Population encoding of stimulus features along the visual hierarchy.

The retina and primary visual cortex (V1) both exhibit diverse neural populations sensitive to diverse visual features. Yet it remains unclear how neural populations in each area partition stimulus space to span these features. One possibility is that neural populations are organized into discrete groups of neurons, with each group signaling a particular constellation of features. Alternatively, neurons could be continuously distributed across feature-encoding space. To distinguish these possibilities, we presented a battery of visual stimuli to the mouse retina and V1 while measuring neural responses with multi-electrode arrays. Using machine learning approaches, we developed a manifold embedding technique that captures how neural populations partition feature space and how visual responses correlate with physiological and anatomical properties of individual neurons. We show that retinal populations discretely encode features, while V1 populations provide a more continuous representation. Applying the same analysis approach to convolutional neural networks that model visual processing, we demonstrate that they partition features much more similarly to the retina, indicating they are more like big retinas than little brains.

Optimal reaching subject to computational and physical constraints reveals structure of the sensorimotor control system.

Optimal feedback control provides an abstract framework describing the architecture of the sensorimotor system without prescribing implementation details such as what coordinate system to use, how feedback is incorporated, or how to accommodate changing task complexity. We investigate how such details are determined by computational and physical constraints by creating a model of the upper limb sensorimotor system in which all connection weights between neurons, feedback, and muscles are unknown. By optimizing these parameters with respect to an objective function, we find that the model exhibits a preference for an intrinsic (joint angle) coordinate representation of inputs and feedback and learns to calculate a weighted feedforward and feedback error. We further show that complex reaches around obstacles can be achieved by augmenting our model with a path-planner based on via points. The path-planner revealed "avoidance" neurons that encode directions to reach around obstacles and "placement" neurons that make fine-tuned adjustments to via point placement. Our results demonstrate the surprising capability of computationally constrained systems and highlight interesting characteristics of the sensorimotor system.

Chaotic neural dynamics facilitate probabilistic computations through sampling.

Cortical neurons exhibit highly variable responses over trials and time. Theoretical works posit that this variability arises potentially from chaotic network dynamics of recurrently connected neurons. Here, we demonstrate that chaotic neural dynamics, formed through synaptic learning, allow networks to perform sensory cue integration in a sampling-based implementation. We show that the emergent chaotic dynamics provide neural substrates for generating samples not only of a static variable but also of a dynamical trajectory, where generic recurrent networks acquire these abilities with a biologically plausible learning rule through trial and error. Furthermore, the networks generalize their experience in the stimulus-evoked samples to the inference without partial or all sensory information, which suggests a computational role of spontaneous activity as a representation of the priors as well as a tractable biological computation for marginal distributions. These findings suggest that chaotic neural dynamics may serve for the brain function as a Bayesian generative model.

Common population codes produce extremely nonlinear neural manifolds.

Populations of neurons represent sensory, motor, and cognitive variables via patterns of activity distributed across the population. The size of the population used to encode a variable is typically much greater than the dimension of the variable itself, and thus, the corresponding neural population activity occupies lower-dimensional subsets of the full set of possible activity states. Given population activity data with such lower-dimensional structure, a fundamental question asks how close the low-dimensional data lie to a linear subspace. The linearity or nonlinearity of the low-dimensional structure reflects important computational features of the encoding, such as robustness and generalizability. Moreover, identifying such linear structure underlies common data analysis methods such as Principal Component Analysis (PCA). Here, we show that for data drawn from many common population codes the resulting point clouds and manifolds are exceedingly nonlinear, with the dimension of the best-fitting linear subspace growing at least exponentially with the true dimension of the data. Consequently, linear methods like PCA fail dramatically at identifying the true underlying structure, even in the limit of arbitrarily many data points and no noise.

Nonquantal transmission at the vestibular hair cell-calyx synapse: KLV currents modulate fast electrical and slow K+ potentials.

