SEARCH · Search NASA

Results for “Neurons”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

136 records · Page 8

All sources NASA NTRS DOE OSTI

DOE OSTI2025

Mapping Spiking Neural Networks to Heterogeneous Crossbar Architectures using Integer Linear Programming

Advances in novel hardware devices and architectures allow Spiking Neural Network (SNN) evaluation using ultra-low power, mixed-signal, memristor crossbar arrays. As individual network sizes quickly scale beyond the dimensional capabilities of single crossbars, networks must be mapped onto multiple crossbars. Crossbar sizes within modern Memristor Crossbar Architectures (MCAs) are determined predominately not by device technology but by network topology; more, smaller crossbars consume less area thanks to the high structural sparsity found in larger, brain-inspired SNNs. Motivated by continuing increases in SNN sparsity due to improvements in training methods, we propose utilizing heterogeneous crossbar sizes to further reduce area consumption. This approach was previously unachievable as prior compiler studies only explored solutions targeting homogeneous MCAs. Our work improves on the state-of-the-art by providing Integer Linear Programming (ILP) formulations supporting arbitrarily heterogeneous architectures. By modeling axonal interactions between neurons, our methods produce better mappings while removing inhibitive a priori knowledge requirements. We first show a 16.7-27.6% reduction in area consumption for square-crossbar homogeneous architectures. Then, we demonstrate 66.9-72.7% further reduction when using a reasonable configuration of heterogeneous crossbar dimensions. Next, we present a new optimization formulation capable of minimizing the number of inter-crossbar routes. When applied to solutions already near-optimal in area, an 11.9-26.4% routing reduction is observed without impacting area consumption. Finally, we present a profile-guided optimization capable of minimizing the number of runtime spikes between crossbars. Compared to the best-area-then-route optimized solutions, we observe a further 0.5-14.8% inter-crossbar spike reduction while requiring 1–3 orders of magnitude less solver time.

Pohl, Devin [ORNL] (ORCID:0009000040149027)↗

DOE OSTI2024

Waveguides for neurostimulation in the cochlea

Introduction Optical stimulation has been suggested for neural stimulation to improve cochlear implants. Light allows for more spatially selective activation of neuron populations than electrical current, offering more independent frequency bands along the spiral ganglion. These bands are available to encode acoustic information with anticipated better frequency resolution, improving cochlear implant user performance in noisy listening environments, tonal languages, and music perception. Methods Optical cochlear implants (oCIs) can deliver light either directly via small emitters within the cochlea or via waveguides from external optical sources. We investigated three waveguide designs made from OrmoComp ® , a polymer that cures through ultraviolet (UV) radiation. Waveguides were fabricated via injection molding and coated using dip-coating or thermal reflow, or through aspiration of OrmoComp ® into polyimide tubing that served as the cladding of the waveguide. The choice of fabrication technique directly determined the waveguides' total diameter: thermal reflow yielded ≈940 μm, dip-coating produced ≈306 μm, and aspiration resulted in ≈132 μm core diameter waveguides. Given the human cochlea's small size, we focused on analyzing the 306-μm and 132-μm waveguides, evaluating their optical performance (propagation and bending losses) and mechanical properties (bending stiffness and insertion forces). Furthermore, we evaluated some of these designs in in-vivo guinea pigs experiments. Results For the 100-μm core diameter waveguides, the propagation losses were 12.34 ± 1.26, 1.18 ± 0.88, 1.49 ± 0.58, and 3.43 ± 0.68 dB/cm at 534, 1,375, 1,460, and 1,550 nm, respectively. The respective bending losses at a 2 mm radius of curvature were 5.50 ± 1.32, 0.56 ± 0.26, 0.79 ± 0.18, and 0.64 ± 0.23 dB, and at 1 mm 8.54 ± 1.30, 2.05 ± 0.84, 2.11 ± 0.50, and 1.44 ± 0.37 dB. The bending stiffness of a 1 mm segment of the 100-μm-diameter waveguides was 18.9 ± 2.2 N/m. Insertion forces for the 100-μm-diameter waveguides into an acrylic human-size scala tympani model were < 25 mN. For the waveguides with 306 and 940 μm total diameter, the propagation losses ranged between 0.43 and 2.40 dB/cm at 534, 680, 1,375, and 1,550 nm, between 2.19 and 3.78 dB/cm at 450 and 1,460 nm. Bending losses for 360 degrees at 1,375 nm were 5.0, 2.4, and 0.46 for a bending radius of 2.5-, 3-, and 4-mm. Discussion Our study demonstrated that the polymer OrmoComp ® is suitable for fabricating waveguides to transmit near-infrared radiation. In-vivo experiments showed optically evoked auditory responses originating from optical stimulation in the guinea pigs' first cochlear turn while radiation was delivered. Incoming experiments will focus on evaluating long-term performance of these waveguides in guinea pigs and cats. This study will be designed to provide insights into the waveguides' performance and biocompatibility over extended periods, essential for their potential clinical application in future oCIs.

