has emerged as a potential solution to tackle the growing need for more efficient data processing, but progress toward practical systems has been stymied due to a lack of materials and devices with the appropriate attributes. Recently, solid state electrochemical ion-insertion, also known as electrochemical random access memory (ECRAM) has emerged as a promising approach to realize the needed device characteristics. ECRAM is a three terminal device that operates by tuning electronic conductance in functional materials through solid-state electrochemical redox reactions. This mechanism can be considered as a gate-controlled bulk modulation of dopants and/or phases in the channel. Early work demonstrating that ECRAM can achieve nearly ideal analogue synaptic characteristics has sparked tremendous interest in this approach. More recently, the realization that electrochemical ion insertion can be used to tune the electronic properties of many types of materials including transition metal oxides, layered two-dimensional materials, organic and coordination polymers, and that the changes in conductance can span orders of magnitude has further attracted interest in ECRAM as the basis for analogue synaptic elements for inference accelerators as well as for dynamical devices that can emulate a wide range of neuronal characteristics for implementation in analogue spiking neural networks. At its core, ECRAM shares many fundamental aspects with rechargeable batteries, where ion insertion materials are used extensively for their ability to reversibly store charge and energy. Computing applications, however, present drastically different requirements: systems will require many millions of devices, scaled down to tens of nanometers, all while achieving reliable electronic-state tuning at scaled-up rates and endurances, and with minimal energy dissipation and noise. Further, in this review, we discuss the history, basic concepts, recent progress, as well as the challenges and opportunities for different types of ECRAM, broadly grouped by their primary mobile ionic charge carrier, including Li, protons, and oxygen vacancies.
The molecular mechanisms underlying neuronal dysfunction in Alzheimer’s disease (AD) remain uncharacterized. Here, we identify genes, molecular pathways and cellular components associated with whole-brain dysregulation caused by amyloid-beta (Aβ) and tau deposits in the living human brain. We obtained in-vivo resting-state functional MRI (rs-fMRI), Aβ- and tau-PET for 47 cognitively unimpaired and 16 AD participants from the Translational Biomarkers in Aging and Dementia cohort. Adverse neuronal activity impacts by Aβ and tau were quantified with personalized dynamical models by fitting pathology-mediated computational signals to the participant’s real rs-fMRIs. Then, we detected robust brain-wide associations between the spatial profiles of Aβ-tau impacts and gene expression in the neurotypical transcriptome (Allen Human Brain Atlas). Within the obtained distinctive signature of in-vivo neuronal dysfunction, several genes have prominent roles in microglial activation and in interactions with Aβ and tau. Moreover, cellular vulnerability estimations revealed strong association of microglial expression patterns with Aβ and tau’s synergistic impact on neuronal activity (q < 0.001). These results further support the central role of the immune system and neuroinflammatory pathways in AD pathogenesis. Neuronal dysregulation by AD pathologies also associated with neurotypical synaptic and developmental processes. In addition, we identified drug candidates from the vast LINCS library to halt or reduce the observed Aβ-tau effects on neuronal activity. Top-ranked pharmacological interventions target inflammatory, cancer and cardiovascular pathways, including specific medications undergoing clinical evaluation in AD. Our findings, based on the examination of molecular-pathological-functional interactions in humans, may accelerate the process of bringing effective therapies into clinical practice.