Vestibular hair cells transmit information about head position and motion across synapses to primary afferent neurons. At some of these synapses, the afferent neuron envelopes the hair cell, forming an enlarged synaptic terminal called a calyx. The vestibular hair cell-calyx synapse supports a mysterious form of electrical transmission that does not involve gap junctions, termed nonquantal transmission (NQT). The NQT mechanism is thought to involve the flow of ions from the presynaptic hair cell to the postsynaptic calyx through low-voltage-activated channels driven by changes in cleft [K+] as K+ exits the hair cell. However, this hypothesis has not been tested with a quantitative model and the possible role of an electrical potential in the cleft has remained speculative. Here, we present a computational model that captures experimental observations of NQT and identifies features that support the existence of an electrical potential (ϕ) in the synaptic cleft. We show that changes in cleft ϕ reduce transmission latency and illustrate the relative contributions of both cleft [K+] and ϕ to the gain and phase of NQT. We further demonstrate that the magnitude and speed of NQT depend on calyx morphology and that increasing calyx height reduces action potential latency in the calyx afferent. These predictions are consistent with the idea that the calyx evolved to enhance NQT and speed up vestibular signals that drive neural circuits controlling gaze, balance, and orientation.

Scalable Gaussian process inference of neural responses to natural images.

Predicting the responses of sensory neurons is a long-standing neuroscience goal. However, while there has been much progress in modeling neural responses to simple and/or artificial stimuli, predicting responses to natural stimuli remains an ongoing challenge. On the one hand, deep neural networks perform very well on certain datasets but can fail when data are limited. On the other hand, Gaussian processes (GPs) perform well on limited data but are poor at predicting responses to high-dimensional stimuli, such as natural images. Here, we show how structured priors, e.g., for local and smooth receptive fields, can be used to scale up GPs to model neural responses to high-dimensional stimuli. With this addition, GPs largely outperform a deep neural network trained to predict retinal responses to natural images, with the largest differences observed when both models are trained on a small dataset. Further, since they allow us to quantify the uncertainty in their predictions, GPs are well suited to closed-loop experiments, where stimuli are chosen actively so as to collect "informative" neural data. We show how GPs can be used to actively select which stimuli to present, so as to i) efficiently learn a model of retinal responses to natural images, using few data, and ii) rapidly distinguish between competing models (e.g., a linear vs. a nonlinear model). In the future, our approach could be applied to other sensory areas, beyond the retina.

Intrinsic neural diversity quenches the dynamic volatility of neural networks.

Heterogeneity is the norm in biology. The brain is no different: Neuronal cell types are myriad, reflected through their cellular morphology, type, excitability, connectivity motifs, and ion channel distributions. While this biophysical diversity enriches neural systems' dynamical repertoire, it remains challenging to reconcile with the robustness and persistence of brain function over time (resilience). To better understand the relationship between excitability heterogeneity (variability in excitability within a population of neurons) and resilience, we analyzed both analytically and numerically a nonlinear sparse neural network with balanced excitatory and inhibitory connections evolving over long time scales. Homogeneous networks demonstrated increases in excitability, and strong firing rate correlations-signs of instability-in response to a slowly varying modulatory fluctuation. Excitability heterogeneity tuned network stability in a context-dependent way by restraining responses to modulatory challenges and limiting firing rate correlations, while enriching dynamics during states of low modulatory drive. Excitability heterogeneity was found to implement a homeostatic control mechanism enhancing network resilience to changes in population size, connection probability, strength and variability of synaptic weights, by quenching the volatility (i.e., its susceptibility to critical transitions) of its dynamics. Together, these results highlight the fundamental role played by cell-to-cell heterogeneity in the robustness of brain function in the face of change.

Topological insights into the neural basis of flexible behavior.

It is widely accepted that there is an inextricable link between neural computations, biological mechanisms, and behavior, but it is challenging to simultaneously relate all three. Here, we show that topological data analysis (TDA) provides an important bridge between these approaches to studying how brains mediate behavior. We demonstrate that cognitive processes change the topological description of the shared activity of populations of visual neurons. These topological changes constrain and distinguish between competing mechanistic models, are connected to subjects' performance on a visual change detection task, and, via a link with network control theory, reveal a tradeoff between improving sensitivity to subtle visual stimulus changes and increasing the chance that the subject will stray off task. These connections provide a blueprint for using TDA to uncover the biological and computational mechanisms by which cognition affects behavior in health and disease.