Kim, Matthew↗

DOE OSTI2025

Foveal vision reduces neural resources in agent-based game learning

Efficient processing of information is crucial for the optimization of neural resources in both biological and artificial visual systems. In this paper, we study the efficiency that may be obtained via the use of a fovea. Using biologically-motivated agents, we study visual information processing, learning, and decision making in a controlled artificial environment, namely the Atari Pong video game. We compare the resources necessary to play Pong between agents with and without a fovea. Our study shows that a fovea can significantly reduce the neural resources, in the form of number of neurons, number of synapses, and number of computations, while at the same time maintaining performance at playing Pong. To our knowledge, this is the first study in which an agent must simultaneously optimize its visual system, along with its decision making and action generation capabilities. That is, the visual system is integral to a complete agent.

60 APPLIED LIFE SCIENCES↗

DOE OSTI2025

The key role of the ferroptosis mechanism in neurological diseases and prospects for targeted therapy

Neurological disorders represent a major global health concern owing to their intricate pathological processes. Ferroptosis, defined as a form of cell death that is reliant on iron, has been closely linked to various neurological conditions. The fundamental process underlying ferroptosis is defined by the excessive buildup of iron ions, which initiates lipid peroxidation processes leading to cellular demise. Neurons, as highly metabolically active cells, are susceptible to oxidative stress, and imbalances in iron metabolism can directly initiate the ferroptosis process. In neurodegenerative disorders like Alzheimer’s disease and Parkinson’s disease, ferroptosis driven by iron accumulation represents a fundamental pathological connection. Although the connection between ferroptosis and neurological diseases is clear, clinical application still faces challenges, such as precise regulation of iron metabolism, development of specific drugs, and assessment of efficacy. The limited comprehension of the ferroptosis mechanism hinders the development of personalized treatment approaches. Consequently, subsequent investigations must tackle these obstacles to facilitate the clinical application of ferroptosis-associated therapies in neurological disorders. This article provides a comprehensive overview of the most recent advancements regarding the underlying mechanisms of ferroptosis. Subsequently, the study investigates the mechanistic contributions of ferroptosis within the nervous system. In conclusion, we evaluate and deliberate on targeted therapeutic strategies associated with ferroptosis and neurological disorders.

Xie, Chenyu↗

DOE OSTI2025

Composite Behavior of Nanopore Array Large Memristors

Synthetic nanopores were recently demonstrated with memristive and nonlinear voltage-current behaviors, akin to ion channels in a cell membrane. Such ionic devices are considered a promising candidate for the development of brain-inspired neuromorphic computing techniques. In this work, we show the composite behavior of nanopore-array large memristors, formed with different membrane materials, pore sizes, electrolytes, and device arrangements. Anodic aluminum oxide (AAO) membranes with 5 nm and 20 nm diameter pores and track-etched polycarbonate (PCTE) membranes with 10 nm diameter pores are tested and shown to demonstrate memristive and nonlinear behaviors with approximately 107–1010 pores in parallel when electrolyte concentration across the membranes is asymmetric. Ion diffusion through the large number of channels induces time-dependent electrolyte asymmetry that drives the system through different memristive states. The behaviors of series composite memristors with different configurations are also presented. In addition to helping understand fluidic devices and circuits for neuromorphic computing, the results also shed light on the development of field-assisted ion-selection-membrane filtration techniques as well as the investigations of large neurons and giant synapses. Further work is needed to de-embed parasitic components of the measurement setup to obtain intrinsic large memristor properties.

Chemistry↗

DOE OSTI2025

Astrocyte FABP7 Modulates Seizure Activity-Dependent Protein Expression in Mouse Brain

Background/Objectives: Patients with epilepsy commonly experience patterns of seizures that change with sleep/wake behavior or diurnal rhythms. The cellular and molecular mechanisms that underlie these patterns in seizure activity are not well understood but may involve non-neuronal cells, such as astrocytes. Our previous studies show the critical importance of one specific astrocyte factor, the brain-type fatty acid binding protein Fabp7, in the regulation of time-of-day-dependent electroshock seizure threshold and neural activity-dependent gene expression in mice. Here, we examined whether Fabp7 influences differential seizure activity-dependent protein expression, by comparing Fabp7 knockout (KO) to wild-type (WT) mice under control conditions and after reaching the maximal electroshock seizure threshold (MEST). Methods: We analyzed the proteome in cortical–hippocampal extracts from MEST and SHAM groups of WT and KO mice using mass spectrometry (MS), followed by Gene Ontology (GO) and pathway analyses. GO and pathway analyses of all groups revealed a diverse set of up- and downregulated differentially expressed proteins (DEPs). Results: We identified 65 significant DEPs in the comparison of KO SHAM versus WT SHAM; 33 proteins were upregulated and 32 were downregulated. We found downregulation in mitochondrial-associated proteins in WT MEST compared to WT SHAM controls, including Slc1a4, Slc25a27, Cox7a2, Cox8a, Micos10, and Atp5mk. Several upregulated DEPs in the KO SHAM versus WT SHAM comparison were associated with the 20S proteasomal subunit, suggesting proteasomal activity is elevated in the absence of Fabp7 expression. We also observed 92 DEPs significantly altered in the KO MEST versus WT MEST, with 49 proteins upregulated and 43 downregulated. Conclusions: Together, these data suggest that the astrocyte Fabp7 regulation of time-of-day-mediated neural excitability is modulated by multiple cellular mechanisms, which include proteasomal pathways, independent of its role in activity-dependent gene expression.