Invariant odor recognition with ON-OFF neural ensembles.

Invariant stimulus recognition is a challenging pattern-recognition problem that must be dealt with by all sensory systems. Since neural responses evoked by a stimulus are perturbed in a multitude of ways, how can this computational capability be achieved? We examine this issue in the locust olfactory system. We find that locusts trained in an appetitive-conditioning assay robustly recognize the trained odorant independent of variations in stimulus durations, dynamics, or history, or changes in background and ambient conditions. However, individual- and population-level neural responses vary unpredictably with many of these variations. Our results indicate that linear statistical decoding schemes, which assign positive weights to ON neurons and negative weights to OFF neurons, resolve this apparent confound between neural variability and behavioral stability. Furthermore, simplification of the decoder using only ternary weights ({+1, 0, -1}) (i.e., an "ON-minus-OFF" approach) does not compromise performance, thereby striking a fine balance between simplicity and robustness.

A computational neural model that incorporates both intrinsic dynamics and sensory feedback in the Aplysia feeding network.

Studying the nervous system underlying animal motor control can shed light on how animals can adapt flexibly to a changing environment. We focus on the neural basis of feeding control in Aplysia californica. Using the Synthetic Nervous System framework, we developed a model of Aplysia feeding neural circuitry that balances neurophysiological plausibility and computational complexity. The circuitry includes neurons, synapses, and feedback pathways identified in existing literature. We organized the neurons into three layers and five subnetworks according to their functional roles. Simulation results demonstrate that the circuitry model can capture the intrinsic dynamics at neuronal and network levels. When combined with a simplified peripheral biomechanical model, it is sufficient to mediate three animal-like feeding behaviors (biting, swallowing, and rejection). The kinematic, dynamic, and neural responses of the model also share similar features with animal data. These results emphasize the functional roles of sensory feedback during feeding.

Age-dependent increased sag amplitude in human pyramidal neurons dampens baseline cortical activity.

Aging involves various neurobiological changes, although their effect on brain function in humans remains poorly understood. The growing availability of human neuronal and circuit data provides opportunities for uncovering age-dependent changes of brain networks and for constraining models to predict consequences on brain activity. Here we found increased sag voltage amplitude in human middle temporal gyrus layer 5 pyramidal neurons from older subjects and captured this effect in biophysical models of younger and older pyramidal neurons. We used these models to simulate detailed layer 5 microcircuits and found lower baseline firing in older pyramidal neuron microcircuits, with minimal effect on response. We then validated the predicted reduced baseline firing using extracellular multielectrode recordings from human brain slices of different ages. Our results thus report changes in human pyramidal neuron input integration properties and provide fundamental insights into the neuronal mechanisms of altered cortical excitability and resting-state activity in human aging.

Modular origins of high-amplitude cofluctuations in fine-scale functional connectivity dynamics.

The topology of structural brain networks shapes brain dynamics, including the correlation structure of brain activity (functional connectivity) as estimated from functional neuroimaging data. Empirical studies have shown that functional connectivity fluctuates over time, exhibiting patterns that vary in the spatial arrangement of correlations among segregated functional systems. Recently, an exact decomposition of functional connectivity into frame-wise contributions has revealed fine-scale dynamics that are punctuated by brief and intermittent episodes (events) of high-amplitude cofluctuations involving large sets of brain regions. Their origin is currently unclear. Here, we demonstrate that similar episodes readily appear in silico using computational simulations of whole-brain dynamics. As in empirical data, simulated events contribute disproportionately to long-time functional connectivity, involve recurrence of patterned cofluctuations, and can be clustered into distinct families. Importantly, comparison of event-related patterns of cofluctuations to underlying patterns of structural connectivity reveals that modular organization present in the coupling matrix shapes patterns of event-related cofluctuations. Our work suggests that brief, intermittent events in functional dynamics are partly shaped by modular organization of structural connectivity.

The neural architecture of language: Integrative modeling converges on predictive processing.

The neuroscience of perception has recently been revolutionized with an integrative modeling approach in which computation, brain function, and behavior are linked across many datasets and many computational models. By revealing trends across models, this approach yields novel insights into cognitive and neural mechanisms in the target domain. We here present a systematic study taking this approach to higher-level cognition: human language processing, our species' signature cognitive skill. We find that the most powerful "transformer" models predict nearly 100% of explainable variance in neural responses to sentences and generalize across different datasets and imaging modalities (functional MRI and electrocorticography). Models' neural fits ("brain score") and fits to behavioral responses are both strongly correlated with model accuracy on the next-word prediction task (but not other language tasks). Model architecture appears to substantially contribute to neural fit. These results provide computationally explicit evidence that predictive processing fundamentally shapes the language comprehension mechanisms in the human brain.

An ecologically motivated image dataset for deep learning yields better models of human vision.

Deep neural networks provide the current best models of visual information processing in the primate brain. Drawing on work from computer vision, the most commonly used networks are pretrained on data from the ImageNet Large Scale Visual Recognition Challenge. This dataset comprises images from 1,000 categories, selected to provide a challenging testbed for automated visual object recognition systems. Moving beyond this common practice, we here introduce ecoset, a collection of >1.5 million images from 565 basic-level categories selected to better capture the distribution of objects relevant to humans. Ecoset categories were chosen to be both frequent in linguistic usage and concrete, thereby mirroring important physical objects in the world. We test the effects of training on this ecologically more valid dataset using multiple instances of two neural network architectures: AlexNet and vNet, a novel architecture designed to mimic the progressive increase in receptive field sizes along the human ventral stream. We show that training on ecoset leads to significant improvements in predicting representations in human higher-level visual cortex and perceptual judgments, surpassing the previous state of the art. Significant and highly consistent benefits are demonstrated for both architectures on two separate functional magnetic resonance imaging (fMRI) datasets and behavioral data, jointly covering responses to 1,292 visual stimuli from a wide variety of object categories. These results suggest that computational visual neuroscience may take better advantage of the deep learning framework by using image sets that reflect the human perceptual and cognitive experience. Ecoset and trained network models are openly available to the research community.

NSF DARE-transforming modeling in neurorehabilitation: perspectives and opportunities from US funding agencies.

In recognition of the importance and timeliness of computational models for accelerating progress in neurorehabilitation, the U.S. National Science Foundation (NSF) and the National Institutes of Health (NIH) sponsored a conference in March 2023 at the University of Southern California that drew global participation from engineers, scientists, clinicians, and trainees. This commentary highlights promising applications of computational models to understand neurorehabilitation ("Using computational models to understand complex mechanisms in neurorehabilitation" section), improve rehabilitation care in the context of digital twin frameworks ("Using computational models to improve delivery and implementation of rehabilitation care" section), and empower future interdisciplinary workforces to deliver higher-quality clinical care using computational models ("Using computational models in neurorehabilitation requires an interdisciplinary workforce" section). The authors describe near-term gaps and opportunities, all of which encourage interdisciplinary team science. Four major opportunities were identified including (1) deciphering the relationship between engineering figures of merit-a term commonly used by engineers to objectively quantify the performance of a device, system, method, or material relative to existing state of the art-and clinical outcome measures, (2) validating computational models from engineering and patient perspectives, (3) creating and curating datasets that are made publicly accessible, and (4) developing new transdisciplinary frameworks, theories, and models that incorporate the complexities of the nervous and musculoskeletal systems. This commentary summarizes U.S. funding opportunities by two Federal agencies that support computational research in neurorehabilitation. The NSF has funding programs that support high-risk/high-reward research proposals on computational methods in neurorehabilitation informed by theory- and data-driven approaches. The NIH supports the development of new interventions and therapies for a wide range of nervous system injuries and impairments informed by the field of computational modeling. The conference materials can be found at https://dare2023.usc.edu/ .