Neural Excitability↗

DOE OSTI2025

Machine Learning-Based Process Control for Injection Molding of Recycled Polypropylene

The increased interest in artificial intelligence in manufacturing has driven the adoption of machine learning to optimize processes and improve efficiency. A key challenge in injection molding is the variability of recycled materials, which affects part quality and processing stability. This study presents a novel closed-loop process control approach for injection molding, leveraging machine learning to adaptively predict processing inputs and quality outcomes. The methodology was tested on five blends of recycled polypropylene (rPP), using artificial neural networks (ANNs), linear regression, and polynomial regression to model the relationships between material properties and process parameters. The dataset was split 80/20 into training and testing sets. The ANN model was implemented using TensorFlow and Keras, with six hidden layers of 32 neurons per layer, ReLU activation, and an Adam optimizer. Empirical tuning and early stopping were used to optimize performance and prevent overfitting. Predictions were evaluated based on mean absolute error (MAE), mean squared error (MSE), and percentage error. The results showed that yield stress, ultimate elongation, and part weight were accurately predicted within a 5% error for linear and polynomial regression models and within a 10% error for the ANN. However, modulus predictions were less reliable, with errors of ~11% for ANN and linear regression and ~40% for polynomial regression, reflecting the inherent variability of this property in rPP blends. Predictions of processing inputs had errors ranging from 3% to 25%, depending on the model and response variable. No single modeling approach was consistently superior across all responses, highlighting the complexity of the relationship between material properties, process parameters, and quality metrics. Overall, the work demonstrates that closed-loop process control, powered by machine learning, can effectively predict key quality parameters in injection molding of recycled materials. The proposed approach can improve process stability and material utilization, facilitating increased adoption of sustainable materials.

Krantz, Joshua↗

DOE OSTI2024

Convolutional double copy in (anti) de Sitter space

The double copy is a remarkable relationship between gauge theory and gravity that has been explored in a number of contexts, most notably scattering amplitudes and classical solutions. The convolutional double copy provides a straightforward method to bridge the two theories via a precise map for the fields and symmetries at the linearised level. This method has been thoroughly investigated in flat space, offering a comprehensive dictionary both with and without fixing the gauge degrees of freedom. In this paper, we extend this to curved space with an (anti) de Sitter background metric. We work in the temporal gauge, and employ a modified convolution that involves the Mellin transformation in the time direction. As an example, we show that the point-like charge in gauge theory double copies to the (dS-) Schwarzschild black hole solution.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

DOE OSTI2025

Shadow masks predictions in SPARC tokamak plasma-facing components using HEAT code and machine learning methods

Here, this work uses machine learning (ML) to complement HEAT (Heat flux Engineering Analysis Toolkit) by developing 3-D footprint surrogate models for fast and accurate heat load calculations in the divertor of the SPARC tokamak. The focus is on shadowed regions, or magnetic shadows, caused by the 3-D geometry of plasma-facing components (PFCs). ML classifiers are employed to create a surrogate model for HEAT generated shadow masks, predicting these shadow masks and divertor heat flux profiles based on a diverse range of equilibria and only the plasma current, safety factor(q95) at the edge, and magnetic flux angles as input parameters. The ultimate goal is to integrate the model for real-time control and future operational decisions.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2024

Overview of RFID Applications Utilizing Neural Networks

As Radio Frequency Identification (RFID) methods continue to evolve to higher levels of complexity, one form of machine learning is making its appearance. The use of Neural Networks (NN) in the RFID field is steadily increasing, and in the fields of localization and activity recognition, promising results are being shown from a variety of research. RFID applications fall primarily under two types of problems including regression and classification. We analyze RIFD localization techniques which fall under regression, and activity recognition which falls under classification. Many works don’t classify themselves as activity recognition methods, but because they fall under the classification category, we still consider them as activity recognition techniques. This research overviews the Neural Network models in the localization field based on whether they can perform independently of the environment in which they were tested. For activity recognition and accessory fields, the major methods involve tag-based and tag-free approaches. In conclusion, after the models are surveyed, a comparison study is given to examine what may be the cause for increased accuracy between different Neural Network models.

42 ENGINEERING↗