NSF DARE-Transforming modeling in neurorehabilitation: Four threads for catalyzing progress.

We present an overview of the Conference on Transformative Opportunities for Modeling in Neurorehabilitation held in March 2023. It was supported by the Disability and Rehabilitation Engineering (DARE) program from the National Science Foundation's Engineering Biology and Health Cluster. The conference brought together experts and trainees from around the world to discuss critical questions, challenges, and opportunities at the intersection of computational modeling and neurorehabilitation to understand, optimize, and improve clinical translation of neurorehabilitation. We organized the conference around four key, relevant, and promising Focus Areas for modeling: Adaptation & Plasticity, Personalization, Human-Device Interactions, and Modeling 'In-the-Wild'. We identified four common threads across the Focus Areas that, if addressed, can catalyze progress in the short, medium, and long terms. These were: (i) the need to capture and curate appropriate and useful data necessary to develop, validate, and deploy useful computational models (ii) the need to create multi-scale models that span the personalization spectrum from individuals to populations, and from cellular to behavioral levels (iii) the need for algorithms that extract as much information from available data, while requiring as little data as possible from each client (iv) the insistence on leveraging readily available sensors and data systems to push model-driven treatments from the lab, and into the clinic, home, workplace, and community. The conference archive can be found at (dare2023.usc.edu). These topics are also extended by three perspective papers prepared by trainees and junior faculty, clinician researchers, and federal funding agency representatives who attended the conference.

Offline memory replay in recurrent neuronal networks emerges from constraints on online dynamics.

During spatial exploration, neural circuits in the hippocampus store memories of sequences of sensory events encountered in the environment. When sensory information is absent during 'offline' resting periods, brief neuronal population bursts can 'replay' sequences of activity that resemble bouts of sensory experience. These sequences can occur in either forward or reverse order, and can even include spatial trajectories that have not been experienced, but are consistent with the topology of the environment. The neural circuit mechanisms underlying this variable and flexible sequence generation are unknown. Here we demonstrate in a recurrent spiking network model of hippocampal area CA3 that experimental constraints on network dynamics such as population sparsity, stimulus selectivity, rhythmicity and spike rate adaptation, as well as associative synaptic connectivity, enable additional emergent properties, including variable offline memory replay. In an online stimulus-driven state, we observed the emergence of neuronal sequences that swept from representations of past to future stimuli on the timescale of the theta rhythm. In an offline state driven only by noise, the network generated both forward and reverse neuronal sequences, and recapitulated the experimental observation that offline memory replay events tend to include salient locations like the site of a reward. These results demonstrate that biological constraints on the dynamics of recurrent neural circuits are sufficient to enable memories of sensory events stored in the strengths of synaptic connections to be flexibly read out during rest and sleep, which is thought to be important for memory consolidation and planning of future behaviour. KEY POINTS: A recurrent spiking network model of hippocampal area CA3 was optimized to recapitulate experimentally observed network dynamics during simulated spatial exploration. During simulated offline rest, the network exhibited the emergent property of generating flexible forward, reverse and mixed direction memory replay events. Network perturbations and analysis of model diversity and degeneracy identified associative synaptic connectivity and key features of network dynamics as important for offline sequence generation. Network simulations demonstrate that population over-representation of salient positions like the site of reward results in biased memory replay.

Benchmarking the generalizability of brain age models: Challenges posed by scanner variance and prediction bias.

Machine learning has been increasingly applied to neuroimaging data to predict age, deriving a personalized biomarker with potential clinical applications. The scientific and clinical value of these models depends on their applicability to independently acquired scans from diverse sources. Accordingly, we evaluated the generalizability of two brain age models that were trained across the lifespan by applying them to three distinct early-life samples with participants aged 8-22 years. These models were chosen based on the size and diversity of their training data, but they also differed greatly in their processing methods and predictive algorithms. Specifically, one brain age model was built by applying gradient tree boosting (GTB) to extracted features of cortical thickness, surface area, and brain volume. The other model applied a 2D convolutional neural network (DBN) to minimally preprocessed slices of T1-weighted scans. Additional model variants were created to understand how generalizability changed when each model was trained with data that became more similar to the test samples in terms of age and acquisition protocols. Our results illustrated numerous trade-offs. The GTB predictions were relatively more accurate overall and yielded more reliable predictions when applied to lower quality scans. In contrast, the DBN displayed the most utility in detecting associations between brain age gaps and cognitive functioning. Broadly speaking, the largest limitations affecting generalizability were acquisition protocol differences and biased brain age estimates. If such confounds could eventually be removed without post-hoc corrections, brain age predictions may have greater utility as personalized biomarkers of healthy aging.

Unpacking the multimodal, multi-scale data of the fast and slow lanes of the cardiac vagus through computational modelling.

NEW FINDINGS: What is the topic of this review? The vagus nerve is a crucial regulator of cardiovascular homeostasis, and its activity is linked to heart health. Vagal activity originates from two brainstem nuclei: the nucleus ambiguus (fast lane) and the dorsal motor nucleus of the vagus (slow lane), nicknamed for the time scales that they require to transmit signals. What advances does it highlight? Computational models are powerful tools for organizing multi-scale, multimodal data on the fast and slow lanes in a physiologically meaningful way. A strategy is laid out for how these models can guide experiments aimed at harnessing the cardiovascular health benefits of differential activation of the fast and slow lanes. ABSTRACT: The vagus nerve is a key mediator of brain-heart signaling, and its activity is necessary for cardiovascular health. Vagal outflow stems from the nucleus ambiguus, responsible primarily for fast, beat-to-beat regulation of heart rate and rhythm, and the dorsal motor nucleus of the vagus, responsible primarily for slow regulation of ventricular contractility. Due to the high-dimensional and multimodal nature of the anatomical, molecular and physiological data on neural regulation of cardiac function, data-derived mechanistic insights have proven elusive. Elucidating insights has been complicated further by the broad distribution of the data across heart, brain and peripheral nervous system circuits. Here we lay out an integrative framework based on computational modelling for combining these disparate and multi-scale data on the two vagal control lanes of the cardiovascular system. Newly available molecular-scale data, particularly single-cell transcriptomic analyses, have augmented our understanding of the heterogeneous neuronal states underlying vagally mediated fast and slow regulation of cardiac physiology. Cellular-scale computational models built from these data sets represent building blocks that can be combined using anatomical and neural circuit connectivity, neuronal electrophysiology, and organ/organismal-scale physiology data to create multi-system, multi-scale models that enable in silico exploration of the fast versus slow lane vagal stimulation. The insights from the computational modelling and analyses will guide new experimental questions on the mechanisms regulating the fast and slow lanes of the cardiac vagus toward exploiting targeted vagal neuromodulatory activity to promote cardiovascular health.

Optimization of data pre-processing methods for time-series classification of electroencephalography data.

The performance of time-series classification of electroencephalographic data varies strongly across experimental paradigms and study participants. Reasons are task-dependent differences in neuronal processing and seemingly random variations between subjects, amongst others. The effect of data pre-processing techniques to ameliorate these challenges is relatively little studied. Here, the influence of spatial filter optimization methods and non-linear data transformation on time-series classification performance is analyzed by the example of high-frequency somatosensory evoked responses. This is a model paradigm for the analysis of high-frequency electroencephalography data at a very low signal-to-noise ratio, which emphasizes the differences of the explored methods. For the utilized data, it was found that the individual signal-to-noise ratio explained up to 74% of the performance differences between subjects. While data pre-processing was shown to increase average time-series classification performance, it could not fully compensate the signal-to-noise ratio differences between the subjects. This study proposes an algorithm to prototype and benchmark pre-processing pipelines for a paradigm and data set at hand. Extreme learning machines, Random Forest, and Logistic Regression can be used quickly to compare a set of potentially suitable pipelines. For subsequent classification, however, machine learning models were shown to provide better accuracy.